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.
4
Klaske M. Schukken1, Niels Kloosterhuis2, Klaas Sjollema3, Petra Bakker1, Bart van de Sluis2, and Floris Foijer1
1European Research Institute for the Biology of Ageing (ERIBA), University of Groningen, University Medical Center Groningen,
9713 AV, Groningen, the Netherlands
2Transgenic Mouse Clinic, University of Groningen, University Medical Center Groningen, 9713 AV, Groningen, the Netherlands
3Imaging Facility, University of Groningen, University Medical Center Groningen, 9713 AV, Groningen, the Netherlands
CIN tracker- a novel mouse model
to monitor mitosis in vivo
Chapter 4
4
4
CIN tracker- a novel mouse model to monitor mitosis
in vivo
Klaske M. Schukken1, Niels Kloosterhuis2, Klaas Sjollema3, Petra Bakker1, Bart
van de Sluis2, and Floris Foijer1
1European Research Institute for the Biology of Ageing (ERIBA), University of
Groningen, University Medical Center Groningen, 9713 AV, Groningen, the Netherlands
2Transgenic Mouse Clinic, University of Groningen, University Medical Center
Groningen, 9713 AV, Groningen, the Netherlands
3 Imaging Facility, University of Groningen, University Medical Center
Groningen, 9713 AV, Groningen, the Netherlands
Abstract Chromosome Instability (CIN) is a hallmark of cancer. Despite this, CIN can have paradoxically variable effects on cell survival and tumor progression. While most cancers are aneuploid, and many types of CIN enhance tumorigenesis, other drivers of CIN appear to inhibit tumor progression. While the rates of chromosome mis‐segregation, the most common form of CIN, are frequently studied in cell culture, the rate of mis‐segregation in vivo is rarely studied due to technical difficulties and a lack of proper mouse models. Here, we engineer and validate a novel mouse model called the “CIN tracker”. This mouse models allows for monitoring nuclear and cytoplasmic markers in vivo, which will allow researchers to observe in vivo CIN rates. The CIN tracker mouse model uses the Cre‐LoxP system to induce expression of fluorescent fusion proteins, allowing for fluorescence expression at a time‐point of choice, tissue specific activation, and long term cell tracking. This novel mouse model labels both chromatin (H2B‐ GFP) and α‐tubulin (mTurquoise2173‐αTubulin), so both DNA (mis‐ )segregation and the mitotic spindle can be monitored during mitosis. Going forward, the CIN tracker mouse model can be used to visualize CIN rates in various tissues and tumors, to better understand which types and rates of chromosome mis‐segregation enhance or inhibit tumor progression. Keywords: CIN, fluorescence, mouse model, cancer Introduction Chromosome INstability (CIN) is most accurately measured by monitoring and quantifying chromosome mis‐segregation rates in living cells1. While mis‐segregation rates have been measured in all types of cell lines, and even in organoids174, information about mis‐segregation rates within living tissues is limited71. Chromosome mis‐segregation rates in cell lines are rarely below ~3%, even in diploid cells, and therefore not classified as CIN51,52,57,97 (Also see Chapter 3 of this thesis). However, recent karyotyping methods have shown that there is only very little aneuploidy in healthy human tissues91, and that aneuploidy increases significantly upon tissue expansion in vitro24, indicating that cell culture CIN rates may not be representative of in vivo conditions, or that in vivo aneuploid cells are actively being cleared after their formation. Since CIN is a hallmark of both cancer39 and aging103, observing rates of CIN within living tissue, and
tracking the fate of the resulting cells may give more insight into the roles of CIN and aneuploidy within an organism. So far, mitosis has most commonly been monitored in vivo by injecting fluorescent cells into immune compromised mice71, this method ignores many effects of the immune system and only allows for the visualization of cells that have been cultured and transduced ex vivo. While mouse models exist in which we can monitor cell division via transgenic fluorescent fusion proteins constructs, however these models are either not tissue specific175, or use a tet‐TRE promoter, which requires continuous doxycycline application to maintain fluorescent protein expression176,177. In this chapter we describe the development of a new mouse model henceforth called the “CIN Tracker” mouse that can be used to visualize chromosome mis‐segregation in vivo, and track the fate of the resulting cells. The advantage of this mouse model is that the mitotic reporters can be induced in a tissue specific manner at a time‐point of choice. After induction, the switched cells and all their progeny remain fluorescent. With this mouse model, it becomes possible to monitor cell division in vivo without it being an endpoint measurement for either the cell or the mouse, allowing for the tracking of cell fate over time. Additionally, this mouse model fluorescently labels multiple parts of a cell, including the chromatin, spindle network, and kinetochores, allowing us to better visualize and quantify cell division.
