University of Groningen
The consequences of aneuploidy and chromosome instability
Schukken, Klaske Marijke
DOI:
10.33612/diss.135392967
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2020
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Schukken, K. M. (2020). The consequences of aneuploidy and chromosome instability: Survival, cell death and cancer. University of Groningen. https://doi.org/10.33612/diss.135392967
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
2
Bioessays, 2018 Jan; 40(1). Doi: 10.1002/bies.201700147
Klaske M. Schukken1 and Floris Foijer1
1European Research Institute for the Biology of Ageing (ERIBA), University of Groningen, University Medical Center Groningen, 9713 AV,
Groningen, the Netherlands
CIN and aneuploidy: different concepts,
different consequences
Chapter 2
2
2
CIN and aneuploidy: different concepts, different consequences
Klaske M. Schukken1 and Floris Foijer1
1European Research Institute for the Biology of Ageing (ERIBA), University of
Groningen, University Medical Center Groningen, 9713 AV, Groningen, the Netherlands
Bioessays, 2018 Jan; 40(1). Doi: 10.1002/bies.201700147
Abstract
Chromosomal instability (CIN) and aneuploidy are similar concepts but not synonymous. CIN is the process that leads to chromosome copy number alterations, and aneuploidy is the result. While CIN and resulting aneuploidy often cause growth defects, they are also selected for in cancer cells. Although such contradicting fates may seem
paradoxical at first, they can be better understood when CIN and
aneuploidy are assessed separately, taking into account the in vitro or in
vivo context, the rate of CIN, and severity of the aneuploid karyotype. As
CIN can only be measured in living cells, which proves to be technically challenging in vivo, aneuploidy is more frequently quantified. However, CIN rates might be more predictive for tumor outcome than assessing aneuploidy rates alone. In reviewing the literature, we therefore conclude that there is an urgent need for new models in which we can monitor chromosome mis-segregation and its consequences in vivo. Keywords: Aneuploidy; CIN; cancer; aging; in vivo; in vitro
1. Introduction
Chromosomal instability (CIN) is a process in which cells suffer from chromosome segregation defects in mitosis, which leads to cells with chromosome copy number alterations and structural changes of the chromosomes, a state called aneuploidy. While aneuploidy typically causes a growth defect, it can also enhance cell growth and survival3. In
fact, more than two thirds of all human cancers are aneuploid8. The
reduced fitness that CIN and the resulting aneuploidy instigate on cells can trigger a reduced proliferation potential, apoptosis and/or
senescence. The consequences of CIN likely depend on the severity of the chromosomal abnormality, the cell’s microenvironment, and the cell type in which CIN took place16,39,40. While CIN and aneuploidy are used
interchangeably, it is important to understand the difference.
Aneuploidy is a genetic state, whereas CIN is the cellular behavior that leads to aneuploidy, see Figure 1 for further clarification. Furthermore, cells can be aneuploid without exhibiting CIN, in which case the
karyotypes do not change over time, thus resulting in cellular
populations that all display the exact same karyotype. One of the most well-known examples of aneuploidy without CIN is Down syndrome, caused by systemic trisomy of chromosome 21; these cells are aneuploid without an increased mis-segregation rate41. Additionally, stable
aneuploid cell lines have been engineered by introducing one or more extra chromosomes to investigate the consequences of aneuploidy in the absence of CIN42,43.
There are many methods to quantify aneuploidy44, but in order to
quantify CIN actual mis-segregation rates need to be measured in living cells, which is time consuming and challenging, especially in animal models. Therefore, in many cases, when CIN is induced in a cell population or tissue, only the resulting aneuploidy is quantified as a measure of the underlying CIN.
The objective of this review is to emphasize the difference between aneuploidy and CIN, and to discuss the differential responses to CIN in in
vitro and in vivo conditions. While at first this vast variability in cellular
fates may seem paradoxical, we will discuss some of the factors that cause these differences and ultimately argue why we need better models to measure CIN in an in vivo setting.
1
2
Abstract
Chromosomal instability (CIN) and aneuploidy are similar concepts but not synonymous. CIN is the process that leads to chromosome copy number alterations, and aneuploidy is the result. While CIN and resulting aneuploidy often cause growth defects, they are also selected for in cancer cells. Although such contradicting fates may seem
paradoxical at first, they can be better understood when CIN and
aneuploidy are assessed separately, taking into account the in vitro or in
vivo context, the rate of CIN, and severity of the aneuploid karyotype. As
CIN can only be measured in living cells, which proves to be technically challenging in vivo, aneuploidy is more frequently quantified. However, CIN rates might be more predictive for tumor outcome than assessing aneuploidy rates alone. In reviewing the literature, we therefore conclude that there is an urgent need for new models in which we can monitor chromosome mis-segregation and its consequences in vivo. Keywords: Aneuploidy; CIN; cancer; aging; in vivo; in vitro
1. Introduction
Chromosomal instability (CIN) is a process in which cells suffer from chromosome segregation defects in mitosis, which leads to cells with chromosome copy number alterations and structural changes of the chromosomes, a state called aneuploidy. While aneuploidy typically causes a growth defect, it can also enhance cell growth and survival3. In
fact, more than two thirds of all human cancers are aneuploid8. The
reduced fitness that CIN and the resulting aneuploidy instigate on cells can trigger a reduced proliferation potential, apoptosis and/or
senescence. The consequences of CIN likely depend on the severity of the chromosomal abnormality, the cell’s microenvironment, and the cell type in which CIN took place16,39,40. While CIN and aneuploidy are used
interchangeably, it is important to understand the difference.
Aneuploidy is a genetic state, whereas CIN is the cellular behavior that leads to aneuploidy, see Figure 1 for further clarification. Furthermore, cells can be aneuploid without exhibiting CIN, in which case the
karyotypes do not change over time, thus resulting in cellular
populations that all display the exact same karyotype. One of the most well-known examples of aneuploidy without CIN is Down syndrome, caused by systemic trisomy of chromosome 21; these cells are aneuploid without an increased mis-segregation rate41. Additionally, stable
aneuploid cell lines have been engineered by introducing one or more extra chromosomes to investigate the consequences of aneuploidy in the absence of CIN42,43.
There are many methods to quantify aneuploidy44, but in order to
quantify CIN actual mis-segregation rates need to be measured in living cells, which is time consuming and challenging, especially in animal models. Therefore, in many cases, when CIN is induced in a cell population or tissue, only the resulting aneuploidy is quantified as a measure of the underlying CIN.
The objective of this review is to emphasize the difference between aneuploidy and CIN, and to discuss the differential responses to CIN in in
vitro and in vivo conditions. While at first this vast variability in cellular
fates may seem paradoxical, we will discuss some of the factors that cause these differences and ultimately argue why we need better models to measure CIN in an in vivo setting.
To study the consequences of CIN, various model systems have been engineered in which CIN can be provoked, for instance via inhibition or inactivation of the spindle assembly checkpoint (SAC)45, by irradiating
cells, inducing centrosome amplification, or by interfering with cohesion attachments, kinetochore-spindle attachments, or the spindle/tubulin network4,40,45–47. These interventions lead to chromosome
mis-segregation events and other mitotic defects, further described in Box 1 and Figure 1c. The effects of CIN have been found to be strongly
dependent on the rate and type of chromosome mis-segregation or mitotic defect48–52. CIN is often measured in a tissue culture setting,
which we, for the purpose of this review, define as in vitro. CIN can be quantified via time-lapse microscopy or immunofluorescence staining of fixed cells in mitosis. Since CIN is a process, the only way to truly
measure it is to observe cells over time and record the mis-segregations that occur during mitosis. Immunohistochemistry can also be used to observe mitotic errors, but this method may underestimate the rates of CIN because it only takes a snapshot of mitosis.
