Aneuploidy in the human brain and cancer
van den Bos, Hilda
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: 2017
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
van den Bos, H. (2017). Aneuploidy in the human brain and cancer: Studying heterogeneity using single-cell sequencing. University of Groningen.
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.
Chapter 6
General discussion
The exploration of the human genome started around 150 years ago when Friedrich Miescher was the first to isolate DNA, which he called nuclein1. Later, Flemming was the first to describe
chromosomes and their behavior during cell division2. A very important next step was the
Boveri-Sutton chromosome theory. This theory states that the chromosomes contain the hereditary material, are present in pairs, and that genes on these chromosomes are responsible for the Mendelian inheritance3,4. This work from the early 1900’s was far ahead
of its time. Moreover, Boveri was the first to suggest that defects in the distribution of these chromosomes during cell division might lead to cancer5. Following the discovery of the
structure of DNA in 19526–8 and the development of the first DNA sequencing methods9,10,
studying the human genome11,12 became possible enabling new ways to study aneuploidy and
chromosomal instability. Continuous and ongoing improvements of sequencing methods have greatly reduced the costs and improved the throughput of sequencing projects13,14.
Unraveling of the human genome has had a major impact on technology development and research into the functionality of the genome, illustrated by the ENCODE project15. This
project identified functional regions in the human genome such as transcription start sites, transcription factor binding sites, regulatory elements and chromatin structural features15. In
general, large scale sequencing projects have greatly improved our understanding of genetics and the role of genetics in diseases16. To study Boveri’s theory, that defects in the distribution
of chromosomes are involved in cancer, several methods have been developed (Chapter 1). All available techniques have advantages and disadvantages and the best method to use in a particular study depends largely on the research question and the material available for study. For example, dividing cells are required for counting chromosomes and detecting sub-chromosomal copy number aberrations (CNAs) by staining condensed chromosomes, e.g. Giemsa staining or SKY17–19. To study non-dividing cells or avoid bias towards dividing cells,
other methods have been developed such as interphase FISH, SNP arrays and array CGH. All methods vary in the type of genomic aberrations that can be detected, their resolution, the number of chromosomes that can be studied simultaneously and whether they can be applied to single cells or DNA extracted from many cells.
Studying aneuploidy, CNAs and mutations with single-cell sequencing
The development of single-cell sequencing techniques has opened up a new field of research. By sequencing the genomes of individual cells, information on the genetic diversity in sample can be gathered which would be lost when the DNA of many cells is pooled and sequenced in one reaction. This genetic diversity is often referred to as the heterogeneity of a sample. The more cells differ from each other, in structural features such as inversions, deletions, amplifications and whole chromosomal CNAs as well as in point mutations, the more heterogeneous the sample is. The first single-cell sequencing protocol was published in 2011, which used a) flow sorting of nuclei, b) whole genome amplification for the construction of sequencing libraries and c) next generation sequencing20,21. To overcome the difficulties of
sequencing the minute amount of DNA present in single cells and to increase the fraction of the genome covered by sequencing, most current protocols still employ various forms of whole genome amplification22. Such pre-amplification has advantages as well as
disadvantages. Advantages include that sufficiently high genomic coverage in sequencing libraries can be obtained to identify point mutations. However, the uneven amplification of DNA, that is almost impossible to avoid using various amplification strategies, limits this approach for questions about copy number variations. For example, the commonly used method of multiple displacement amplification (MDA) results in high but uneven coverage, while degenerative oligonucleotide primer PCR (DOP-PCR) gives more even, but low coverage22. These limitations should be taken into account when choosing a particular
method of library construction for the analysis of DNA in single cells. Recently, our lab (chapter 3) and others have developed single-cell sequencing protocols that do not employ whole genome amplification23,24. The main advantage of producing single-cell sequencing
libraries without pre-amplification is the low but even genome coverage that can be obtained. Without pre-amplification DNA fragments are only amplified by PCR after annealing of sequencing adapters. Therefore, when two fragments with exactly the same sequence are found, they must have originated from the same DNA fragment. Such PCR duplicates can be easily identified and removed prior to further analysis. Additional advantages are the lower costs per cell, as fewer reagents are needed, and the higher throughput with fewer working hours. It is important to note that, due to the even, but low coverage achieved, pre-amplification free library preparation is only suitable for CNA analysis. The genome coverage is typically too low to reliably detect point mutations in individual cells.
Aneuploidy in the human brain
Aneuploidy, known from systemic trisomies such as Down syndrome, is a known hallmark of cancer. Cells from the brain have long been thought to have no chromosomal CNAs, but this view was challenged when large numbers of aneuploid cells were identified in developing and adult mouse brain using interphase fluorescence in situ hybridization (FISH)25–27. Moreover,
aneuploidy was also detected in developing human brain28,29 and adult human brain25,30.
Interestingly, some studies showed that the level of aneuploidy appeared to be even higher in the brain cells of patients with Alzheimer’s disease (AD)31–33. These observations suggested
that aneuploid cells might be involved in the process of neurodegeneration in AD. As the live expectancy increases, more and more people will develop neurodegenerative diseases such as AD. Despite the enormous effort put in developing medication to prevent, stop or delay progression of AD, available treatment options are still very limited. A new starting point for treatment would therefore be very welcome. Aneuploidy is known to cause changes in protein levels as the expression of the genes on extra chromosomes will result in unbalanced protein expression34. The excess proteins have to be folded in the right conformation or be
degraded. This can overload the protein folding pathways and/or protein degradation machinery leaving unfolded or incorrectly folded proteins present in the cells, which could
6
General discussion
The exploration of the human genome started around 150 years ago when Friedrich Miescher was the first to isolate DNA, which he called nuclein1. Later, Flemming was the first to describe
chromosomes and their behavior during cell division2. A very important next step was the
Boveri-Sutton chromosome theory. This theory states that the chromosomes contain the hereditary material, are present in pairs, and that genes on these chromosomes are responsible for the Mendelian inheritance3,4. This work from the early 1900’s was far ahead
of its time. Moreover, Boveri was the first to suggest that defects in the distribution of these chromosomes during cell division might lead to cancer5. Following the discovery of the
structure of DNA in 19526–8 and the development of the first DNA sequencing methods9,10,
studying the human genome11,12 became possible enabling new ways to study aneuploidy and
chromosomal instability. Continuous and ongoing improvements of sequencing methods have greatly reduced the costs and improved the throughput of sequencing projects13,14.
Unraveling of the human genome has had a major impact on technology development and research into the functionality of the genome, illustrated by the ENCODE project15. This
project identified functional regions in the human genome such as transcription start sites, transcription factor binding sites, regulatory elements and chromatin structural features15. In
general, large scale sequencing projects have greatly improved our understanding of genetics and the role of genetics in diseases16. To study Boveri’s theory, that defects in the distribution
of chromosomes are involved in cancer, several methods have been developed (Chapter 1). All available techniques have advantages and disadvantages and the best method to use in a particular study depends largely on the research question and the material available for study. For example, dividing cells are required for counting chromosomes and detecting sub-chromosomal copy number aberrations (CNAs) by staining condensed chromosomes, e.g. Giemsa staining or SKY17–19. To study non-dividing cells or avoid bias towards dividing cells,
other methods have been developed such as interphase FISH, SNP arrays and array CGH. All methods vary in the type of genomic aberrations that can be detected, their resolution, the number of chromosomes that can be studied simultaneously and whether they can be applied to single cells or DNA extracted from many cells.