1
4
Abstract Chromosome Instability (CIN) is a hallmark of cancer. Despite this, CIN can have paradoxically variable effects on cell survival and tumor progression. While most cancers are aneuploid, and many types of CIN enhance tumorigenesis, other drivers of CIN appear to inhibit tumor progression. While the rates of chromosome mis‐segregation, the most common form of CIN, are frequently studied in cell culture, the rate of mis‐segregation in vivo is rarely studied due to technical difficulties and a lack of proper mouse models. Here, we engineer and validate a novel mouse model called the “CIN tracker”. This mouse models allows for monitoring nuclear and cytoplasmic markers in vivo, which will allow researchers to observe in vivo CIN rates. The CIN tracker mouse model uses the Cre‐LoxP system to induce expression of fluorescent fusion proteins, allowing for fluorescence expression at a time‐point of choice, tissue specific activation, and long term cell tracking. This novel mouse model labels both chromatin (H2B‐ GFP) and α‐tubulin (mTurquoise2173‐αTubulin), so both DNA (mis‐ )segregation and the mitotic spindle can be monitored during mitosis. Going forward, the CIN tracker mouse model can be used to visualize CIN rates in various tissues and tumors, to better understand which types and rates of chromosome mis‐segregation enhance or inhibit tumor progression. Keywords: CIN, fluorescence, mouse model, cancer Introduction Chromosome INstability (CIN) is most accurately measured by monitoring and quantifying chromosome mis‐segregation rates in living cells1. While mis‐segregation rates have been measured in all types of cell lines, and even in organoids174, information about mis‐segregation rates within living tissues is limited71. Chromosome mis‐segregation rates in cell lines are rarely below ~3%, even in diploid cells, and therefore not classified as CIN51,52,57,97 (Also see Chapter 3 of this thesis). However, recent karyotyping methods have shown that there is only very little aneuploidy in healthy human tissues91, and that aneuploidy increases significantly upon tissue expansion in vitro24, indicating that cell culture CIN rates may not be representative of in vivo conditions, or that in vivo aneuploid cells are actively being cleared after their formation. Since CIN is a hallmark of both cancer39 and aging103, observing rates of CIN within living tissue, andtracking the fate of the resulting cells may give more insight into the roles of CIN and aneuploidy within an organism. So far, mitosis has most commonly been monitored in vivo by injecting fluorescent cells into immune compromised mice71, this method ignores many effects of the immune system and only allows for the visualization of cells that have been cultured and transduced ex vivo. While mouse models exist in which we can monitor cell division via transgenic fluorescent fusion proteins constructs, however these models are either not tissue specific175, or use a tet‐TRE promoter, which requires continuous doxycycline application to maintain fluorescent protein expression176,177. In this chapter we describe the development of a new mouse model henceforth called the “CIN Tracker” mouse that can be used to visualize chromosome mis‐segregation in vivo, and track the fate of the resulting cells. The advantage of this mouse model is that the mitotic reporters can be induced in a tissue specific manner at a time‐point of choice. After induction, the switched cells and all their progeny remain fluorescent. With this mouse model, it becomes possible to monitor cell division in vivo without it being an endpoint measurement for either the cell or the mouse, allowing for the tracking of cell fate over time. Additionally, this mouse model fluorescently labels multiple parts of a cell, including the chromatin, spindle network, and kinetochores, allowing us to better visualize and quantify cell division.
Results Creating CIN tracking cell lines To visualize mitosis, we set out to fluorescently label chromatin (H2B), the spindle network (α tubulin) and either kinetochores (CenpB) or centrioles (Centrin3) (Figure 1A). To minimize the number of independent genetic constructs needed to visualize multiple aspects of chromosome segregation, three genes were cloned into a single vector with a single promoter. Genes were linked together with T2A or P2A178,179 sequences, a self‐cleaving sequence which cleaves at the protein level allowing for polycistronic expression from a single promoter. Two combinations of fluorescent proteins were created to visualize mitosis: the CenpB tri‐ fluorescent vector labeling chromatin (H2B), the spindle (α‐tubulin) and kinetochores (CenpB), and the Centrin3 tri‐fluorescent vector labeling chromatin (H2B), the spindle (α‐tubulin) and the centrioles (Centrin3). To test functionality of the engineered constructs, RPE1 cells were transduced with rtTA & Dox inducible fluorescence expression system and used to visualize fluorescence localization using time‐lapse imaging of cultured cells (Figure 1B and C). This confirmed that all three fluorescent proteins localized correctly for both the CenpB tri‐fluorescent vector and the Centrin3 tri‐fluorescent vector. We also quantified chromosome mis‐ segregation events of the labeled RPE1 cells to rule out any effects of the constructs on mitotic fidelity. While the CenpB tri‐fluorescent marker showed no significant increase in chromosome mis‐segregation rates relative to the control (p‐value= 0.06, ns), the Centrin3 tri‐fluorescent marker did show