CIN is most often measured in vitro, for instance when assessing the effects of overexpression or inhibition of proteins or the effects of drugs48–52. In vitro CIN measurements can also be used to characterize
animal models. For instance, some studies make use of mouse embryonic fibroblasts (MEFs) from mouse models of genetically induced CIN to estimate the in vivo rate of CIN. While MEFs have the same genotype as the mouse model, cell culture conditions may impact the CIN behavior and different cell types might respond differently to CIN in vitro as compared to in vivo (also see below, “the differing effects of CIN”).
Organoids are a novel in vitro method that not only model the genetic condition of the organism, but also allow for several cell types to interact in a 3D context that mimic the structure and function of in vivo organs, and thus more closely mimic in vivo conditions than cell lines grown in 2D53. Indeed, CIN has also been induced and observed in
organoids54, but so far, side by side comparisons of CIN rates between
2D and 3D cultures in an isogenic background are still lacking.
In vitro CIN measurements are useful because these models can be
imaged long-term, increasing the number of mitotic events to be tracked, and allowing for cell fate tracking following chromosome mis-segregation events. Furthermore, many cell types and genetic
interventions to induce CIN are available for in vitro studies.
Figure 1: The difference between CIN and aneuploidy (A-B) Cells displaying the copy
numbers of a single chromosome before, during and after mitosis. A) A cell exhibiting CIN starting out as an euploid cell, mis-segregating chromosomes to become aneuploid. B) A non-CIN, aneuploid cell undergoing faithful mitosis. C) Diagrams for various chromosome mis-segregations and other effects of CIN. | Chromosomes displayed in red, mitotic spindle displayed in green.
CIN CIN CIN Ploidy A) No CIN No CIN Mitosis: no mis-segregation CIN
Ploidy 3 copiesof chromosome XAneuploid 3 copies of chromosome XAneuploid No CIN B)
Lagging chromosome mis-alignmentChromosome Anaphase bridges C) Effect
of CIN Monopolar spindle
Multipolar spindle Micronuclei Effect
of CIN Multinucleated cells
Several effects of CIN are only quantifiable if observed over time. Other effects CIN Mitosis: missegregation of chromosome(s) Euploidy
1
2
To study the consequences of CIN, various model systems have been engineered in which CIN can be provoked, for instance via inhibition or inactivation of the spindle assembly checkpoint (SAC)45, by irradiating
cells, inducing centrosome amplification, or by interfering with cohesion attachments, kinetochore-spindle attachments, or the spindle/tubulin network4,40,45–47. These interventions lead to chromosome
mis-segregation events and other mitotic defects, further described in Box 1 and Figure 1c. The effects of CIN have been found to be strongly
dependent on the rate and type of chromosome mis-segregation or mitotic defect48–52. CIN is often measured in a tissue culture setting,
which we, for the purpose of this review, define as in vitro. CIN can be quantified via time-lapse microscopy or immunofluorescence staining of fixed cells in mitosis. Since CIN is a process, the only way to truly
measure it is to observe cells over time and record the mis-segregations that occur during mitosis. Immunohistochemistry can also be used to observe mitotic errors, but this method may underestimate the rates of CIN because it only takes a snapshot of mitosis.
CIN is most often measured in vitro, for instance when assessing the effects of overexpression or inhibition of proteins or the effects of drugs48–52. In vitro CIN measurements can also be used to characterize
animal models. For instance, some studies make use of mouse embryonic fibroblasts (MEFs) from mouse models of genetically induced CIN to estimate the in vivo rate of CIN. While MEFs have the same genotype as the mouse model, cell culture conditions may impact the CIN behavior and different cell types might respond differently to CIN in vitro as compared to in vivo (also see below, “the differing effects of CIN”).
Organoids are a novel in vitro method that not only model the genetic condition of the organism, but also allow for several cell types to interact in a 3D context that mimic the structure and function of in vivo organs, and thus more closely mimic in vivo conditions than cell lines grown in 2D53. Indeed, CIN has also been induced and observed in
organoids54, but so far, side by side comparisons of CIN rates between
2D and 3D cultures in an isogenic background are still lacking.
In vitro CIN measurements are useful because these models can be
imaged long-term, increasing the number of mitotic events to be tracked, and allowing for cell fate tracking following chromosome mis-segregation events. Furthermore, many cell types and genetic
interventions to induce CIN are available for in vitro studies.
Figure 1: The difference between CIN and aneuploidy (A-B) Cells displaying the copy
numbers of a single chromosome before, during and after mitosis. A) A cell exhibiting CIN starting out as an euploid cell, mis-segregating chromosomes to become aneuploid. B) A non-CIN, aneuploid cell undergoing faithful mitosis. C) Diagrams for various chromosome mis-segregations and other effects of CIN. | Chromosomes displayed in red, mitotic spindle displayed in green.
CIN CIN CIN Ploidy A) No CIN No CIN Mitosis: no mis-segregation CIN
Ploidy 3 copiesof chromosome XAneuploid 3 copies of chromosome XAneuploid No CIN B)
Lagging chromosome mis-alignmentChromosome Anaphase bridges C) Effect
of CIN Monopolar spindle
Multipolar spindle Micronuclei Effect
of CIN Multinucleated cells
Several effects of CIN are only quantifiable if observed over time. Other effects CIN Mitosis: missegregation of chromosome(s) Euploidy
Additionally, the cost, time and labor which goes into creating and testing in vitro models is much lower than creating animal models. Unfortunately, not all primary tissue types can be cultured easily, and cell culture conditions might not be fully representative of their in vivo counterparts, as for instance the immune, nervous and endocrine systems are absent in cell and organoid cultures.
Box 1:A list of chromosome mis-segregations and other mitotic defects 1. Lagging chromosomes can be observed during anaphase as
chromosomes that are separated later than most others, which can lead to mis-segregation events. Laggards can be caused by mutations in the SAC, abnormal sister chromatid cohesion, disturbed kinetochore-microtubule attachments (e.g. merotelic attachments) and centrosome replication defects such as multipolar spindles55. Laggards can lead to aneuploidy, and
micronuclei (see B. Micronuclei).
2. Chromosome mis‐alignment can yield polar chromosomes. Such mis-segregating chromosomes fail to align at the metaphase plate, and instead remain behind at one of the spindle poles. Chromosome mis-alignment can be caused by SAC defects56,57,
and can form micronuclei (see B. Micronuclei).
3. Anaphase bridges and Ultrafine Anaphase Bridges 58 occur
when a chromosome is stretched between two spindle poles during anaphase. The ultrafine bridges can be difficult to visualize using live-cell imaging, particularly when the stretched DNA is single stranded. Anaphase bridges can be caused by replication stress or stalling59, and can lead to chromosome breaks and
structural abnormalities.
4. Monopolar spindles occur when the spindle poles fail to separate. Monopolar spindles can be caused by poorly separated centrioles, tubulin network defects, or motor or centrosome protein mutations60.