Studying aneuploidy, CNAs and mutations with single-cell sequencing
The development of single-cell sequencing techniques has opened up a new field of research. By sequencing the genomes of individual cells, information on the genetic diversity in sample can be gathered which would be lost when the DNA of many cells is pooled and sequenced in one reaction. This genetic diversity is often referred to as the heterogeneity of a sample. The more cells differ from each other, in structural features such as inversions, deletions, amplifications and whole chromosomal CNAs as well as in point mutations, the more heterogeneous the sample is. The first single-cell sequencing protocol was published in 2011, which used a) flow sorting of nuclei, b) whole genome amplification for the construction of sequencing libraries and c) next generation sequencing20,21. To overcome the difficulties of
sequencing the minute amount of DNA present in single cells and to increase the fraction of the genome covered by sequencing, most current protocols still employ various forms of whole genome amplification22. Such pre-amplification has advantages as well as
disadvantages. Advantages include that sufficiently high genomic coverage in sequencing libraries can be obtained to identify point mutations. However, the uneven amplification of DNA, that is almost impossible to avoid using various amplification strategies, limits this approach for questions about copy number variations. For example, the commonly used method of multiple displacement amplification (MDA) results in high but uneven coverage, while degenerative oligonucleotide primer PCR (DOP-PCR) gives more even, but low coverage22. These limitations should be taken into account when choosing a particular
method of library construction for the analysis of DNA in single cells. Recently, our lab (chapter 3) and others have developed single-cell sequencing protocols that do not employ whole genome amplification23,24. The main advantage of producing single-cell sequencing
libraries without pre-amplification is the low but even genome coverage that can be obtained. Without pre-amplification DNA fragments are only amplified by PCR after annealing of sequencing adapters. Therefore, when two fragments with exactly the same sequence are found, they must have originated from the same DNA fragment. Such PCR duplicates can be easily identified and removed prior to further analysis. Additional advantages are the lower costs per cell, as fewer reagents are needed, and the higher throughput with fewer working hours. It is important to note that, due to the even, but low coverage achieved, pre-amplification free library preparation is only suitable for CNA analysis. The genome coverage is typically too low to reliably detect point mutations in individual cells.
Aneuploidy in the human brain
Aneuploidy, known from systemic trisomies such as Down syndrome, is a known hallmark of cancer. Cells from the brain have long been thought to have no chromosomal CNAs, but this view was challenged when large numbers of aneuploid cells were identified in developing and adult mouse brain using interphase fluorescence in situ hybridization (FISH)25–27. Moreover,
aneuploidy was also detected in developing human brain28,29 and adult human brain25,30.
Interestingly, some studies showed that the level of aneuploidy appeared to be even higher in the brain cells of patients with Alzheimer’s disease (AD)31–33. These observations suggested
that aneuploid cells might be involved in the process of neurodegeneration in AD. As the live expectancy increases, more and more people will develop neurodegenerative diseases such as AD. Despite the enormous effort put in developing medication to prevent, stop or delay progression of AD, available treatment options are still very limited. A new starting point for treatment would therefore be very welcome. Aneuploidy is known to cause changes in protein levels as the expression of the genes on extra chromosomes will result in unbalanced protein expression34. The excess proteins have to be folded in the right conformation or be
degraded. This can overload the protein folding pathways and/or protein degradation machinery leaving unfolded or incorrectly folded proteins present in the cells, which could
form the protein aggregates characteristic of AD. The possibility that aneuploidy in brain cells could therefore be involved in AD seemed an interesting possibility.
However, the level of reported aneuploidy varied widely from no aneuploidy35 up to
approximately 50%32. It must be noted that most of these studies used interphase FISH and
therefore studied only a few chromosomes in a single cell at a time, with the total aneuploidy rate extrapolated from the chromosomes that were labeled. Chapter 2 reviews the studies investigating the presence or absence of aneuploid cells in the normal human brain and brain affected by AD. In contrast to the FISH studies, small scale single-cell sequencing studies found much lower levels of aneuploidy in the human brain36–38. Our study, described in chapter 3, is
the first large scale study to use single-cell sequencing to study aneuploidy in normal and AD brain. We used brain cells from an individual with Down syndrome as a control to validate our method and found three copies of chromosome 21 in all cells analyzed and no other aneuploidies. In contrast to the FISH-based studies but in line with the sequencing studies36– 38, we found low levels of aneuploidy and no evidence for increased aneuploidy in human
(ageing) brain and AD-affected brain samples. Moreover, very low levels of aneuploidy were found in all samples analyzed, including neurons and non-neuronal cells. Our results thus challenge the previous data generated using FISH approaches. The inconsistent results can be explained in several ways. Either FISH-based methods overestimate the level of aneuploidy, or our single-cell sequencing approach somehow ‘missed’ the aneuploid cells picked up by the FISH-based methods. We could have missed aneuploid cells if they were only present in specific regions of the brain. However, this explanation seems unlikely since we analyzed the frontal cortex, the same region used in the FISH-based studies that identified aneuploid cells, in which amyloid plaques were detected by immunostaining indicating diseased tissue. Another possibility for a false negative result in our study would be the presence of micronuclei. When cells missegregate chromosomes, such chromosomes can end up in micronuclei, outside the main nucleus. Since we isolate and sort nuclei only, and not whole cells, information on the presence of micronuclei and the DNA within was lost. However, in none of the studies using FISH, the presence of micronuclei is reported which makes this perhaps an unlikely explanation. Also, we would still identify chromosome loss when micronuclei were not included in our analysis. On the other hand, FISH can lead to overestimation of aneuploidy in several ways, e.g. by non-specific binding of the probes, by hybridization failure. Moreover, single-cell sequencing will sample the copy number state of each individual chromosome hundreds of times, while interphase FISH will only sample a few chromosomes per single cell and only detect one fragment of the sampled chromosomes. In conclusion, the most likely explanation is that aneuploidy is less common in the human brain than previously thought, and not noticeably increased in AD.
Although the data presented in chapter 3 provide important evidence, a direct comparison between methods is still lacking. By performing both FISH and single-cell sequencing on cells isolated from the same samples, more conclusive results regarding the level of aneuploidy and the most reliable method to study aneuploidy in the human brain could be obtained. In such experiments, neuronal and non-neuronal cells from various brain areas should be
included as well as brain cells with a known aneuploidy. If FISH indeed is inherently noisier, one would expect to find higher aneuploidy rates with this method. Analysis of brain cells with a known, single, aneuploidy such as a trisomy 21, can reveal the sensitivity, whether the aneuploidy is detected in each cell, and specificity, the background noise, of both methods in a direct comparison. From these experiments, we will be able to conclude which method is the most suitable for the detection of low-level aneuploidy.
Besides establishing the best method for aneuploidy detection in the brain and determining the actual aneuploidy rate and variation in the normal and diseased human brain, it is important to continue research on other factors that are thought to play a smaller or bigger role in the development or progression of AD, such as inflammation, oxidative damage and impaired protein folding or clearance. Additionally, there is accumulating evidence that several lifestyle aspects can increase or decrease the risk of developing AD39. Maintaining a
normal weight, engaging in enough exercise, having healthy eating habits, being socially active and keeping the brain active are all thought to decrease the chance of developing AD. A lot can be gained by (better) implementation of these lifestyle habits.