a significant increase (p‐value ≤ 1E‐5, Figure 1D). Expression of the Centrin3 tri‐fluorescent marker provoked notably more multi‐polar spindles and lead to extended mitotic timing (Figure 1D i), as well as to an increased number of micronuclei in interphase cells (p‐value≤ 1E‐5, Figure 1D ii). Since we want to quantify naturally occurring mis‐ segregation rates in vivo, we decided to not continue with the Centrin3 tri‐ fluorescent cells. Creating CIN tracker mice Many mouse models use the Cre ‐ loxP system to activate or delete genes. The advantage of Cre is that it can be expressed in a tissue specific manner, or can be induced via drug‐mediated activation; this allows for tissue Figure 1: CIN tracking cell lines. A) Diagram of the two CIN tracker cell lines expressing H2B‐ GFP, mTurquoise2‐αTubulin and i) CenpB‐mCherry or ii) mCherry‐Centrin3. B) RPE1 cells expressing CenpB‐mCherry, H2B‐GFP and mTurquoise2‐ αTubulin. Cells in i) interphase or ii) Mitosis. C) RPE1 cells expressing mCherry‐Centrin3, H2B‐GFP and mTurquoise2‐ αTubulin. H2B(Histones) Tubulin CenpB (kinetochore) Centrin3 (centrioles) A) i) ii) B) i) C) i) D) i) ii) 0.00 0.25 0.50 0.75 1.00 � � Induc ible C enpB Induc ible C entrin 3 freq (int er pha se ph en otyp e) Type of interphase good interphase micronuclei m�nuclei tetraploid dying/dead cell Nuclear morphology ns0.00001 0.00 0.25 0.50 0.75 1.00 � � Induc ible C enpB Induc ible C entrin 3 freq(in ter pha se p hen oty
pe) Type of Mitosis
good mitosis anaphase bridge extended mitosis multipolar spindle other ns0.00001 n= 244 180 59 n=5510 1616 1644 ii) ii)
CenpB H2B Tubulin Merge
1
4
Results Creating CIN tracking cell lines To visualize mitosis, we set out to fluorescently label chromatin (H2B), the spindle network (α tubulin) and either kinetochores (CenpB) or centrioles (Centrin3) (Figure 1A). To minimize the number of independent genetic constructs needed to visualize multiple aspects of chromosome segregation, three genes were cloned into a single vector with a single promoter. Genes were linked together with T2A or P2A178,179 sequences, a self‐cleaving sequence which cleaves at the protein level allowing for polycistronic expression from a single promoter. Two combinations of fluorescent proteins were created to visualize mitosis: the CenpB tri‐ fluorescent vector labeling chromatin (H2B), the spindle (α‐tubulin) and kinetochores (CenpB), and the Centrin3 tri‐fluorescent vector labeling chromatin (H2B), the spindle (α‐tubulin) and the centrioles (Centrin3). To test functionality of the engineered constructs, RPE1 cells were transduced with rtTA & Dox inducible fluorescence expression system and used to visualize fluorescence localization using time‐lapse imaging of cultured cells (Figure 1B and C). This confirmed that all three fluorescent proteins localized correctly for both the CenpB tri‐fluorescent vector and the Centrin3 tri‐fluorescent vector. We also quantified chromosome mis‐ segregation events of the labeled RPE1 cells to rule out any effects of the constructs on mitotic fidelity. While the CenpB tri‐fluorescent marker showed no significant increase in chromosome mis‐segregation rates relative to the control (p‐value= 0.06, ns), the Centrin3 tri‐fluorescent marker did show a significant increase (p‐value ≤ 1E‐5, Figure 1D). Expression of the Centrin3 tri‐fluorescent marker provoked notably more multi‐polar spindles and lead to extended mitotic timing (Figure 1D i), as well as to an increased number of micronuclei in interphase cells (p‐value≤ 1E‐5, Figure 1D ii). Since we want to quantify naturally occurring mis‐ segregation rates in vivo, we decided to not continue with the Centrin3 tri‐ fluorescent cells. Creating CIN tracker mice Many mouse models use the Cre ‐ loxP system to activate or delete genes. The advantage of Cre is that it can be expressed in a tissue specific manner, or can be induced via drug‐mediated activation; this allows for tissue Figure 1: CIN tracking cell lines. A) Diagram of the two CIN tracker cell lines expressing H2B‐ GFP, mTurquoise2‐αTubulin and i) CenpB‐mCherry or ii) mCherry‐Centrin3. B) RPE1 cells expressing CenpB‐mCherry, H2B‐GFP and mTurquoise2‐ αTubulin. Cells in i) interphase or ii) Mitosis. C) RPE1 cells expressing mCherry‐Centrin3, H2B‐GFP and mTurquoise2‐ αTubulin. H2B(Histones) Tubulin CenpB (kinetochore) Centrin3 (centrioles) A) i) ii) B) i) C) i) D) i) ii) 0.00 0.25 0.50 0.75 1.00 H2B- GFP ControlInduci ble CenpB Induci ble Cent rin3 freq(inte rphase phenotype) Type of interphase good interphase micronuclei multi- nuclei tetraploid dying/dead cell Nuclear morphology ns 0.00001 0.00 0.25 0.50 0.75 1.00 H2B- GFP ControlInduci ble CenpB Induci ble Cent rin3 freq(inte rphase phenotype) Type of Mitosis good mitosis anaphase bridge extended mitosis multipolar spindle other ns 0.00001 n= 244 180 59 n=5510 1616 1644 ii) ii)CenpB H2B Tubulin Merge