5. Multipolar spindles occur when a mitotic cell forms more than two spindle poles. They are often caused by centrosome amplification61, additional centrosomes due to cytokinesis
failure/ tetraploidization, and/or loss of spindle pole integrity62.
Multipolar spindles can lead to more than 2 daughter cells, or daughter cell(s) containing multiple spindle poles63, polyploidy,
cell death and aneuploidy61, as well as lagging chromosomes and
incorrectly attached chromosomes (see 1. Lagging Chromosomes).
6. Cytokinesis failure is when a cell segregates its chromosomes but fails to produce two daughter cells. Cytokinesis is a multi-step, highly regulated process which can fail at multiple points64.
Cytokinesis failure can lead to cell death, polyploidy, and/or centrosome amplification, and multinucleated cells (see C. Multinucleated cells)65.
Furthermore, the following events can be the result of CIN, but are not direct measurements of chromosome mis-segregation:
A. A lack of a properly aligned metaphase plate (failure to align) and an extended or shortened mitosis are signs of CIN that can only be visualized by time-lapse microscopy. Errors in chromosome alignment will cause the SAC to delay anaphase until all chromosomes are properly aligned and attached to the spindle45. However, when the SAC is (partially) alleviated, time in
mitosis is reduced57. Sustained SAC activation, for instance due to
spindle poison exposure or increased Mad2 levels21, will block
cells in mitosis. When cells are stuck in mitosis for extended periods of time, they might undergo apoptosis66.
B. Micronuclei are not visible during mitosis, but they are the result of chromosome mis-segregation. Micronuclei are primarily caused by lagging chromosomes, multi-polar spindles and mis-aligned chromosomes67,68.
C. Multinucleated cells are typically the result of cytokinesis failure (also see 6. Cytokinesis failure) and can lead to chromosome mis-segregation events in the next mitosis.
\\Box1 \\
3. Can CIN be measured in vivo?
One way to estimate CIN rates in an in vivo setting is by quantifying chromosome copy number heterogeneity within a primary tumor, for instance by single-cell sequencing27. In this case, the heterogeneity
resulting from the mis-segregation events is used as a measure to estimate CIN. However, single-cell aneuploidy assessments are still end-point measurements, which are likely skewed by strong selection pressures for or against certain karyotypes, as well as by the length of time a cell population has been chromosomal instable. Furthermore,
1
2
Additionally, the cost, time and labor which goes into creating and testing in vitro models is much lower than creating animal models. Unfortunately, not all primary tissue types can be cultured easily, and cell culture conditions might not be fully representative of their in vivo counterparts, as for instance the immune, nervous and endocrine systems are absent in cell and organoid cultures.
Box 1:A list of chromosome mis-segregations and other mitotic defects 1. Lagging chromosomes can be observed during anaphase as
chromosomes that are separated later than most others, which can lead to mis-segregation events. Laggards can be caused by mutations in the SAC, abnormal sister chromatid cohesion, disturbed kinetochore-microtubule attachments (e.g. merotelic attachments) and centrosome replication defects such as multipolar spindles55. Laggards can lead to aneuploidy, and
micronuclei (see B. Micronuclei).
2. Chromosome mis‐alignment can yield polar chromosomes. Such mis-segregating chromosomes fail to align at the metaphase plate, and instead remain behind at one of the spindle poles. Chromosome mis-alignment can be caused by SAC defects56,57,
and can form micronuclei (see B. Micronuclei).
3. Anaphase bridges and Ultrafine Anaphase Bridges 58 occur
when a chromosome is stretched between two spindle poles during anaphase. The ultrafine bridges can be difficult to visualize using live-cell imaging, particularly when the stretched DNA is single stranded. Anaphase bridges can be caused by replication stress or stalling59, and can lead to chromosome breaks and
structural abnormalities.
4. Monopolar spindles occur when the spindle poles fail to separate. Monopolar spindles can be caused by poorly separated centrioles, tubulin network defects, or motor or centrosome protein mutations60.
5. Multipolar spindles occur when a mitotic cell forms more than two spindle poles. They are often caused by centrosome amplification61, additional centrosomes due to cytokinesis
failure/ tetraploidization, and/or loss of spindle pole integrity62.
Multipolar spindles can lead to more than 2 daughter cells, or daughter cell(s) containing multiple spindle poles63, polyploidy,
cell death and aneuploidy61, as well as lagging chromosomes and
incorrectly attached chromosomes (see 1. Lagging Chromosomes).
6. Cytokinesis failure is when a cell segregates its chromosomes but fails to produce two daughter cells. Cytokinesis is a multi-step, highly regulated process which can fail at multiple points64.
Cytokinesis failure can lead to cell death, polyploidy, and/or centrosome amplification, and multinucleated cells (see C. Multinucleated cells)65.
Furthermore, the following events can be the result of CIN, but are not direct measurements of chromosome mis-segregation:
A. A lack of a properly aligned metaphase plate (failure to align) and an extended or shortened mitosis are signs of CIN that can only be visualized by time-lapse microscopy. Errors in chromosome alignment will cause the SAC to delay anaphase until all chromosomes are properly aligned and attached to the spindle45. However, when the SAC is (partially) alleviated, time in
mitosis is reduced57. Sustained SAC activation, for instance due to
spindle poison exposure or increased Mad2 levels21, will block
cells in mitosis. When cells are stuck in mitosis for extended periods of time, they might undergo apoptosis66.
B. Micronuclei are not visible during mitosis, but they are the result of chromosome mis-segregation. Micronuclei are primarily caused by lagging chromosomes, multi-polar spindles and mis-aligned chromosomes67,68.
C. Multinucleated cells are typically the result of cytokinesis failure (also see 6. Cytokinesis failure) and can lead to chromosome mis-segregation events in the next mitosis.
\\Box1 \\
3. Can CIN be measured in vivo?
One way to estimate CIN rates in an in vivo setting is by quantifying chromosome copy number heterogeneity within a primary tumor, for instance by single-cell sequencing27. In this case, the heterogeneity
resulting from the mis-segregation events is used as a measure to estimate CIN. However, single-cell aneuploidy assessments are still end-point measurements, which are likely skewed by strong selection pressures for or against certain karyotypes, as well as by the length of time a cell population has been chromosomal instable. Furthermore,
heterogeneity measurements do not give information about the type of mis-segregation that has occurred, fail to detect cells that die as a result of the mis-segregation, and therefore do not provide ultimate proof that CIN is still ongoing.
Other methods that have been used to estimate CIN in vivo are the use of organoid cultures, MEFs and other primary cell cultures directly derived from in vivo models (also see above 2. Measuring CIN in vitro).
As explained above, the most reliable evidence for CIN in vivo is catching mis-segregating cells in the act, for instance using intra-vital imaging. However, intra-vital imaging at sufficient resolution to monitor mitosis is technically challenging and requires dedicated animal models in which mitotic components are fluorescently labeled. Unfortunately, only a few mouse models have been engineered for this purpose (e.g. mice expressing H2B-GFP)69. While mouse models have been employed to
monitor the rate of cell division70, they have not been used to monitor
chromosome (mis)segregation per se. Other studies have used intra-vital imaging to measure CIN in fluorescently labeled xenografted tumor cells71.