Aneuploidy and chromosomal instability in cancer
In contrast to the brain, aneuploidy and CNAs are repeatedly shown to be important in cancer. Chromosomal instability (CIN) is known to be an important accelerating factor in cancer40.
Single-cell sequencing can be applied to any normal or diseased tissue from which single cells or nuclei can be isolated. Especially in cancer research, single-cell sequencing can make a big contribution41. Chapter 4 highlights the added value of single-cell sequencing for the
diagnosis, prognosis and monitoring of cancer. Single-cell sequencing can reveal important information on the evolution of a cancer. Does a tumor experience ongoing genomic and chromosomal instability or does heterogeneity develop in one event, followed by clonal outgrowth of selected clones?20,42,43 The degree of genomic heterogeneity may influence
whether a tumor will develop resistant clones upon treatment. Sequencing cells from primary tumors and its metastases can also provide insight in the metastatic processes. Which cells are released from the tumor, are capable of invading other tissues and grow out to form a metastasis? The clonality of a metastasis can reveal its origin: whether it was it formed from one cell, or a clump of different cells. This is emphasized in chapter 5, in which we describe the whole genome sequencing of single cells from small cell lung cancer. This study is unique in the number of cells sequenced from one patient and the number of sites, primary and metastatic, from which cells were isolated. Sequencing cells from two sites of the primary tumor and from liver, adrenal gland and lymph node metastases revealed both monoclonal and polyclonal metastatic seeding patterns. Also in colorectal, ovarian and prostate cancer mono- and polyclonal seeding patterns have been found44–46. Moreover, we identified
potential founder cells of the metastases in the primary tumor. These cells showed a CNA pattern very similar to the metastasis. When all cells are clustered based on CNA pattern similarity, these founder cells clustered with the metastasis and not with the primary tumor where they were isolated from. Also, the heterogeneity within the primary tumor and the
6
form the protein aggregates characteristic of AD. The possibility that aneuploidy in brain cellscould therefore be involved in AD seemed an interesting possibility.
However, the level of reported aneuploidy varied widely from no aneuploidy35 up to
approximately 50%32. It must be noted that most of these studies used interphase FISH and
therefore studied only a few chromosomes in a single cell at a time, with the total aneuploidy rate extrapolated from the chromosomes that were labeled. Chapter 2 reviews the studies investigating the presence or absence of aneuploid cells in the normal human brain and brain affected by AD. In contrast to the FISH studies, small scale single-cell sequencing studies found much lower levels of aneuploidy in the human brain36–38. Our study, described in chapter 3, is
the first large scale study to use single-cell sequencing to study aneuploidy in normal and AD brain. We used brain cells from an individual with Down syndrome as a control to validate our method and found three copies of chromosome 21 in all cells analyzed and no other aneuploidies. In contrast to the FISH-based studies but in line with the sequencing studies36– 38, we found low levels of aneuploidy and no evidence for increased aneuploidy in human
(ageing) brain and AD-affected brain samples. Moreover, very low levels of aneuploidy were found in all samples analyzed, including neurons and non-neuronal cells. Our results thus challenge the previous data generated using FISH approaches. The inconsistent results can be explained in several ways. Either FISH-based methods overestimate the level of aneuploidy, or our single-cell sequencing approach somehow ‘missed’ the aneuploid cells picked up by the FISH-based methods. We could have missed aneuploid cells if they were only present in specific regions of the brain. However, this explanation seems unlikely since we analyzed the frontal cortex, the same region used in the FISH-based studies that identified aneuploid cells, in which amyloid plaques were detected by immunostaining indicating diseased tissue. Another possibility for a false negative result in our study would be the presence of micronuclei. When cells missegregate chromosomes, such chromosomes can end up in micronuclei, outside the main nucleus. Since we isolate and sort nuclei only, and not whole cells, information on the presence of micronuclei and the DNA within was lost. However, in none of the studies using FISH, the presence of micronuclei is reported which makes this perhaps an unlikely explanation. Also, we would still identify chromosome loss when micronuclei were not included in our analysis. On the other hand, FISH can lead to overestimation of aneuploidy in several ways, e.g. by non-specific binding of the probes, by hybridization failure. Moreover, single-cell sequencing will sample the copy number state of each individual chromosome hundreds of times, while interphase FISH will only sample a few chromosomes per single cell and only detect one fragment of the sampled chromosomes. In conclusion, the most likely explanation is that aneuploidy is less common in the human brain than previously thought, and not noticeably increased in AD.
Although the data presented in chapter 3 provide important evidence, a direct comparison between methods is still lacking. By performing both FISH and single-cell sequencing on cells isolated from the same samples, more conclusive results regarding the level of aneuploidy and the most reliable method to study aneuploidy in the human brain could be obtained. In such experiments, neuronal and non-neuronal cells from various brain areas should be
included as well as brain cells with a known aneuploidy. If FISH indeed is inherently noisier, one would expect to find higher aneuploidy rates with this method. Analysis of brain cells with a known, single, aneuploidy such as a trisomy 21, can reveal the sensitivity, whether the aneuploidy is detected in each cell, and specificity, the background noise, of both methods in a direct comparison. From these experiments, we will be able to conclude which method is the most suitable for the detection of low-level aneuploidy.
Besides establishing the best method for aneuploidy detection in the brain and determining the actual aneuploidy rate and variation in the normal and diseased human brain, it is important to continue research on other factors that are thought to play a smaller or bigger role in the development or progression of AD, such as inflammation, oxidative damage and impaired protein folding or clearance. Additionally, there is accumulating evidence that several lifestyle aspects can increase or decrease the risk of developing AD39. Maintaining a
normal weight, engaging in enough exercise, having healthy eating habits, being socially active and keeping the brain active are all thought to decrease the chance of developing AD. A lot can be gained by (better) implementation of these lifestyle habits.
Aneuploidy and chromosomal instability in cancer
In contrast to the brain, aneuploidy and CNAs are repeatedly shown to be important in cancer. Chromosomal instability (CIN) is known to be an important accelerating factor in cancer40.
Single-cell sequencing can be applied to any normal or diseased tissue from which single cells or nuclei can be isolated. Especially in cancer research, single-cell sequencing can make a big contribution41. Chapter 4 highlights the added value of single-cell sequencing for the
diagnosis, prognosis and monitoring of cancer. Single-cell sequencing can reveal important information on the evolution of a cancer. Does a tumor experience ongoing genomic and chromosomal instability or does heterogeneity develop in one event, followed by clonal outgrowth of selected clones?20,42,43 The degree of genomic heterogeneity may influence
whether a tumor will develop resistant clones upon treatment. Sequencing cells from primary tumors and its metastases can also provide insight in the metastatic processes. Which cells are released from the tumor, are capable of invading other tissues and grow out to form a metastasis? The clonality of a metastasis can reveal its origin: whether it was it formed from one cell, or a clump of different cells. This is emphasized in chapter 5, in which we describe the whole genome sequencing of single cells from small cell lung cancer. This study is unique in the number of cells sequenced from one patient and the number of sites, primary and metastatic, from which cells were isolated. Sequencing cells from two sites of the primary tumor and from liver, adrenal gland and lymph node metastases revealed both monoclonal and polyclonal metastatic seeding patterns. Also in colorectal, ovarian and prostate cancer mono- and polyclonal seeding patterns have been found44–46. Moreover, we identified
potential founder cells of the metastases in the primary tumor. These cells showed a CNA pattern very similar to the metastasis. When all cells are clustered based on CNA pattern similarity, these founder cells clustered with the metastasis and not with the primary tumor where they were isolated from. Also, the heterogeneity within the primary tumor and the
metastases varied: especially the liver metastasis showed markedly less heterogeneity. Several possibilities could explain this observation. The cells could be optimally suited to grow at the selected site, showing less chromosome missegregation, or chromosomally instable cells could be selected against more stringently. It has been shown that in SCLC patients clumps of tumors cells are often present in the blood47. Metastatic spread can take place via
multiple routes. Metastases can form from one or multiple cells from the primary tumor resulting in mono- or polyclonal seeding, but cells from metastases can also give rise to additional metastases. Also, metastatic cells might migrate back into the primary tumor, called self-seeding, or into another metastasis48. It is more difficult to reconstruct
evolutionary trees based on CNAs than using mutations since it is not a linear process; in contrast to mutations, chromosomes can be gained during one cell division but the same chromosome might be lost again at a later point. Therefore, determining the order of events is often difficult to elucidate.