Cells in i) interphase or ii) Mitosis. Zoom in of centrioles displayed on the right. D) Quantification of CIN phenotypes from live cell imaging experiments of RPE1 cells expressing either H2B‐GFP, or an inducible form of the tri‐fluorescent vectors. “Inducible CenpB”= H2B‐ GFP‐T2A‐CenpB‐mCherry‐T2A‐mTorquoise2‐Tubulin, “inducible centrin3”= H2B‐GFP‐T2A‐ mCherry‐Centrin3‐T2A‐mTorquoise2‐Tubulin. P‐values from chi squared test displayed. Number of (i) mitotic cells or (ii) interphase cells counter per condition displayed. specific and /or temporal control of genetic switching. When wildtype loxP sites are used, Cre recombinase leads to a permanent change in the genomic DNA, so the ‘switched’ cells pass the genomic alteration on to their daughter cells. As we want to label cells permanently with the mitotic marker, but only in a selection of cells (tissue/time specific), we engineered a vector with a Cre inducible fluorescent marker. The marker contained fluorescent fusions of H2B, CenpB and tubulin (Figure 2A), and genes were separated by a self‐cleaving T2A sequence, similar to what is described above. To generate mouse ES cells for blastocyst injection and transgenic mouse derivation, the vector was linearized, electorporated into mouse Embryonic Stem Cells (mESC) and screened for correct gene expression (Figure 2B). A total of 96 single cell‐derived mESC colonies were plated, grown, split and transduced with Cre adenovirus to test expression of the transgenic construct. Adeno‐Cre‐transduced cells were subjected to time‐ lapse imaging to quantify GFP expression, and GFP positive cells were evaluated for H2B‐eGFP, CenpB‐mCherry and mTurquoise2173‐Tubulin protein expression (Figure 2C, D I and ii not all clones shown). While many clones showed correct expression and localization of H2B and tubulin, the expression of CenpB was weak and sporadic. Even within a single clone, CenpB expression was highly variable between cells. While multiple transgene integrations would lead to higher fluorescence expression, this increased expression would be lost in mice as the transgene would be integrated in multiple chromosomes, which would partly be lost due to asymmetric inheritance in offspring. Therefore, qPCR was used to identify ES cell clones with a single vector integration, and the clones with multiple integrations were discarded (Figure 2E). These analyses lead us to select clone A10, a mESC clone with a single vector integration, correct fluorescence localization and high fluorescent protein expression. An early passage of this mESC clone that did not yet express Cre recombinase was injected into mouse blastocysts, followed by transplantation into pseudo‐ pregnant female mice (Figure 2B, see material and methods). The resulting chimeric offspring were used to start the “CIN tracker” mouse strain. Figure 2: Creating and validating tri‐fluorescent mESC. A) Diagram of Cre‐inducible tri‐ fluorescent vector used to create mouse model. B) Schematic of method used to create mouse model. Inducible vector was transduced into mouse ESC, which was injected into a mouse blastocyst, which was in turn injected into a pregnant mother mouse. The offspring of which were chimeras containing the construct. C) A western blot of some of the mESC screened for multi‐protein expression. mESC clones were activated via Adenoviral Cre expression. RPE1 cells were used as controls. Alpha‐tubulin was used as loading control. D) mESC clone A10 imaged via live cell imaging in GFP, mCherry and CFP channels. Images in i) interphase and ii) mitosis. E) qPCR results for vector integrations per mESC clone. All delta‐ CP values were set relative to clone A10, clones with single vector integrations were colored green. A) E) C)
H2B CenpB alpha-Tubulin Merge
i)
ii)
loxP loxP
pChicken B-actin Neo Stop H2B-GFP-T2A-CenpB-mCherry-T2A-mTorq-Tubulin PolyA
Offspring Vector transduction mESC injection injection Mother mouse Blastocyst B) stop codon CenpB-mCherry H2B-eGFP mTorquoise2-Tubulin Alpha-Tubulin
A1 A6 A8 A10 B1 B3 B6 B9 B10 C2 C4 C11 neg. pos.
Triple positive? + - + + - - - + + - - + - +
mouse ESC clones RPE1
D) 0 1 2 3 4 5 6 7 A8 A10 B9 B10 C11 D10 E8 E10 G6 G9 mESC clone gD N A m ar ke r i nt eg ra tio ns (R el at iv e t o A 10) Single integration FALSE TRUE
1
4
Cells in i) interphase or ii) Mitosis. Zoom in of centrioles displayed on the right. D) Quantification of CIN phenotypes from live cell imaging experiments of RPE1 cells expressing either H2B‐GFP, or an inducible form of the tri‐fluorescent vectors. “Inducible CenpB”= H2B‐ GFP‐T2A‐CenpB‐mCherry‐T2A‐mTorquoise2‐Tubulin, “inducible centrin3”= H2B‐GFP‐T2A‐ mCherry‐Centrin3‐T2A‐mTorquoise2‐Tubulin. P‐values from chi squared test displayed. Number of (i) mitotic cells or (ii) interphase cells counter per condition displayed. specific and /or temporal control of genetic switching. When wildtype loxP sites are used, Cre recombinase leads to a permanent change in the genomic DNA, so the ‘switched’ cells pass the genomic alteration on to their daughter cells. As we want to label cells permanently with the mitotic marker, but only in a selection of cells (tissue/time specific), we engineered a vector with a Cre inducible fluorescent marker. The marker contained fluorescent fusions of H2B, CenpB and tubulin (Figure 2A), and genes were separated by a self‐cleaving T2A sequence, similar to what is described above. To generate mouse ES cells for blastocyst injection and transgenic mouse derivation, the vector was linearized, electorporated into mouse Embryonic Stem Cells (mESC) and screened for correct gene expression (Figure 2B). A total of 96 single cell‐derived mESC colonies were plated, grown, split and transduced with Cre adenovirus to test expression of the transgenic construct. Adeno‐Cre‐transduced cells were subjected to time‐ lapse imaging to quantify GFP expression, and GFP positive cells were evaluated for H2B‐eGFP, CenpB‐mCherry and mTurquoise2173‐Tubulin protein expression (Figure 2C, D I and ii not all clones shown). While many clones showed correct expression and localization of H2B and tubulin, the expression of CenpB was weak and sporadic. Even within a single clone, CenpB expression