Measuring CIN in vivo would more accurately show the effect of CIN for instance during development, to assess the role of the immune system and inter-tissue interactions, and the effect of CIN in the context of an unperturbed tissue. However, drawbacks of in vivo CIN measurement are the limited time available for imaging, the high cost, technical difficulty of creating and imaging animal models, and the relatively low rate of cell division in vivo. Therefore in most cases, aneuploidy is measured in vivo and not CIN. However, as explained above, the level of aneuploidy within a cell population may not be representative of the level of CIN.
4. Does aneuploidy lead to CIN?
As explained earlier, CIN leads to aneuploidy, but does aneuploidy also lead to CIN? Additionally, does the level of aneuploidy in a cell
population automatically correlate with the level of CIN? It is still contested in the field whether aneuploidy itself induces CIN. Most studies concerning the relationship of aneuploidy with CIN look at single chromosome gains and focus of the direct effects of aneuploidy, leaving the effect of complex karyotypes and chromosome losses unstudied. For
this purpose, aneuploid cells have been engineered via microcell-mediated chromosome transfer (MMCT)72,73.
Indeed, several trisomies, tetrasomies and a few double-trisomies were shown to cause further mis-segregation events in both cancer and non-cancer cell lines51,52. A trisomy of chromosome 13 in human cells can
lead to failing cytokinesis, while a trisomy of chromosome 7 does not, showing that different aneuploid karyotypes can cause different forms of CIN51. However, in fully triploid fibroblasts (3 copies of each
chromosome) the percentage of aneuploid cells increases, but the rate of chromosome mis-segregation (CIN), relative to a euploid control, does not41. This can be explained by the fact that triploid cells that
encountered chromosome losses had a growth advantage relative to their fully-triploid counterparts, whereas chromosome losses were selected against in the diploid cells; this shows that the increased rate of aneuploidy was due to selection, not an increased CIN rate. Additionally, while aneuploidy can lead to increased CIN in the short term, whether this is a long-tern effect remains controversial due to relatively short-term CIN endpoint measurements. Therefore, the level of aneuploidy is not always an accurate representation of the level of CIN, nor does an aneuploid state always increase chromosome mis-segregation frequency.
Other reports suggest that aneuploidy does not necessarily increase the rate of CIN. For instance, patients with trisomies for chromosome 13, 18 or 21 were shown to not display significantly higher levels of
aneuploidies of other chromosomes compared to euploid fibroblasts in
vitro41. Together, these findings suggest that some aneuploidies can
provoke further CIN, but not in all settings.
5. Aneuploidy can have highly differential effects on cells. 5.1. What are the effects of aneuploidy in vitro?
In general, aneuploidy has been shown to be detrimental to cell growth2,42,43,52. This growth defect is caused by the deregulation of the
proteasome3,74–77, cell metabolism78, transcriptome73, and the increase in
cellular senescence2,79. However, not all aneuploidies have the same
effect on cells. For example, a trisomy of chromosome 1 confers a much larger growth defect than a trisomy of the smaller chromosome 19, and in general the size of the chromosome anomaly is inversely correlated to the resulting fitness2.
1
2
heterogeneity measurements do not give information about the type of mis-segregation that has occurred, fail to detect cells that die as a result of the mis-segregation, and therefore do not provide ultimate proof that CIN is still ongoing.
Other methods that have been used to estimate CIN in vivo are the use of organoid cultures, MEFs and other primary cell cultures directly derived from in vivo models (also see above 2. Measuring CIN in vitro).
As explained above, the most reliable evidence for CIN in vivo is catching mis-segregating cells in the act, for instance using intra-vital imaging. However, intra-vital imaging at sufficient resolution to monitor mitosis is technically challenging and requires dedicated animal models in which mitotic components are fluorescently labeled. Unfortunately, only a few mouse models have been engineered for this purpose (e.g. mice expressing H2B-GFP)69. While mouse models have been employed to
monitor the rate of cell division70, they have not been used to monitor
chromosome (mis)segregation per se. Other studies have used intra-vital imaging to measure CIN in fluorescently labeled xenografted tumor cells71.
Measuring CIN in vivo would more accurately show the effect of CIN for instance during development, to assess the role of the immune system and inter-tissue interactions, and the effect of CIN in the context of an unperturbed tissue. However, drawbacks of in vivo CIN measurement are the limited time available for imaging, the high cost, technical difficulty of creating and imaging animal models, and the relatively low rate of cell division in vivo. Therefore in most cases, aneuploidy is measured in vivo and not CIN. However, as explained above, the level of aneuploidy within a cell population may not be representative of the level of CIN.
4. Does aneuploidy lead to CIN?
As explained earlier, CIN leads to aneuploidy, but does aneuploidy also lead to CIN? Additionally, does the level of aneuploidy in a cell
population automatically correlate with the level of CIN? It is still contested in the field whether aneuploidy itself induces CIN. Most studies concerning the relationship of aneuploidy with CIN look at single chromosome gains and focus of the direct effects of aneuploidy, leaving the effect of complex karyotypes and chromosome losses unstudied. For
this purpose, aneuploid cells have been engineered via microcell-mediated chromosome transfer (MMCT)72,73.
Indeed, several trisomies, tetrasomies and a few double-trisomies were shown to cause further mis-segregation events in both cancer and non-cancer cell lines51,52. A trisomy of chromosome 13 in human cells can
lead to failing cytokinesis, while a trisomy of chromosome 7 does not, showing that different aneuploid karyotypes can cause different forms of CIN51. However, in fully triploid fibroblasts (3 copies of each
chromosome) the percentage of aneuploid cells increases, but the rate of chromosome mis-segregation (CIN), relative to a euploid control, does not41. This can be explained by the fact that triploid cells that
encountered chromosome losses had a growth advantage relative to their fully-triploid counterparts, whereas chromosome losses were selected against in the diploid cells; this shows that the increased rate of aneuploidy was due to selection, not an increased CIN rate. Additionally, while aneuploidy can lead to increased CIN in the short term, whether this is a long-tern effect remains controversial due to relatively short-term CIN endpoint measurements. Therefore, the level of aneuploidy is not always an accurate representation of the level of CIN, nor does an aneuploid state always increase chromosome mis-segregation frequency.
Other reports suggest that aneuploidy does not necessarily increase the rate of CIN. For instance, patients with trisomies for chromosome 13, 18 or 21 were shown to not display significantly higher levels of
aneuploidies of other chromosomes compared to euploid fibroblasts in
vitro41. Together, these findings suggest that some aneuploidies can
provoke further CIN, but not in all settings.
5. Aneuploidy can have highly differential effects on cells. 5.1. What are the effects of aneuploidy in vitro?
In general, aneuploidy has been shown to be detrimental to cell growth2,42,43,52. This growth defect is caused by the deregulation of the
proteasome3,74–77, cell metabolism78, transcriptome73, and the increase in
cellular senescence2,79. However, not all aneuploidies have the same
effect on cells. For example, a trisomy of chromosome 1 confers a much larger growth defect than a trisomy of the smaller chromosome 19, and in general the size of the chromosome anomaly is inversely correlated to the resulting fitness2.
Cancer cells, however, are frequently highly aneuploid. For instance, 61% of all cancerous mouse cell lines tested were aneuploid80, and most
human cancer cell lines are aneuploid as well81–83. While single
chromosome trisomies can act as tumor suppressors (in the euploid HCT 116 colorectal cancer cell line and in mouse xenografts),
spontaneous karyotype evolution in these cell lines leads to selection for karyotypes that grow better than their euploid counterparts2. Indeed,
the frequency of aneuploidy among tumors and cancer cell lines shows that aneuploidy in a cancer setting is quite well tolerated, with certain aneuploidies even being beneficial to cancer cell growth.