Aneuploidy in cancer development
It is still not known why tumors are so often aneuploid. Cells can become aneuploid when chromosomes missegregate during cell division. In normal cells mechanisms such as the spindle assembly checkpoint (SAC) are in place to ensure correct chromosome segregation49.
When these mechanisms do not function properly, chromosome missegregation can occur. Also, telomere shortening is known to cause aneuploidy. When telomeres become too short cells will generally exit the cell cycle50. With loss of the tumor suppressor p53 cells can
continue to proliferate, even with critically short telomeres51. This results in end-to-end fusion
of sister chromatid telomeres, and the formation of dicentric chromosomes. Dicentric chromosomes are likely to missegregate during mitosis thus resulting in DNA breaks and aneuploidy, and trigger so-called breakage-fusion-bridge (BFB) cycles. This can continue over many cell divisions, leading to a heterogeneous and aneuploid population of cells with large duplications and deletions52. Aneuploidy is known to inhibit growth, but at the same time it
seems to stimulate growth of cancer cells34,53,54. Studying aneuploidy in cancer will provide
clues on how and at which stage of the tumor it develops and how aneuploidy can stimulate the growth of tumor cells.
A study by the group of Amon emphasizes how aneuploidy might play a dual role in growth suppression and stimulation56. They found that cells harboring a single extra chromosome
showed growth inhibition and increased senescence. Even the activation of growth stimulating oncogenic pathways in combination with single chromosome aneuploidy showed reduced tumorigenic potential as they formed smaller tumors compared to euploid cells with the oncogene activation only. So even in the context of oncogene activation, single chromosome aneuploidy functions as a tumor suppressor. On the other hand, the single chromosome aneuploidy can cause chromosomal instability leading to additional structural and whole chromosome copy number abnormalities. These additional changes compensate for the growth defect and resulted in improved fitness and tumorigenicity suggesting that single aneuploidies could facilitate selection of cells with highly rearranged genomes56.
Indeed, a very recent study by the group of Medema describes the activation of p53 in untransformed cells with whole chromosome or structural aneuploidy. They found that structural aneuploidy does activate p53 and induces cell cycle arrest, but whole chromosome aneuploidy does not induce p53 activation in all cells; many are still able to proliferate with mild whole chromosome imbalances57. However, the fact that the number of whole
chromosome abnormalities decreased in the proliferating population does suggest selection against high levels of aneuploidy in normal cells. This is in line with the observation that complex karyotypes will induce a cell cycle arrest, while cells with low amounts of chromosomal imbalances are often able to continue to proliferate.58 These data suggest that
aneuploidy can trigger a tumor suppressor mechanism by inducing p53 activation and cell cycle arrest in cells with severe karyotype abnormalities.
Tumors induced by overexpression of an oncogene will often display extensive cell death upon removal of the oncogene. This is called oncogene addiction, which has been shown for instance for the KRASV12 oncogene55. Sotillo et al. combined inducible expression of the
oncogene KRAS with overexpression of the mitotic checkpoint gene MAD2, which leads to chromosomal instability and aneuploidy thus combining oncogene addiction with chromosomal instability. Both KRASV12 as well as KRASV12; Mad2 overexpressing mice rapidly
developed lung tumors that disappeared upon removal of the KRAS V12 and MAD2
overexpression. However, only the mice in which the KRASV12 oncogene was combined with
chromosomal instability through Mad2 overexpression displayed tumor relapse within 4-11 months after removal of the oncogene and MAD2 overexpression55. Both the KRASV12; Mad2
primary tumors as well as the recurrent tumors were found to be highly aneuploidy and heterogeneous. This suggests that the genetic heterogeneity induced by MAD2 overexpression in the primary tumor provides these tumors with a growth advantage over oncogene-only-induced tumors, even after oncogene and MAD2 overexpression removal. Studies like these shed light on the aneuploidy paradox: aneuploidy can both function as a tumor suppressor by inhibiting growth, as well as promote tumorigenesis and stimulate growth.
Future perspectives
Although single-cell sequencing is causing a revolution in cancer research, we are only at the start of exploring its possibilities. Studies performed so far are mainly case studies, or based on small numbers of patients. Also, limited numbers of cells are typically analyzed for specific cancer types. Therefore, larger scale studies on various cancers are needed to get a clear view on how aneuploidy, CNAs and heterogeneity in the development, progression and treatment resistance of cancer cells. By comparing karyotypes and heterogeneity before and after treatment, we can identify which cells respond and which are resistant to therapy. Recently, a large study investigating intra-tumor heterogeneity in over 300 samples from 100 patients with non-small cell lung cancer was published43. Using multiregion whole exome sequencing
they found widespread regional intra-tumor heterogeneity in both CNAs as well as mutations. An interesting the next step will be to now assess these individual regions at the single cell
6
metastases varied: especially the liver metastasis showed markedly less heterogeneity.Several possibilities could explain this observation. The cells could be optimally suited to grow at the selected site, showing less chromosome missegregation, or chromosomally instable cells could be selected against more stringently. It has been shown that in SCLC patients clumps of tumors cells are often present in the blood47. Metastatic spread can take place via
multiple routes. Metastases can form from one or multiple cells from the primary tumor resulting in mono- or polyclonal seeding, but cells from metastases can also give rise to additional metastases. Also, metastatic cells might migrate back into the primary tumor, called self-seeding, or into another metastasis48. It is more difficult to reconstruct
evolutionary trees based on CNAs than using mutations since it is not a linear process; in contrast to mutations, chromosomes can be gained during one cell division but the same chromosome might be lost again at a later point. Therefore, determining the order of events is often difficult to elucidate.
Aneuploidy in cancer development
It is still not known why tumors are so often aneuploid. Cells can become aneuploid when chromosomes missegregate during cell division. In normal cells mechanisms such as the spindle assembly checkpoint (SAC) are in place to ensure correct chromosome segregation49.