was highly variable between cells. While multiple transgene integrations would lead to higher fluorescence expression, this increased expression would be lost in mice as the transgene would be integrated in multiple chromosomes, which would partly be lost due to asymmetric inheritance in offspring. Therefore, qPCR was used to identify ES cell clones with a single vector integration, and the clones with multiple integrations were discarded (Figure 2E). These analyses lead us to select clone A10, a mESC clone with a single vector integration, correct fluorescence localization and high fluorescent protein expression. An early passage of this mESC clone that did not yet express Cre recombinase was injected into mouse blastocysts, followed by transplantation into pseudo‐ pregnant female mice (Figure 2B, see material and methods). The resulting chimeric offspring were used to start the “CIN tracker” mouse strain. Figure 2: Creating and validating tri‐fluorescent mESC. A) Diagram of Cre‐inducible tri‐ fluorescent vector used to create mouse model. B) Schematic of method used to create mouse model. Inducible vector was transduced into mouse ESC, which was injected into a mouse blastocyst, which was in turn injected into a pregnant mother mouse. The offspring of which were chimeras containing the construct. C) A western blot of some of the mESC screened for multi‐protein expression. mESC clones were activated via Adenoviral Cre expression. RPE1 cells were used as controls. Alpha‐tubulin was used as loading control. D) mESC clone A10 imaged via live cell imaging in GFP, mCherry and CFP channels. Images in i) interphase and ii) mitosis. E) qPCR results for vector integrations per mESC clone. All delta‐ CP values were set relative to clone A10, clones with single vector integrations were colored green. A) E) C)H2B CenpB alpha-Tubulin Merge
i)
ii)
loxP loxP
pChicken B-actin Neo Stop H2B-GFP-T2A-CenpB-mCherry-T2A-mTorq-Tubulin PolyA
Offspring Vector transduction mESC injection injection Mother mouse Blastocyst B) stop codon CenpB-mCherry H2B-eGFP mTorquoise2-Tubulin Alpha-Tubulin
A1 A6 A8 A10 B1 B3 B6 B9 B10 C2 C4 C11 neg. pos.
Triple positive? + - + + - - - + + - - + - +
mouse ESC clones RPE1
D) 0 1 2 3 4 5 6 7 A8 A10 B9 B10 C11 D10 E8 E10 G6 G9 mESC clone gD N A m ar ke r i nt eg ra tio ns (R el at iv e t o A 10) Single integration FALSE TRUE
Embryonic Stem Cells express transcripts without a polyA site In our initial attempt to make the CIN tracker mice, transfected and selected mESCs colonies displayed substantial expression of the fluorescently labeled proteins: several mESC clones showed moderate to good fluorescence expression, with all three proteins localizing correctly: chromatin (H2B), kinetochores (CenpB), and tubulin (α‐Tubulin) (Supplementary Figure 1A). The resulting CIN tracker mouse was next subjected to a number of additional validation tests. We harvested MEFs, which were genotyped to check for the inheritance of the “CIN tracker” marker, and positive clones were induced to express the marker by transducing them with Cre Adenovirus. As a positive control, 3T3 MEFs were transduced with the same Cre inducible tri‐fluorescent marker used to make the CRE tracker mouse (Supplementary Table 1). While we succeeded in transfecting 3T3 cells with a control GFP vector, transfection of the CIN tracker construct did not show any expression of the transgenic construct and none of the “CIN tracker” primary MEFs displayed any fluorescence expression after Cre‐treatment(Supplementary Figure 1B). The lack of expression in the transgenic MEFs was not the result of absence of the construct, nor failure of Cre to remove the lox‐STOP‐lox sequence that blocks mitotic marker expression before Cre activation (Data not shown). To check whether the lack of expression was specific to MEFs, the “CIN tracker” mice were crossed with Cre‐ERT2 mice (a tamoxifen‐inducible version, systemically expressed version of Cre‐recombinase), treated with tamoxifen via oral gavaging, to activate of the fluorescence in vivo. Unfortunately, also in this case, two‐photon imaging of the mouse skin revealed that the initial “CIN tracker” mice had no notable fluorescence expression after tamoxifen induction (Supplementary Figure 1C). To investigate the reason for the lack of fluorescence expression, we isolated genomic DNA from the CIN tracker mice and sequenced each section of the transgene, revealing that the transgenic construct was abrogated between the 3’ end of the mTorquoise2‐Tubulin and the PolyA site (Supplementary Figure 1D). We concluded that the fluorescent marker was separated from its downstream polyA site when it was transfected into mESC. One somewhat unrelated finding therefore is that mouse ES cells apparently do not need a polyA sequence to express the fluorescent marker transcript. Thus, while mESCs may be able to translate transcripts without a PolyA tail, many differentiated cells clearly cannot. While this is an intriguing finding, it is outside the scope of this thesis and we therefore decided to not pursue this further. Activating the CIN tracker mouse model Given the failed CIN tracker mice, version 1, we generated new mESCs with the triple fluorescent marker, now ensuring that the polyA sequence would be included, digesting the transgene targeting vector 718bp downstream of the polyA site (Figure 2A). Transgenic ES cells were derived as described above (Figure 2B). Before the ES cells were injected, they were tested for protein expression of H2B‐GFP, CenpB‐mCherry and