Many non-cancer cell lines have also become aneuploid when cultured, showing that certain aneuploid karyotypes can be beneficial to cell growth even in a non-transformed setting2,3,84. Culture conditions have
been shown to lead to low levels of aneuploidy24,25, and these low levels
of aneuploidy can lead to the expansion of a clone with an advantageous karyotype3. For example 25% of non-cancer mouse cell lines are
aneuploid80 and RPE-1 cells (a human, non-cancer cell line which was
initially euploid when harvested) often display clonal gain of
chromosome 10q and 1252. Gains of chromosomes 12 and 17 have also
been reported to be quite common in human pluripotent stem cells (hPSCs)85, with gain of chromosome 12 in hPSCs having been shown to
cause a growth benefit3.
5.2. How well is aneuploidy tolerated in vivo?
While there are very few models that have studied the effect of
aneuploidy in a non-CIN background, one group has studied the effect of defined chromosome trisomies in vivo. In this study, aneuploid and euploid hematopoietic stem cells (HSCs) were transplanted into irradiated recipient mice. Indeed, aneuploid HSCs were shown to have reduced fitness relative to euploid controls, and their recipient mice had a significantly decreased survival relative to euploid HSC-injected littermates. This effect was more obvious for HSCs with trisomies for larger chromosomes, with stem cells containing a trisomy of
chromosome 19 (the smallest autosomal chromosome in the mouse) performing almost as well as wildtype HSCs in their ability to
reconstitute hematopoietic stem cell lineages in vivo with little to no effect on mouse survival or blood cell regeneration86.
Aneuploidy also occurs in healthy human and mouse tissues in vivo at low levels. For instance, significant aneuploidy has been reported for
human brain87,88, and liver89,90, with aneuploidy affecting up to 30% of
the cells in unperturbed tissues in vivo. However, these observations were challenged by more recent studies that made use of single-cell genomics to more reliably measure aneuploidy. These studies suggest that the level of aneuploidy in brain and liver is much lower than previously reported, with aneuploidy only affecting ~2-5% of the cells, similar to the low rates of aneuploidy which were reported for
epidermis91–93. Together these data suggest that many human tissues,
including brain, skin and liver, contain a low level of aneuploidy.
However, the exact contribution of aneuploid cells to in vivo tissues still needs to be more accurately quantified.
While mosaic aneuploidy, an occasional aneuploid cell within a euploid tissue, might affect many tissue types in vivo, systemic aneuploidy, a condition in which every cell in an organism has the same aneuploid karyotype, is rarely tolerated. For instance, in humans, most types of systemic aneuploidy are embryonic lethal.5 The major exceptions are
the trisomy of chromosome 13, 18 and 21 (Patau’s, Edward’s and Down’s syndromes respectively). However, these aneuploidies have a severe impact on development and life expectancy of the affected individuals: the median survival age of children with Edward’s and Patau’s syndrome is less than two weeks94 and Down’s syndrome
individuals suffer from congenital disease, and mental retardation95.
Similarly, in mice, aneuploidy of autosomal chromosomes is not tolerated during embryonic development, with the exception of
chromosome 19 which leads to death shortly after birth96. Thus, unlike
low-grade mosaic CIN, systemic aneuploidies are not tolerated, except for a few trisomies, highlighting the importance of a euploid set of chromosomes, especially in the earliest phases of development. 5.3. The effects of aneuploidy.
In summary, aneuploidy is usually detrimental for cells, tissues and thus organisms, yielding decreased proliferation rates and reduced cellular fitness in tissue culture and developmental defects in organisms. However, aneuploidy can also be tolerated, with low levels of aneuploidy persisting in both cell culture and in vivo conditions. Additionally, aneuploidy can even be beneficial to cell growth and survival, with some trisomies being selected for in tissue culture, and the various aneuploid karyotypes that are selected for in cancers. However, to better understand the contribution of various aneuploidies to cellular, tumor and organismal fitness and proliferation, it is
1
2
Cancer cells, however, are frequently highly aneuploid. For instance, 61% of all cancerous mouse cell lines tested were aneuploid80, and most
human cancer cell lines are aneuploid as well81–83. While single
chromosome trisomies can act as tumor suppressors (in the euploid HCT 116 colorectal cancer cell line and in mouse xenografts),
spontaneous karyotype evolution in these cell lines leads to selection for karyotypes that grow better than their euploid counterparts2. Indeed,
the frequency of aneuploidy among tumors and cancer cell lines shows that aneuploidy in a cancer setting is quite well tolerated, with certain aneuploidies even being beneficial to cancer cell growth.
Many non-cancer cell lines have also become aneuploid when cultured, showing that certain aneuploid karyotypes can be beneficial to cell growth even in a non-transformed setting2,3,84. Culture conditions have
been shown to lead to low levels of aneuploidy24,25, and these low levels
of aneuploidy can lead to the expansion of a clone with an advantageous karyotype3. For example 25% of non-cancer mouse cell lines are
aneuploid80 and RPE-1 cells (a human, non-cancer cell line which was
initially euploid when harvested) often display clonal gain of
chromosome 10q and 1252. Gains of chromosomes 12 and 17 have also
been reported to be quite common in human pluripotent stem cells (hPSCs)85, with gain of chromosome 12 in hPSCs having been shown to
cause a growth benefit3.
5.2. How well is aneuploidy tolerated in vivo?
While there are very few models that have studied the effect of
aneuploidy in a non-CIN background, one group has studied the effect of defined chromosome trisomies in vivo. In this study, aneuploid and euploid hematopoietic stem cells (HSCs) were transplanted into irradiated recipient mice. Indeed, aneuploid HSCs were shown to have reduced fitness relative to euploid controls, and their recipient mice had a significantly decreased survival relative to euploid HSC-injected littermates. This effect was more obvious for HSCs with trisomies for larger chromosomes, with stem cells containing a trisomy of
chromosome 19 (the smallest autosomal chromosome in the mouse) performing almost as well as wildtype HSCs in their ability to
reconstitute hematopoietic stem cell lineages in vivo with little to no effect on mouse survival or blood cell regeneration86.
Aneuploidy also occurs in healthy human and mouse tissues in vivo at low levels. For instance, significant aneuploidy has been reported for
human brain87,88, and liver89,90, with aneuploidy affecting up to 30% of
the cells in unperturbed tissues in vivo. However, these observations were challenged by more recent studies that made use of single-cell genomics to more reliably measure aneuploidy. These studies suggest that the level of aneuploidy in brain and liver is much lower than previously reported, with aneuploidy only affecting ~2-5% of the cells, similar to the low rates of aneuploidy which were reported for
epidermis91–93. Together these data suggest that many human tissues,
including brain, skin and liver, contain a low level of aneuploidy.
However, the exact contribution of aneuploid cells to in vivo tissues still needs to be more accurately quantified.
While mosaic aneuploidy, an occasional aneuploid cell within a euploid tissue, might affect many tissue types in vivo, systemic aneuploidy, a condition in which every cell in an organism has the same aneuploid karyotype, is rarely tolerated. For instance, in humans, most types of systemic aneuploidy are embryonic lethal.5 The major exceptions are
the trisomy of chromosome 13, 18 and 21 (Patau’s, Edward’s and Down’s syndromes respectively). However, these aneuploidies have a severe impact on development and life expectancy of the affected individuals: the median survival age of children with Edward’s and Patau’s syndrome is less than two weeks94 and Down’s syndrome
individuals suffer from congenital disease, and mental retardation95.