When these mechanisms do not function properly, chromosome missegregation can occur. Also, telomere shortening is known to cause aneuploidy. When telomeres become too short cells will generally exit the cell cycle50. With loss of the tumor suppressor p53 cells can
continue to proliferate, even with critically short telomeres51. This results in end-to-end fusion
of sister chromatid telomeres, and the formation of dicentric chromosomes. Dicentric chromosomes are likely to missegregate during mitosis thus resulting in DNA breaks and aneuploidy, and trigger so-called breakage-fusion-bridge (BFB) cycles. This can continue over many cell divisions, leading to a heterogeneous and aneuploid population of cells with large duplications and deletions52. Aneuploidy is known to inhibit growth, but at the same time it
seems to stimulate growth of cancer cells34,53,54. Studying aneuploidy in cancer will provide
clues on how and at which stage of the tumor it develops and how aneuploidy can stimulate the growth of tumor cells.
A study by the group of Amon emphasizes how aneuploidy might play a dual role in growth suppression and stimulation56. They found that cells harboring a single extra chromosome
showed growth inhibition and increased senescence. Even the activation of growth stimulating oncogenic pathways in combination with single chromosome aneuploidy showed reduced tumorigenic potential as they formed smaller tumors compared to euploid cells with the oncogene activation only. So even in the context of oncogene activation, single chromosome aneuploidy functions as a tumor suppressor. On the other hand, the single chromosome aneuploidy can cause chromosomal instability leading to additional structural and whole chromosome copy number abnormalities. These additional changes compensate for the growth defect and resulted in improved fitness and tumorigenicity suggesting that single aneuploidies could facilitate selection of cells with highly rearranged genomes56.
Indeed, a very recent study by the group of Medema describes the activation of p53 in untransformed cells with whole chromosome or structural aneuploidy. They found that structural aneuploidy does activate p53 and induces cell cycle arrest, but whole chromosome aneuploidy does not induce p53 activation in all cells; many are still able to proliferate with mild whole chromosome imbalances57. However, the fact that the number of whole
chromosome abnormalities decreased in the proliferating population does suggest selection against high levels of aneuploidy in normal cells. This is in line with the observation that complex karyotypes will induce a cell cycle arrest, while cells with low amounts of chromosomal imbalances are often able to continue to proliferate.58 These data suggest that
aneuploidy can trigger a tumor suppressor mechanism by inducing p53 activation and cell cycle arrest in cells with severe karyotype abnormalities.
Tumors induced by overexpression of an oncogene will often display extensive cell death upon removal of the oncogene. This is called oncogene addiction, which has been shown for instance for the KRASV12 oncogene55. Sotillo et al. combined inducible expression of the
oncogene KRAS with overexpression of the mitotic checkpoint gene MAD2, which leads to chromosomal instability and aneuploidy thus combining oncogene addiction with chromosomal instability. Both KRASV12 as well as KRASV12; Mad2 overexpressing mice rapidly
developed lung tumors that disappeared upon removal of the KRAS V12 and MAD2
overexpression. However, only the mice in which the KRASV12 oncogene was combined with
chromosomal instability through Mad2 overexpression displayed tumor relapse within 4-11 months after removal of the oncogene and MAD2 overexpression55. Both the KRASV12; Mad2
primary tumors as well as the recurrent tumors were found to be highly aneuploidy and heterogeneous. This suggests that the genetic heterogeneity induced by MAD2 overexpression in the primary tumor provides these tumors with a growth advantage over oncogene-only-induced tumors, even after oncogene and MAD2 overexpression removal. Studies like these shed light on the aneuploidy paradox: aneuploidy can both function as a tumor suppressor by inhibiting growth, as well as promote tumorigenesis and stimulate growth.
Future perspectives
Although single-cell sequencing is causing a revolution in cancer research, we are only at the start of exploring its possibilities. Studies performed so far are mainly case studies, or based on small numbers of patients. Also, limited numbers of cells are typically analyzed for specific cancer types. Therefore, larger scale studies on various cancers are needed to get a clear view on how aneuploidy, CNAs and heterogeneity in the development, progression and treatment resistance of cancer cells. By comparing karyotypes and heterogeneity before and after treatment, we can identify which cells respond and which are resistant to therapy. Recently, a large study investigating intra-tumor heterogeneity in over 300 samples from 100 patients with non-small cell lung cancer was published43. Using multiregion whole exome sequencing
they found widespread regional intra-tumor heterogeneity in both CNAs as well as mutations. An interesting the next step will be to now assess these individual regions at the single cell
level. Moreover, chromosomal instability mediated copy number heterogeneity was associated with increased risk of recurrence or death, underlining the potential value of heterogeneity in predicting prognosis43. This study emphasized the importance of analyzing
multiple regions of a tumor, both for bulk as well as single cell analysis. One region of the tumor might contain different cells, with different CNAs or mutations than another. Therefore, even single-cell sequencing might not be able to capture the full heterogeneity within a tumor when only cells isolated from one region of the tumor are analyzed.
Circulating tumor cells
Over the recent years, research efforts have begun to transition towards studies on circulating tumor cells (see also Chapter 4). Such studies provide a non-invasive method to identify, characterize and monitor cancer. Methods are being developed to count and isolate circulating tumor cells from the blood, with so far one, the CellSearch system61, clinically
approved by the FDA. For several cancer types it has been shown that the number of circulating tumor cells has prognostic value62. But sequencing these cells can reveal a wealth
of additional information. In combination with analyzing primary tumor and metastasis cells, one can assess whether circulating tumor cells are representative of the primary tumor or possibly more similar to metastases. This is still under debate as some studies concluded that the CNA and mutation patterns of CTCs are similar to the primary tumor63–65, while another
found the CNA pattern to be more homogeneous than the primary tumor and more similar to the metastases66. The CNA pattern found in circulating tumor cells from various cancers
was found to cluster by cancer type66. Therefore, sequencing circulating tumor cells, possibly
in combination with RNA sequencing, can potentially reveal the tissue from which the tumor originated. This information can be used to direct further tests such as imaging to identify the exact location of the tumor. Also, treatment can become more focused as patient groups can be identified which are likely or unlikely to respond to a certain therapy59. A start has been
made to explore this by sequencing circulating tumor cells before and after therapy. This resulted in the identification of resistant clones in prostate cancer60, and prediction of the
treatment response on the basis of CNA pattern of circulating tumor cells in lung cancer59. Technical advances
The combination of decreasing price, increasing throughput and sample ‘types’ suitable for single-cell sequencing, e.g frozen, fixed, or even FFPE samples67, will enable single-cell
sequencing to become a mainstream method. One can imagine that in the future everyone will have regularly some blood drawn to check for circulating tumor cells, or circulating tumor DNA68. The resulting earlier detection, before metastases have developed, will thus lead to
improved survival. If circulating tumor cells are present, they can be sequenced revealing the origin, heterogeneity and mutational landscape of the tumor cells. Besides knowledge about the location of a tumor, this will also give information about treatments that will be most efficient at attacking the tumor and potential metastases. Furthermore, during treatment,
regular blood draws will allow monitoring of treatment efficacy and identification of emerging resistant clones.
In addition to improving single-cell sequencing platforms, the development of better sequencing methods continues as well. Although not (yet) applicable for single-cell sequencing, methods such as nanopore sequencing holds great potential for the (near) future.69 With this method long stretches of DNA are sequenced by having a motor protein
‘pull’ single stranded DNA through a pore in a membrane. A voltage is applied over this membrane. By measuring the voltage changes with every nucleotide passing through the pore, the DNA sequence is read70. A major advantage of this method is the possibility to
sequence long stretches of DNA at low cost. Moreover, the massive reduction in size of sequencers, from ‘room-filling’ to table-top and even the size of a USB stick will enable sequencing to be performed all over the world, and even in space71.