mTurq‐Tubulin (Figure 2C), imaged for fluorescent protein localization (Figure 2D), and qPCR was used to check for clones with only a single integration of the vector construct (Figure 2E), followed by blastocyst injection and injection of chimeric blastocysts into pseudo‐pregnant female mice. Before activation and assessment of CIN tracker expression in mice, we tested the CIN tracker construct in cell culture. For this purpose, CIN tracker mice were bred with Cre‐ERT2 homozygous mice, from which we isolated primary mouse embryonic fibroblasts (MEFs) containing the CIN tracker construct and Cre‐ERT2. A genomic PCR was used to identify which MEF clones had inherited the CIN tracker construct (Data not shown). Clones that were positive for the construct were either transduced with Cre adenovirus, or cultured with 4‐hydroxy‐tamoxifen to activate Cre‐ERT2 to remove the lox‐STOP‐lox cassette 5’ of the fluorescent construct. CIN tracker MEFs were monitored for fluorescence expression and correct localization of the mitotic reporters. Indeed, 90% of the MEFs treated with Cre‐adenovirus expressed the fluorescent proteins, while 4‐hydroxy‐ tamoxifen treated cells had ~ 5% fluorescence expression. Importantly, switched MEFs showed expression and correct localization of all three fluorescent proteins expressed: H2B‐eGFP, CenpB‐mCherry, and mTurquoise2‐Tubulin (Figure 3A). However, while H2B and tubulin fluorescence was always co‐expressed, CenpB loci were only sporadically visible, despite all three proteins being translated from the same mRNA, in line with our earlier observations in mouse ES cells.
1
4
Embryonic Stem Cells express transcripts without a polyA site In our initial attempt to make the CIN tracker mice, transfected and selected mESCs colonies displayed substantial expression of the fluorescently labeled proteins: several mESC clones showed moderate to good fluorescence expression, with all three proteins localizing correctly: chromatin (H2B), kinetochores (CenpB), and tubulin (α‐Tubulin) (Supplementary Figure 1A). The resulting CIN tracker mouse was next subjected to a number of additional validation tests. We harvested MEFs, which were genotyped to check for the inheritance of the “CIN tracker” marker, and positive clones were induced to express the marker by transducing them with Cre Adenovirus. As a positive control, 3T3 MEFs were transduced with the same Cre inducible tri‐fluorescent marker used to make the CRE tracker mouse (Supplementary Table 1). While we succeeded in transfecting 3T3 cells with a control GFP vector, transfection of the CIN tracker construct did not show any expression of the transgenic construct and none of the “CIN tracker” primary MEFs displayed any fluorescence expression after Cre‐treatment(Supplementary Figure 1B). The lack of expression in the transgenic MEFs was not the result of absence of the construct, nor failure of Cre to remove the lox‐STOP‐lox sequence that blocks mitotic marker expression before Cre activation (Data not shown). To check whether the lack of expression was specific to MEFs, the “CIN tracker” mice were crossed with Cre‐ERT2 mice (a tamoxifen‐inducible version, systemically expressed version of Cre‐recombinase), treated with tamoxifen via oral gavaging, to activate of the fluorescence in vivo. Unfortunately, also in this case, two‐photon imaging of the mouse skin revealed that the initial “CIN tracker” mice had no notable fluorescence expression after tamoxifen induction (Supplementary Figure 1C). To investigate the reason for the lack of fluorescence expression, we isolated genomic DNA from the CIN tracker mice and sequenced each section of the transgene, revealing that the transgenic construct was abrogated between the 3’ end of the mTorquoise2‐Tubulin and the PolyA site (Supplementary Figure 1D). We concluded that the fluorescent marker was separated from its downstream polyA site when it was transfected into mESC. One somewhat unrelated finding therefore is that mouse ES cells apparently do not need a polyA sequence to express the fluorescent marker transcript. Thus, while mESCs may be able to translate transcripts without a PolyA tail, many differentiated cells clearly cannot. While this is an intriguing finding, it is outside the scope of this thesis and we therefore decided to not pursue this further. Activating the CIN tracker mouse model Given the failed CIN tracker mice, version 1, we generated new mESCs with the triple fluorescent marker, now ensuring that the polyA sequence would be included, digesting the transgene targeting vector 718bp downstream of the polyA site (Figure 2A). Transgenic ES cells were derived as described above (Figure 2B). Before the ES cells were injected, they were tested for protein expression of H2B‐GFP, CenpB‐mCherry and mTurq‐Tubulin (Figure 2C), imaged for fluorescent protein localization (Figure 2D), and qPCR was used to check for clones with only a single integration of the vector construct (Figure 2E), followed by blastocyst injection and injection of chimeric blastocysts into pseudo‐pregnant female mice. Before activation and assessment of CIN tracker expression in mice, we tested the CIN tracker construct in cell culture. For this purpose, CIN tracker mice were bred with Cre‐ERT2 homozygous mice, from which we isolated primary mouse embryonic fibroblasts (MEFs) containing the CIN tracker construct and Cre‐ERT2. A genomic PCR was used to identify which MEF clones had inherited the CIN tracker construct (Data not shown). Clones that were positive for the