Similarly, in mice, aneuploidy of autosomal chromosomes is not tolerated during embryonic development, with the exception of
chromosome 19 which leads to death shortly after birth96. Thus, unlike
low-grade mosaic CIN, systemic aneuploidies are not tolerated, except for a few trisomies, highlighting the importance of a euploid set of chromosomes, especially in the earliest phases of development. 5.3. The effects of aneuploidy.
In summary, aneuploidy is usually detrimental for cells, tissues and thus organisms, yielding decreased proliferation rates and reduced cellular fitness in tissue culture and developmental defects in organisms. However, aneuploidy can also be tolerated, with low levels of aneuploidy persisting in both cell culture and in vivo conditions. Additionally, aneuploidy can even be beneficial to cell growth and survival, with some trisomies being selected for in tissue culture, and the various aneuploid karyotypes that are selected for in cancers. However, to better understand the contribution of various aneuploidies to cellular, tumor and organismal fitness and proliferation, it is
necessary to better quantify the specific karyotypes that are selected for in cancers or cultured cells, including quantification cell-to-cell
heterogeneity for instance using single-cell genomics. 6. CIN can have highly differential effects on cells.
As discussed above, low rates of aneuploidy appear to be tolerated in
vivo, depending on the tissue and chromosomes affected. Is this also true
for cells displaying high-grade CIN when karyotypes are much more dynamic? It is difficult to make broad statements about CIN tolerance, as there are many different forms and rates of CIN. However, aneuploid cells with elevated levels of CIN51 had a growth benefit relative to
euploid control cells in non-standard, stressful culture conditions (serum-free conditions, and exposure to chemotherapeutic drug 5-Fluorouracil), but a growth defect under normal conditions.6 This
suggests that CIN may have selective advantages specifically in stressful, or non-standard growth conditions.
It is important to keep in mind, that when one quantifies chromosome mis-segregation frequencies in tissue culture cells, most cell lines already exhibit low levels of chromosome mis-segregation, even in the absence of CIN-inducing mutations16,51,52,57,97. These mis-segregation
events will lead to karyotype heterogeneity and karyotype evolution in the cell population and might explain the low rate of aneuploidy and karyotype evolution in cell cultures in which no (additional) CIN is induced (see Figure 2a for an example of karyotype evolution with low CIN). However, whether this ‘low-grade background’ CIN rate is representative for the in vivo situation still needs to be assessed.
While this ‘low-grade background’ level of CIN appears to be tolerated in cell culture, high-grade CIN causes growth defects and increased
senescence when provoked in vitro98. Furthermore, high levels of CIN
are usually lethal to cells grown in vitro 56, presumably due to the fact
that single cell karyotypes evolve too fast to be selected for (also see Figure 2b). Transient high levels of CIN, however, result in proliferation defects, but are not always lethal2.
Another important parameter to consider when predicting the response to CIN is the genetic background. For example, CIN induced by SAC alleviation through Mad2 inactivation was reported to be lethal in human cancer cell lines (HeLa and T98G)56, and in mouse embryonic
fibroblasts (MEFs). However, MEFs become much more tolerant of CIN
when p53 was also deleted57, although long term cultures of
SAC-deficient MEFs are difficult to maintain38.
Similar to what is observed in vitro, high-grade CIN is not tolerated when provoked early in development: for instance full SAC abrogation through Mad2 inactivation in the mouse, with or without p53 deletion, yields early embryonic death57,99.
However, high-grade CIN is not always lethal in vivo. For instance, in Drosophila, high-grade CIN in neural stem cells causes a reduction in brain size due to apoptosis100, and/or premature differentiation101,
without completely eliminating neural tissue. Additionally, when high-grade CIN is provoked in developing murine epidermis by inactivation of Mad2, hair follicle stem cells disappear, but the basal cells in the epidermis cope remarkably well with the resulting aneuploidy15. The
same is true for T-cells and hepatocytes. While in this setting CIN alone was not sufficient to provoke lymphoma or hepatocellular carcinoma, respectively, concomitant p53 loss yielded aggressive cancers exhibiting high-grade CIN. This indicates that aneuploid cells, while not cancerous initially, are prone to adopt a malignant fate when combined with other predisposing mutations27,38,50.
Interestingly, a recent study suggests that aneuploid cells might be cleared by the immune system. In this study CIN is induced to create senescent, aneuploid cells that were then co-cultured with natural killer cells (NK92) in vitro. Strikingly, the aneuploidy-induced senescent cells were selectively eliminated compared to the euploid control cells, suggesting that aneuploid cells can trigger an immune response.97
Furthermore, the innate immune system was recently found to detect DNA in micronuclei through cytosolic nucleic acid sensor (cGAS)-mediated surveillance102, which could be another way our immune
system detects aneuploid cells.
While these observations collectively indicate that CIN might be tolerated by some tissues when provoked later during development, further work is required to investigate the fate of CIN cells over time and their possible clearance by the immune system in vivo. However, clearly, the consequences of CIN are different in vivo and in vitro; and therefore, we cannot assume in vitro data to be representative of how human cells respond to CIN in vivo. Creating better models to observe the
1
2
necessary to better quantify the specific karyotypes that are selected for in cancers or cultured cells, including quantification cell-to-cell
heterogeneity for instance using single-cell genomics. 6. CIN can have highly differential effects on cells.
As discussed above, low rates of aneuploidy appear to be tolerated in
vivo, depending on the tissue and chromosomes affected. Is this also true
for cells displaying high-grade CIN when karyotypes are much more dynamic? It is difficult to make broad statements about CIN tolerance, as there are many different forms and rates of CIN. However, aneuploid cells with elevated levels of CIN51 had a growth benefit relative to
euploid control cells in non-standard, stressful culture conditions (serum-free conditions, and exposure to chemotherapeutic drug 5-Fluorouracil), but a growth defect under normal conditions.6 This
suggests that CIN may have selective advantages specifically in stressful, or non-standard growth conditions.
It is important to keep in mind, that when one quantifies chromosome mis-segregation frequencies in tissue culture cells, most cell lines already exhibit low levels of chromosome mis-segregation, even in the absence of CIN-inducing mutations16,51,52,57,97. These mis-segregation
events will lead to karyotype heterogeneity and karyotype evolution in the cell population and might explain the low rate of aneuploidy and karyotype evolution in cell cultures in which no (additional) CIN is induced (see Figure 2a for an example of karyotype evolution with low CIN). However, whether this ‘low-grade background’ CIN rate is representative for the in vivo situation still needs to be assessed.
While this ‘low-grade background’ level of CIN appears to be tolerated in cell culture, high-grade CIN causes growth defects and increased
senescence when provoked in vitro98. Furthermore, high levels of CIN
are usually lethal to cells grown in vitro 56, presumably due to the fact
that single cell karyotypes evolve too fast to be selected for (also see Figure 2b). Transient high levels of CIN, however, result in proliferation defects, but are not always lethal2.
Another important parameter to consider when predicting the response to CIN is the genetic background. For example, CIN induced by SAC alleviation through Mad2 inactivation was reported to be lethal in human cancer cell lines (HeLa and T98G)56, and in mouse embryonic
fibroblasts (MEFs). However, MEFs become much more tolerant of CIN
when p53 was also deleted57, although long term cultures of
SAC-deficient MEFs are difficult to maintain38.