Another important hurdle that soon will be taken is the combination of multiple single cell analysis methods on the same individual cells. For example the development of methods combining RNA and DNA sequencing of the same cell holds great promise,72,73 and will provide
direct insight in how aneuploidy and CNAs translate to expression level changes. Furthermore, combining transcriptome with genome and methylome sequencing74–76, will provide further
insight into the functioning of single cells. Recently, a start has been made in combining transcriptomic and proteomic analysis in single cells.77 While RNA levels give a rough
indication of protein expression, post-transcriptional regulation complicates the correlation between RNA and protein levels. Although the number of RNA-protein combinations is still limited, this type of analysis will lead to more insight into the correlation between RNA and protein levels and the regulatory mechanisms. Together, these methods will provide the possibility to generate a complete picture of individual cells, from aneuploidy, CNA profiles and mutation, to transcription (regulation) and protein expression, although further optimization is needed to reduce background noise and increase coverage.
In conclusion, we are only at the start of exploring the genetic heterogeneity in normal and cancer cells. Continuing development of technical and analytical tools will help to generate a wealth of information on the diversity, functioning and predicting response of cell populations.
References
1. Dahm, R. Friedrich Miescher and the discovery of DNA. Developmental Biology 278, 274–288 (2005).
2. Flemming, W. Ueber das Verhalten des Kerns bei der Zelltheilung, und ??ber die Bedeutung mehrkerniger Zellen. Arch. f??r Pathol. Anat. und Physiol. und f??r Klin. Med. 77, 1–29 (1879). 3. Boveri, T. Ueber mehrpolige Mitosen als Mittel zur Analyse des Zellkerns. Verh Phys-med Ges
Würzbg. NF 35, 67–90 (1902).
4. SUTTON, W. S. ON THE MORPHOLOGY OF THE CHROMOSO GROUP IN BRACHYSTOLA MAGNA. Biol. Bull. 4, 24–39 (1902).
5. Boveri, T. Zur Frage der Entstehung maligner Tumoren. Science (80-. ). 40, 857–859 (1914). 6. WATSON, J. D. & CRICK, F. H. Molecular structure of nucleic acids; a structure for deoxyribose
6
level. Moreover, chromosomal instability mediated copy number heterogeneity wasassociated with increased risk of recurrence or death, underlining the potential value of heterogeneity in predicting prognosis43. This study emphasized the importance of analyzing
multiple regions of a tumor, both for bulk as well as single cell analysis. One region of the tumor might contain different cells, with different CNAs or mutations than another. Therefore, even single-cell sequencing might not be able to capture the full heterogeneity within a tumor when only cells isolated from one region of the tumor are analyzed.
Circulating tumor cells
Over the recent years, research efforts have begun to transition towards studies on circulating tumor cells (see also Chapter 4). Such studies provide a non-invasive method to identify, characterize and monitor cancer. Methods are being developed to count and isolate circulating tumor cells from the blood, with so far one, the CellSearch system61, clinically
approved by the FDA. For several cancer types it has been shown that the number of circulating tumor cells has prognostic value62. But sequencing these cells can reveal a wealth
of additional information. In combination with analyzing primary tumor and metastasis cells, one can assess whether circulating tumor cells are representative of the primary tumor or possibly more similar to metastases. This is still under debate as some studies concluded that the CNA and mutation patterns of CTCs are similar to the primary tumor63–65, while another
found the CNA pattern to be more homogeneous than the primary tumor and more similar to the metastases66. The CNA pattern found in circulating tumor cells from various cancers
was found to cluster by cancer type66. Therefore, sequencing circulating tumor cells, possibly
in combination with RNA sequencing, can potentially reveal the tissue from which the tumor originated. This information can be used to direct further tests such as imaging to identify the exact location of the tumor. Also, treatment can become more focused as patient groups can be identified which are likely or unlikely to respond to a certain therapy59. A start has been
made to explore this by sequencing circulating tumor cells before and after therapy. This resulted in the identification of resistant clones in prostate cancer60, and prediction of the
treatment response on the basis of CNA pattern of circulating tumor cells in lung cancer59. Technical advances
The combination of decreasing price, increasing throughput and sample ‘types’ suitable for single-cell sequencing, e.g frozen, fixed, or even FFPE samples67, will enable single-cell
sequencing to become a mainstream method. One can imagine that in the future everyone will have regularly some blood drawn to check for circulating tumor cells, or circulating tumor DNA68. The resulting earlier detection, before metastases have developed, will thus lead to
improved survival. If circulating tumor cells are present, they can be sequenced revealing the origin, heterogeneity and mutational landscape of the tumor cells. Besides knowledge about the location of a tumor, this will also give information about treatments that will be most efficient at attacking the tumor and potential metastases. Furthermore, during treatment,
regular blood draws will allow monitoring of treatment efficacy and identification of emerging resistant clones.
In addition to improving single-cell sequencing platforms, the development of better sequencing methods continues as well. Although not (yet) applicable for single-cell sequencing, methods such as nanopore sequencing holds great potential for the (near) future.69 With this method long stretches of DNA are sequenced by having a motor protein
‘pull’ single stranded DNA through a pore in a membrane. A voltage is applied over this membrane. By measuring the voltage changes with every nucleotide passing through the pore, the DNA sequence is read70. A major advantage of this method is the possibility to
sequence long stretches of DNA at low cost. Moreover, the massive reduction in size of sequencers, from ‘room-filling’ to table-top and even the size of a USB stick will enable sequencing to be performed all over the world, and even in space71.
Another important hurdle that soon will be taken is the combination of multiple single cell analysis methods on the same individual cells. For example the development of methods combining RNA and DNA sequencing of the same cell holds great promise,72,73 and will provide
direct insight in how aneuploidy and CNAs translate to expression level changes. Furthermore, combining transcriptome with genome and methylome sequencing74–76, will provide further
insight into the functioning of single cells. Recently, a start has been made in combining transcriptomic and proteomic analysis in single cells.77 While RNA levels give a rough
indication of protein expression, post-transcriptional regulation complicates the correlation between RNA and protein levels. Although the number of RNA-protein combinations is still limited, this type of analysis will lead to more insight into the correlation between RNA and protein levels and the regulatory mechanisms. Together, these methods will provide the possibility to generate a complete picture of individual cells, from aneuploidy, CNA profiles and mutation, to transcription (regulation) and protein expression, although further optimization is needed to reduce background noise and increase coverage.
In conclusion, we are only at the start of exploring the genetic heterogeneity in normal and cancer cells. Continuing development of technical and analytical tools will help to generate a wealth of information on the diversity, functioning and predicting response of cell populations.
References
1. Dahm, R. Friedrich Miescher and the discovery of DNA. Developmental Biology 278, 274–288 (2005).
2. Flemming, W. Ueber das Verhalten des Kerns bei der Zelltheilung, und ??ber die Bedeutung mehrkerniger Zellen. Arch. f??r Pathol. Anat. und Physiol. und f??r Klin. Med. 77, 1–29 (1879). 3. Boveri, T. Ueber mehrpolige Mitosen als Mittel zur Analyse des Zellkerns. Verh Phys-med Ges
Würzbg. NF 35, 67–90 (1902).