construct were either transduced with Cre adenovirus, or cultured with 4‐hydroxy‐tamoxifen to activate Cre‐ERT2 to remove the lox‐STOP‐lox cassette 5’ of the fluorescent construct. CIN tracker MEFs were monitored for fluorescence expression and correct localization of the mitotic reporters. Indeed, 90% of the MEFs treated with Cre‐adenovirus expressed the fluorescent proteins, while 4‐hydroxy‐ tamoxifen treated cells had ~ 5% fluorescence expression. Importantly, switched MEFs showed expression and correct localization of all three fluorescent proteins expressed: H2B‐eGFP, CenpB‐mCherry, and mTurquoise2‐Tubulin (Figure 3A). However, while H2B and tubulin fluorescence was always co‐expressed, CenpB loci were only sporadically visible, despite all three proteins being translated from the same mRNA, in line with our earlier observations in mouse ES cells.Figure 3: Mice express fluorescent proteins. A) MEF cells from “CIN tracker” mice, induced to become fluorescent with Cre Adenovirus. B) 2‐photon imaging on live “CIN tracker” mouse skin with mosaic switching. i) Single z‐image with 10x lens. Since the mouse is alive during imaging, the mouse’s breathing causes out‐of‐focus stripes. ii) Z‐stack projection of fluorescent cells at 40x. Breathing stripes disappear due to z‐stack projection due to breathing stripes not syncing up between different z‐stack images. C) 3D z‐stack reconstruction of mouse skin with fluorescent cells. Green= H2B‐eGFP, Blue = mTorquoise2‐ Tubulin, Purple= collagen (2nd harmonics). A) B) M ou se s ki n, 1 00 x Mo us e sk in, 40 0x hair follicle
Tubulin H2B mCherry Merge
C re -in du ced M EF s Zo om Mous e ski n, 40 0x , 3 D C)
H2B CenpB Tubulin Merge
i) ii) Collagen layer To now test the construct in vivo, we applied 4‐hydroxy‐tamoxifen (4OHT) topically to the skin of mice in order to activate Cre‐ Loxp switching and induce fluorescence expression (Figure 2A). Mice were imaged at one week and three weeks after first 4OHT application. Mouse skin revealed mosaic cell switching, with fluorescent cells clustered together in groups (Figure 3B). While tubulin and H2B were clearly visible (Figure 3B), we failed to detect kinetochore (CenpB) fluorescence in epidermal cells. Our imaging method also picked up an auto‐fluorescent signal from the adjacent dermal layer. This stems from the fact that collagen, a dense protein which makes up the majority of the dermal layer in skin180, converts light to its second harmonic during 2‐photon imaging181. Therefore, despite having no fluorescent protein labeling, the collagen layer of the dermis is clearly visible in our 2‐photon imaging experiments. The fluorescent cells rest on top of the dermis’ collagen layer, forming a cellular epidermal layer measured to be ~40uM thick (Figure 3C). The fraction of fluorescent cells present in the skin of 4OHT‐treated mice did not noticeably change between one and three weeks after tamoxifen application. However, fluorescent cells tended to appear in small groups, suggesting that Cre activation switched transiently amplifying cells, generating adjacent fluorescent clones, although formally, we cannot exclude the possibility that 4OHT only acted very locally. Together, these data show that our tamoxifen activated, Cre based genetic switching can provoke expression of the CIN tracker, and will allow us to monitor labeled cells for several weeks, if not months, after tamoxifen induction. Discussion Absence of fluorescent CenpB While both H2B, a chromatin marker, and tubulin, a marker for the mitotic spindle, were fluorescently labeled and visible within the skin cells of the CIN tracker mice, we failed to detect the CenpB kinetochore marker. While disappointing, this fits well with our observations in the transgenic mouse ES cells and MEFs, which are genetically identical to the mice model. This inconsistent expression may be explained by the genetic positioning of the CenpB‐mCherry gene. The kinetochore marker is located between H2B and Tubulin on a single open reading frame separated by T2A self‐cleaving
1
4
Figure 3: Mice express fluorescent proteins. A) MEF cells from “CIN tracker” mice, induced to become fluorescent with Cre Adenovirus. B) 2‐photon imaging on live “CIN tracker” mouse skin with mosaic switching. i) Single z‐image with 10x lens. Since the mouse is alive during imaging, the mouse’s breathing causes out‐of‐focus stripes. ii) Z‐stack projection of fluorescent cells at 40x. Breathing stripes disappear due to z‐stack projection due to breathing stripes not syncing up between different z‐stack images. C) 3D z‐stack reconstruction of mouse skin with fluorescent cells. Green= H2B‐eGFP, Blue = mTorquoise2‐ Tubulin, Purple= collagen (2nd harmonics). A) B) M ou se s ki n, 1 00 x Mo us e sk in, 40 0x hair follicleTubulin H2B mCherry Merge
C re -in du ce d M EF s Zo om Mous e ski n, 40 0x , 3 D C)
H2B CenpB Tubulin Merge