Similar to what is observed in vitro, high-grade CIN is not tolerated when provoked early in development: for instance full SAC abrogation through Mad2 inactivation in the mouse, with or without p53 deletion, yields early embryonic death57,99.
However, high-grade CIN is not always lethal in vivo. For instance, in Drosophila, high-grade CIN in neural stem cells causes a reduction in brain size due to apoptosis100, and/or premature differentiation101,
without completely eliminating neural tissue. Additionally, when high-grade CIN is provoked in developing murine epidermis by inactivation of Mad2, hair follicle stem cells disappear, but the basal cells in the epidermis cope remarkably well with the resulting aneuploidy15. The
same is true for T-cells and hepatocytes. While in this setting CIN alone was not sufficient to provoke lymphoma or hepatocellular carcinoma, respectively, concomitant p53 loss yielded aggressive cancers exhibiting high-grade CIN. This indicates that aneuploid cells, while not cancerous initially, are prone to adopt a malignant fate when combined with other predisposing mutations27,38,50.
Interestingly, a recent study suggests that aneuploid cells might be cleared by the immune system. In this study CIN is induced to create senescent, aneuploid cells that were then co-cultured with natural killer cells (NK92) in vitro. Strikingly, the aneuploidy-induced senescent cells were selectively eliminated compared to the euploid control cells, suggesting that aneuploid cells can trigger an immune response.97
Furthermore, the innate immune system was recently found to detect DNA in micronuclei through cytosolic nucleic acid sensor (cGAS)-mediated surveillance102, which could be another way our immune
system detects aneuploid cells.
While these observations collectively indicate that CIN might be tolerated by some tissues when provoked later during development, further work is required to investigate the fate of CIN cells over time and their possible clearance by the immune system in vivo. However, clearly, the consequences of CIN are different in vivo and in vitro; and therefore, we cannot assume in vitro data to be representative of how human cells respond to CIN in vivo. Creating better models to observe the
Figure 2: The possible effects of CIN on cell ploidy over time. A) The different fates of
aneuploid cells generated in a CIN population over time, as a percent of total cell number. B) High-grade CIN will lead to cell death. Some karyotypes may be tolerated, but ongoing CIN will cause further mis-segregations and eventually, cell death. C) Selection of favorable karyotypes in a low-grade CIN background can lead to population growth and tumorigenesis over time.
* Death for disadvantageous karyotypes.
** Plateau for slow growth or no growth karyotypes. *** Selection for advantageous karyotypes.
Euploid cells CIN CIN is ongoing To tal c ell s (%) A) Time * *** ** Aneuploid kary otypes Eu ploid ka ryo typ e * Euploid cells High CIN
High-grade CIN is ongoing
To ta l ce lls (qu an tit y) Time B) * Euploid cells CIN
Low-grade CIN is ongoing
To ta l cell s (q ua ntity) Time C) * ** ***
especially since genomic instability is a hallmark of both cancer and aging9,103.
7. What effect do CIN and aneuploidy have on cancer?
90% of solid tumors, and 75% of hematopoietic tumors have been reported to be aneuploid8. Furthermore, aneuploidy is associated with
poor patient survival12,13,104–111. The karyotypes observed in cancer are
not random, as different cancer types display specific karyotypes17,27,38.
While certain forms of aneuploidy can be beneficial to tumor growth, the relationship between cancer and CIN remains paradoxical. On the one hand CIN itself can lead to tumorigenesis in some models14,112, but
on the other hand, CIN has been found to act as a tumor suppressor in other models31,101,113. Furthermore, yet other CIN models are suggesting
that CIN only enhances tumorigenesis in a tumor-prone background19– 21,38,50. These inconsistent findings may be explained by different rates
and types of CIN in the various models. Indeed, it has been suggested that high rates of CIN inhibit tumor progression, while lower CIN rates allow for tumor evolution16,31,113 (also see Figs. 2b and 2c). Importantly,
CIN does not only seem to affect cancer growth, but also the ability of cancer to adapt and spread as transient CIN induction can lead to tumor recurrence29,114. In addition, CIN can facilitate metastasis115, lead to drug
resistance26 and lowers survival rate in patients116.
However, to better understand which CIN rates in cancer drive further evolution towards therapy resistance and which CIN rates are toxic for tumor progression, we need better tools to quantify CIN rates in vivo. While measuring karyotype heterogeneity to estimate CIN rates might provide part of the answer27, new models in which chromosome
mis-segregation can be monitored live in vivo39 will be an important step
forward towards linking chromosome mis-segregation frequency to tumor outcome.
8. What effect do CIN and aneuploidy have on aging?
In addition to a cancer hallmark, genomic instability is also considered a hallmark of aging103. and indeed aneuploidy has been found to increase
with age in mouse tissues117 and in human blood cells118. In line with
these findings, induced high-grade CIN has been found to accelerate aging in some mouse models. For instance decreased expression of the SAC protein BubR1 in vivo through a hypomorphic allele results in high-grade CIN, severe premature aging phenotypes and a significantly shorter lifespan119. Conversely, BubR1 overexpression was found to
1
2
Figure 2: The possible effects of CIN on cell ploidy over time. A) The different fates of
aneuploid cells generated in a CIN population over time, as a percent of total cell number. B) High-grade CIN will lead to cell death. Some karyotypes may be tolerated, but ongoing CIN will cause further mis-segregations and eventually, cell death. C) Selection of favorable karyotypes in a low-grade CIN background can lead to population growth and tumorigenesis over time.
* Death for disadvantageous karyotypes.
** Plateau for slow growth or no growth karyotypes. *** Selection for advantageous karyotypes.
Euploid cells CIN CIN is ongoing To tal c ell s (%) A) Time * *** ** Aneuploid kary otypes Eu ploid ka ryo typ e * Euploid cells High CIN
High-grade CIN is ongoing
To ta l ce lls (qu an tit y) Time B) * Euploid cells CIN
Low-grade CIN is ongoing
To ta l cell s (q ua ntity) Time C) * ** ***
especially since genomic instability is a hallmark of both cancer and aging9,103.
7. What effect do CIN and aneuploidy have on cancer?
90% of solid tumors, and 75% of hematopoietic tumors have been reported to be aneuploid8. Furthermore, aneuploidy is associated with
poor patient survival12,13,104–111. The karyotypes observed in cancer are
not random, as different cancer types display specific karyotypes17,27,38.
While certain forms of aneuploidy can be beneficial to tumor growth, the relationship between cancer and CIN remains paradoxical. On the one hand CIN itself can lead to tumorigenesis in some models14,112, but
on the other hand, CIN has been found to act as a tumor suppressor in other models31,101,113. Furthermore, yet other CIN models are suggesting
that CIN only enhances tumorigenesis in a tumor-prone background19– 21,38,50. These inconsistent findings may be explained by different rates
and types of CIN in the various models. Indeed, it has been suggested that high rates of CIN inhibit tumor progression, while lower CIN rates allow for tumor evolution16,31,113 (also see Figs. 2b and 2c). Importantly,
CIN does not only seem to affect cancer growth, but also the ability of cancer to adapt and spread as transient CIN induction can lead to tumor recurrence29,114. In addition, CIN can facilitate metastasis115, lead to drug
resistance26 and lowers survival rate in patients116.