4. SUTTON, W. S. ON THE MORPHOLOGY OF THE CHROMOSO GROUP IN BRACHYSTOLA MAGNA. Biol. Bull. 4, 24–39 (1902).
5. Boveri, T. Zur Frage der Entstehung maligner Tumoren. Science (80-. ). 40, 857–859 (1914). 6. WATSON, J. D. & CRICK, F. H. Molecular structure of nucleic acids; a structure for deoxyribose
nucleic acid. Nature 171, 737–8 (1953).
7. WILKINS, M. H., SEEDS, W. E., STOKES, A. R. & WILSON, H. R. Helical structure of crystalline deoxypentose nucleic acid. Nature 172, 759–62 (1953).
8. FRANKLIN, R. E. & GOSLING, R. G. Evidence for 2-chain helix in crystalline structure of sodium deoxyribonucleate. Nature 172, 156–7 (1953).
9. Sanger, F., Nicklen, S. & Coulson, A. R. DNA sequencing with chain-terminating inhibitors.
Proc. Natl. Acad. Sci. U. S. A. 74, 5463–7 (1977).
10. Anderson, S. Shotgun DNA sequencing using cloned DNase I-generated fragments. Nucleic
Acids Res. 9, 3015–27 (1981).
11. Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860–921 (2001).
12. Venter, J. C. et al. The Sequence of the Human Genome. Science (80-. ). 291, 1304–1351 (2001).
13. Margulies, M., Egholm, M., Altman, W., Attiya, S. & Bader, J. Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380 (2005).
14. Voelkerding, K. V., Dames, S. A. & Durtschi, J. D. Next-Generation Sequencing: From Basic Research to Diagnostics. Clin. Chem. 55, 641–658 (2009).
15. Dunham, I. et al. An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57–74 (2012).
16. The Cancer Genome Atlas. Available at: https://cancergenome.nih.gov/. (Accessed: 3rd July 2017)
17. Speicher, M. R. & Carter, N. P. The new cytogenetics: blurring the boundaries with molecular biology. Nat. Rev. Genet. 6, 782–92 (2005).
18. Pellestor, F., Anahory, T. & Hamamah, S. The chromosomal analysis of human oocytes. An overview of established procedures. Hum. Reprod. Update 11, 15–32 (2005).
19. Riegel, M. Human molecular cytogenetics : From cells to nucleotides. Genet. Mol. Biol. 37, 194–209 (2014).
20. Navin, N. et al. Tumour evolution inferred by single-cell sequencing. Nature 472, 90–94 (2011).
21. Baslan, T. et al. Genome-wide copy number analysis of single cells. Nat. Protoc. 7, 1024–1041 (2012).
22. Gawad, C., Koh, W. & Quake, S. R. Single-cell genome sequencing: current state of the science. Nat. Rev. Genet. 17, 175–188 (2016).
23. van den Bos, H. et al. Single-cell whole genome sequencing reveals no evidence for common aneuploidy in normal and Alzheimer’s disease neurons. Genome Biol. 17, 116 (2016). 24. Zahn, H. et al. Scalable whole-genome single-cell library preparation without
pre-amplification. Nat. Methods In press, 1–39 (2017).
25. Rehen, S. K. et al. Chromosomal variation in neurons of the developing and adult mammalian nervous system. Proc. Natl. Acad. Sci. U. S. A. 98, 13361–6 (2001).
26. Yang, A. H. et al. Chromosome segregation defects contribute to aneuploidy in normal neural progenitor cells. J. Neurosci. 23, 10454–10462 (2003).
27. Kingsbury, M. et al. Aneuploid neurons are functionally active and integrated into brain circuitry. Proc. Natl. Acad. Sci. U. S. A. 102, 6143–7 (2005).
28. Yurov, Y. B. et al. The variation of aneuploidy frequency in the developing and adult human brain revealed by an interphase FISH study. J. Histochem. Cytochem. 53, 385–390 (2005).
29. Yurov, Y. B. et al. Aneuploidy and confined chromosomal mosaicism in the developing human brain. PLoS One 2, (2007).
30. Rehen, S. K. et al. Constitutional Aneuploidy in the Normal Human Brain. 25, 2176–2180 (2005).
31. Yang, Y., Geldmacher, D. S. & Herrup, K. DNA replication precedes neuronal cell death in Alzheimer’s disease. J. Neurosci. 21, 2661–2668 (2001).
32. Pack, S. D. et al. Individual adult human neurons display aneuploidy: Detection by fluorescence in situ hybridization and single neuron PCR. Cell Cycle 4, 1758–1760 (2005). 33. Mosch, B. et al. Aneuploidy and DNA replication in the normal human brain and Alzheimer’s
disease. J. Neurosci. 27, 6859–67 (2007).
34. Sheltzer, J. M. & Amon, A. The aneuploidy paradox: Costs and benefits of an incorrect
karyotype. Trends in Genetics 27, 446–453 (2011).
35. Yurov, Y. B., Vostrikov, V. M., Vorsanova, S. G., Monakhov, V. V. & Iourov, I. Y. Multicolor fluorescent in situ hybridization on post-mortem brain in schizophrenia as an approach for identification of low-level chromosomal aneuploidy in neuropsychiatric diseases. in Brain and
Development 23, S186–S190 (2001).
36. McConnell, M. J. et al. Mosaic Copy Number Variation in Human Neurons. Science (80-. ). 342, 632–633 (2013).
37. Cai, X. et al. Single-Cell, Genome-wide Sequencing Identifies Clonal Somatic Copy-Number Variation in the Human Brain. Cell Rep. 8, 1280–1289 (2014).
38. Knouse, K. a., Wu, J., Whittaker, C. a. & Amon, A. Single cell sequencing reveals low levels of aneuploidy across mammalian tissues. Proc. Natl. Acad. Sci. 111, 1–6 (2014).
39. Lista, S., Dubois, B. & Hampel, H. Paths to Alzheimer’s disease prevention: From modifiable risk factors to biomarker enrichment strategies. J. Nutr. Heal. Aging 19, 154–163 (2015). 40. Foijer, F., Draviam, V. M. & Sorger, P. K. Studying chromosome instability in the mouse.
Biochim. Biophys. Acta - Rev. Cancer 1786, 73–82 (2008).
41. Navin, N. E. The first five years of single-cell cancer genomics and beyond. Genome Research 25, 1499–1507 (2015).
42. Gao, R. et al. Punctuated copy number evolution and clonal stasis in triple-negative breast cancer. Nat. Genet. 48, 1–15 (2016).
43. Jamal-Hanjani, D. et al. Tracking the Evolution of Non–Small-Cell Lung Cancer. (2017). doi:10.1056/NEJMoa1616288
44. Leung, M. L. et al. Single cell DNA sequencing reveals a late-dissemination model in
metastatic colorectal cancer. Genome Res. gr.209973.116 (2017). doi:10.1101/gr.209973.116 45. Gundem, G. et al. The evolutionary history of lethal metastatic prostate cancer. Nature 520,
353–357 (2015).
46. McPherson, A. et al. Divergent modes of clonal spread and intraperitoneal mixing in high-grade serous ovarian cancer. Nat. Genet. 48, 758–67 (2016).
47. Hou, J. M. et al. Clinical significance and molecular characteristics of circulating tumor cells and circulating tumor microemboli in patients with small-cell lung cancer. J. Clin. Oncol. 30, 525–532 (2012).