However, to better understand which CIN rates in cancer drive further evolution towards therapy resistance and which CIN rates are toxic for tumor progression, we need better tools to quantify CIN rates in vivo. While measuring karyotype heterogeneity to estimate CIN rates might provide part of the answer27, new models in which chromosome
mis-segregation can be monitored live in vivo39 will be an important step
forward towards linking chromosome mis-segregation frequency to tumor outcome.
8. What effect do CIN and aneuploidy have on aging?
In addition to a cancer hallmark, genomic instability is also considered a hallmark of aging103. and indeed aneuploidy has been found to increase
with age in mouse tissues117 and in human blood cells118. In line with
these findings, induced high-grade CIN has been found to accelerate aging in some mouse models. For instance decreased expression of the SAC protein BubR1 in vivo through a hypomorphic allele results in high-grade CIN, severe premature aging phenotypes and a significantly shorter lifespan119. Conversely, BubR1 overexpression was found to
delay the onset of aging and its associated pathologies120. Similarly,
humans that harbor mutations, among others, in the gene encoding the human homologue for BubR1 (BUB1B) suffer from a rare accelerated aging disease, mosaic variegated aneuploidy or MVA, which leads to a significantly increased number of aneuploid cells in these patients121–123.
The premature aging phenotype observed in BubR1 hypomorphic mice coincides with an accumulation of senescent cells124. Strikingly, clearing
these senescent cells significantly reduces and even partially reverts the aging phenotype125,126. This suggests that CIN might contribute to aging
by increasing the number of senescent cells in vivo. Similarly, Bub1/Rae hypo-morphs, and eIF-5A2 overexpression mice also exhibit increased CIN rates and display accelerated aging127,128. although the exact
mechanism of eIF-5A2 is still unknown.
While there are strong indications that CIN can cause premature aging phenotypes and that organisms accumulate aneuploid cells with aging, the relationship between CIN, aneuploidy and aging is still not fully understood, as only a few of the CIN models show accelerated aging. Therefore, studying the effects of an accumulation of aneuploid and senescent cells in tissues, as well as measuring the rate of chromosome mis-segregation, and the fate of cells after mis-segregation in vivo may lead to a better understanding of what does or does not cause
(accelerated) aging in mammals. 9. Conclusions and outlook
CIN and the resulting aneuploidy play important roles in cancer129 and
aging. To better understand these roles, we need to study the
differential effects of CIN and aneuploidy in vivo. While aneuploidy and CIN are linked, they are not the same thing. Unlike aneuploidy, CIN is rarely measured in vivo, and cell culture CIN rates are often used to estimate in vivo CIN rates. However, CIN can have vastly different effects on cells in vitro and in vivo, and even between different cell lineages in
vivo. Therefore, only using cell culture models to estimate the type of
CIN, mis-segregation rate, and response to CIN in vivo will likely lead to an inaccurate or at least incomplete perspective. Additionally, there are many types of CIN that yield different CIN rates and different karyotypic makeups. The response to CIN depends on a variety of cellular and extracellular factors including the genotype (e.g. mutations in DNA damage and apoptosis genes), the level and underlying defect that causes the increased rate of CIN, the specific aneuploid karyotype, the
period of time a cell exhibits a higher rate of CIN, the cell/tissue type and the in vivo or in vitro context.
Therefore, we cannot assume that cell lines fully represent what occurs in mouse or human tissues, and we cannot (only) use aneuploidy endpoint measurements to estimate in vivo CIN rates. To better understand the effects of CIN, and its role in cancer progression and aging, we need new mouse models in which chromosome mis-segregation rates, and the fate of the resulting daughter cells can be observed in vivo. Such models will allow us to see whether the rates of mis-segregation and resulting cellular fates are comparable in vitro and
in vivo, to thus better understand cancer progression and CIN tolerance in vivo.
Acknowledgements
We would like to thank Bjorn Bakker, Judith Simon and Eleanor Elgood Hunt for critically reading the manuscript and giving helpful feedback. This work was supported by the European Union through an FP7 Marie Curie Innovative Training Network grant, PloidyNet and through support of the Groningen foundation for Pediatric Oncology (SKOG). We declare that we have no conflict of interest.
1
2
delay the onset of aging and its associated pathologies120. Similarly,
humans that harbor mutations, among others, in the gene encoding the human homologue for BubR1 (BUB1B) suffer from a rare accelerated aging disease, mosaic variegated aneuploidy or MVA, which leads to a significantly increased number of aneuploid cells in these patients121–123.
The premature aging phenotype observed in BubR1 hypomorphic mice coincides with an accumulation of senescent cells124. Strikingly, clearing
these senescent cells significantly reduces and even partially reverts the aging phenotype125,126. This suggests that CIN might contribute to aging
by increasing the number of senescent cells in vivo. Similarly, Bub1/Rae hypo-morphs, and eIF-5A2 overexpression mice also exhibit increased CIN rates and display accelerated aging127,128. although the exact
mechanism of eIF-5A2 is still unknown.
While there are strong indications that CIN can cause premature aging phenotypes and that organisms accumulate aneuploid cells with aging, the relationship between CIN, aneuploidy and aging is still not fully understood, as only a few of the CIN models show accelerated aging. Therefore, studying the effects of an accumulation of aneuploid and senescent cells in tissues, as well as measuring the rate of chromosome mis-segregation, and the fate of cells after mis-segregation in vivo may lead to a better understanding of what does or does not cause
(accelerated) aging in mammals. 9. Conclusions and outlook
CIN and the resulting aneuploidy play important roles in cancer129 and
aging. To better understand these roles, we need to study the
differential effects of CIN and aneuploidy in vivo. While aneuploidy and CIN are linked, they are not the same thing. Unlike aneuploidy, CIN is rarely measured in vivo, and cell culture CIN rates are often used to estimate in vivo CIN rates. However, CIN can have vastly different effects on cells in vitro and in vivo, and even between different cell lineages in
vivo. Therefore, only using cell culture models to estimate the type of
CIN, mis-segregation rate, and response to CIN in vivo will likely lead to an inaccurate or at least incomplete perspective. Additionally, there are many types of CIN that yield different CIN rates and different karyotypic makeups. The response to CIN depends on a variety of cellular and extracellular factors including the genotype (e.g. mutations in DNA damage and apoptosis genes), the level and underlying defect that causes the increased rate of CIN, the specific aneuploid karyotype, the
period of time a cell exhibits a higher rate of CIN, the cell/tissue type and the in vivo or in vitro context.
Therefore, we cannot assume that cell lines fully represent what occurs in mouse or human tissues, and we cannot (only) use aneuploidy endpoint measurements to estimate in vivo CIN rates. To better understand the effects of CIN, and its role in cancer progression and aging, we need new mouse models in which chromosome mis-segregation rates, and the fate of the resulting daughter cells can be observed in vivo. Such models will allow us to see whether the rates of mis-segregation and resulting cellular fates are comparable in vitro and
in vivo, to thus better understand cancer progression and CIN tolerance in vivo.
Acknowledgements
We would like to thank Bjorn Bakker, Judith Simon and Eleanor Elgood Hunt for critically reading the manuscript and giving helpful feedback. This work was supported by the European Union through an FP7 Marie Curie Innovative Training Network grant, PloidyNet and through support of the Groningen foundation for Pediatric Oncology (SKOG). We declare that we have no conflict of interest.