48. Turajlic, S. et al. Metastasis as an evolutionary process. Science 352, 169–75 (2016). 49. Musacchio, A. & Salmon, E. D. The spindle-assembly checkpoint in space and time. Mol. Cell
8, 379–393 (2007).
6
nucleic acid. Nature 171, 737–8 (1953).7. WILKINS, M. H., SEEDS, W. E., STOKES, A. R. & WILSON, H. R. Helical structure of crystalline deoxypentose nucleic acid. Nature 172, 759–62 (1953).
8. FRANKLIN, R. E. & GOSLING, R. G. Evidence for 2-chain helix in crystalline structure of sodium deoxyribonucleate. Nature 172, 156–7 (1953).
9. Sanger, F., Nicklen, S. & Coulson, A. R. DNA sequencing with chain-terminating inhibitors.
Proc. Natl. Acad. Sci. U. S. A. 74, 5463–7 (1977).
10. Anderson, S. Shotgun DNA sequencing using cloned DNase I-generated fragments. Nucleic
Acids Res. 9, 3015–27 (1981).
11. Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860–921 (2001).
12. Venter, J. C. et al. The Sequence of the Human Genome. Science (80-. ). 291, 1304–1351 (2001).
13. Margulies, M., Egholm, M., Altman, W., Attiya, S. & Bader, J. Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380 (2005).
14. Voelkerding, K. V., Dames, S. A. & Durtschi, J. D. Next-Generation Sequencing: From Basic Research to Diagnostics. Clin. Chem. 55, 641–658 (2009).
15. Dunham, I. et al. An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57–74 (2012).
16. The Cancer Genome Atlas. Available at: https://cancergenome.nih.gov/. (Accessed: 3rd July 2017)
17. Speicher, M. R. & Carter, N. P. The new cytogenetics: blurring the boundaries with molecular biology. Nat. Rev. Genet. 6, 782–92 (2005).
18. Pellestor, F., Anahory, T. & Hamamah, S. The chromosomal analysis of human oocytes. An overview of established procedures. Hum. Reprod. Update 11, 15–32 (2005).
19. Riegel, M. Human molecular cytogenetics : From cells to nucleotides. Genet. Mol. Biol. 37, 194–209 (2014).
20. Navin, N. et al. Tumour evolution inferred by single-cell sequencing. Nature 472, 90–94 (2011).
21. Baslan, T. et al. Genome-wide copy number analysis of single cells. Nat. Protoc. 7, 1024–1041 (2012).
22. Gawad, C., Koh, W. & Quake, S. R. Single-cell genome sequencing: current state of the science. Nat. Rev. Genet. 17, 175–188 (2016).
23. van den Bos, H. et al. Single-cell whole genome sequencing reveals no evidence for common aneuploidy in normal and Alzheimer’s disease neurons. Genome Biol. 17, 116 (2016). 24. Zahn, H. et al. Scalable whole-genome single-cell library preparation without
pre-amplification. Nat. Methods In press, 1–39 (2017).
25. Rehen, S. K. et al. Chromosomal variation in neurons of the developing and adult mammalian nervous system. Proc. Natl. Acad. Sci. U. S. A. 98, 13361–6 (2001).
26. Yang, A. H. et al. Chromosome segregation defects contribute to aneuploidy in normal neural progenitor cells. J. Neurosci. 23, 10454–10462 (2003).
27. Kingsbury, M. et al. Aneuploid neurons are functionally active and integrated into brain circuitry. Proc. Natl. Acad. Sci. U. S. A. 102, 6143–7 (2005).
28. Yurov, Y. B. et al. The variation of aneuploidy frequency in the developing and adult human brain revealed by an interphase FISH study. J. Histochem. Cytochem. 53, 385–390 (2005).
29. Yurov, Y. B. et al. Aneuploidy and confined chromosomal mosaicism in the developing human brain. PLoS One 2, (2007).
30. Rehen, S. K. et al. Constitutional Aneuploidy in the Normal Human Brain. 25, 2176–2180 (2005).
31. Yang, Y., Geldmacher, D. S. & Herrup, K. DNA replication precedes neuronal cell death in Alzheimer’s disease. J. Neurosci. 21, 2661–2668 (2001).
32. Pack, S. D. et al. Individual adult human neurons display aneuploidy: Detection by fluorescence in situ hybridization and single neuron PCR. Cell Cycle 4, 1758–1760 (2005). 33. Mosch, B. et al. Aneuploidy and DNA replication in the normal human brain and Alzheimer’s
disease. J. Neurosci. 27, 6859–67 (2007).
34. Sheltzer, J. M. & Amon, A. The aneuploidy paradox: Costs and benefits of an incorrect
karyotype. Trends in Genetics 27, 446–453 (2011).
35. Yurov, Y. B., Vostrikov, V. M., Vorsanova, S. G., Monakhov, V. V. & Iourov, I. Y. Multicolor fluorescent in situ hybridization on post-mortem brain in schizophrenia as an approach for identification of low-level chromosomal aneuploidy in neuropsychiatric diseases. in Brain and
Development 23, S186–S190 (2001).
36. McConnell, M. J. et al. Mosaic Copy Number Variation in Human Neurons. Science (80-. ). 342, 632–633 (2013).
37. Cai, X. et al. Single-Cell, Genome-wide Sequencing Identifies Clonal Somatic Copy-Number Variation in the Human Brain. Cell Rep. 8, 1280–1289 (2014).
38. Knouse, K. a., Wu, J., Whittaker, C. a. & Amon, A. Single cell sequencing reveals low levels of aneuploidy across mammalian tissues. Proc. Natl. Acad. Sci. 111, 1–6 (2014).
39. Lista, S., Dubois, B. & Hampel, H. Paths to Alzheimer’s disease prevention: From modifiable risk factors to biomarker enrichment strategies. J. Nutr. Heal. Aging 19, 154–163 (2015). 40. Foijer, F., Draviam, V. M. & Sorger, P. K. Studying chromosome instability in the mouse.
Biochim. Biophys. Acta - Rev. Cancer 1786, 73–82 (2008).
41. Navin, N. E. The first five years of single-cell cancer genomics and beyond. Genome Research 25, 1499–1507 (2015).
42. Gao, R. et al. Punctuated copy number evolution and clonal stasis in triple-negative breast cancer. Nat. Genet. 48, 1–15 (2016).
43. Jamal-Hanjani, D. et al. Tracking the Evolution of Non–Small-Cell Lung Cancer. (2017). doi:10.1056/NEJMoa1616288
44. Leung, M. L. et al. Single cell DNA sequencing reveals a late-dissemination model in
metastatic colorectal cancer. Genome Res. gr.209973.116 (2017). doi:10.1101/gr.209973.116 45. Gundem, G. et al. The evolutionary history of lethal metastatic prostate cancer. Nature 520,
353–357 (2015).
46. McPherson, A. et al. Divergent modes of clonal spread and intraperitoneal mixing in high-grade serous ovarian cancer. Nat. Genet. 48, 758–67 (2016).
47. Hou, J. M. et al. Clinical significance and molecular characteristics of circulating tumor cells and circulating tumor microemboli in patients with small-cell lung cancer. J. Clin. Oncol. 30, 525–532 (2012).
48. Turajlic, S. et al. Metastasis as an evolutionary process. Science 352, 169–75 (2016). 49. Musacchio, A. & Salmon, E. D. The spindle-assembly checkpoint in space and time. Mol. Cell
8, 379–393 (2007).