i) ii) Collagen layer To now test the construct in vivo, we applied 4‐hydroxy‐tamoxifen (4OHT) topically to the skin of mice in order to activate Cre‐ Loxp switching and induce fluorescence expression (Figure 2A). Mice were imaged at one week and three weeks after first 4OHT application. Mouse skin revealed mosaic cell switching, with fluorescent cells clustered together in groups (Figure 3B). While tubulin and H2B were clearly visible (Figure 3B), we failed to detect kinetochore (CenpB) fluorescence in epidermal cells. Our imaging method also picked up an auto‐fluorescent signal from the adjacent dermal layer. This stems from the fact that collagen, a dense protein which makes up the majority of the dermal layer in skin180, converts light to its second harmonic during 2‐photon imaging181. Therefore, despite having no fluorescent protein labeling, the collagen layer of the dermis is clearly visible in our 2‐photon imaging experiments. The fluorescent cells rest on top of the dermis’ collagen layer, forming a cellular epidermal layer measured to be ~40uM thick (Figure 3C). The fraction of fluorescent cells present in the skin of 4OHT‐treated mice did not noticeably change between one and three weeks after tamoxifen application. However, fluorescent cells tended to appear in small groups, suggesting that Cre activation switched transiently amplifying cells, generating adjacent fluorescent clones, although formally, we cannot exclude the possibility that 4OHT only acted very locally. Together, these data show that our tamoxifen activated, Cre based genetic switching can provoke expression of the CIN tracker, and will allow us to monitor labeled cells for several weeks, if not months, after tamoxifen induction. Discussion Absence of fluorescent CenpB While both H2B, a chromatin marker, and tubulin, a marker for the mitotic spindle, were fluorescently labeled and visible within the skin cells of the CIN tracker mice, we failed to detect the CenpB kinetochore marker. While disappointing, this fits well with our observations in the transgenic mouse ES cells and MEFs, which are genetically identical to the mice model. This inconsistent expression may be explained by the genetic positioning of the CenpB‐mCherry gene. The kinetochore marker is located between H2B and Tubulin on a single open reading frame separated by T2A self‐cleaving
peptide sequences, and previous research179 has shown that the second position of a tri‐cistronic construct with 2A cleaving sequences has the lowest expression. Thus, while CenpB‐mCherry is genetically encoded in the cell, it may be expressed below detectable limits. Future work should reveal whether CenpB can be detected at later time points, or with higher resolution imaging or more sensitive detection. Gene expression in mESC without PolyA sites While developing a mouse model containing an inducible genetic transcript without a PolyA wasn’t intentional, it did lead to a potential new discovery about protein translation in embryonic stem cells. Several mESC clones had activatable fluorescent protein expression, while the mice and MEFs resulting from those mESC clones did not, even though they were genetically identical. This implies that mESCs may not need a polyA sequence in order to translate open reading frames. To our knowledge, this has not been previously reported, and may help to illuminate some of the protein expression changes that occur when a cell differentiates. The CIN tracker fluorescent mice The level of aneuploidy within an organism varies between tissues89,92, increases with age182 and can be dramatically altered with genetic modification14,159(also this thesis Chapter 3), and drug treatment in cultured
cells 97,159. Additionally, three out of four cancers are aneuploid7,8, many
cancer cell lines have high rates of CIN137,183, and many cancers have been shown to have multiple aneuploid clones which allow for tumor selection and evolution9,27,183. Despite this, the level of chromosome mis‐segregation within these tumors has rarely been studied, due to the lack of available models to observe CIN in living organisms (See Chapter 2). While cancer cell lines can be used to study CIN in cancer, the frequency of chromosome mis‐ segregation events in cell lines does not necessarily represent the level of CIN in vivo (see also Chapter 2). The CIN tracker mouse model can be used to assess and better understand the rates and types of chromosome mis‐ segregation taking place in vivo within living cells in various tissues or tumors, at multiple time points and within a variety of genetic backgrounds without having to take cells ex vivo. There are several advantages to using the CIN tracker mouse model over other mouse models expressing fluorescently‐tagged H2B to monitor cell division, currently available. First, the CIN tracker does not only express H2B, but α‐tubulin as well, which labels the cell cytoplasm during interphase, and the spindle network during mitosis. This allows for better quantification of different forms of chromosome instability, such as multi‐ polar spindles, can possibly be used to study spindle orientation in mitosis, relevant for asymmetric cell division, and can be used to visualize cell size and shape. The CIN tracker mice express fluorescence expression in vivo, so the fluorescent markers do not need to be introduced ex vivo. Since a CIN phenotype is known to enhance tumor progression9,18,21,50,184, and some forms of CIN have even been reported to initiate tumorigenesis14, visualizing the true level and type of chromosome mis‐segregation rates in vivo is essential to better understanding the biology of cancer initiation and progression. Additionally, while low to intermediate levels of CIN have been reported to enhance tumorigenesis, high levels of CIN appear to inhibit tumor progression16,32,159,164,166, and may be a suitable approach to target CIN cells 159,164 (See chapter 3). Since the CIN tracker induces fluorescence through genetic switching, and does not rely on cell transplantation, there are no requirements to suppress the immune system in these mice. The immune system not only plays an essential, and frequently overlooked, role in tumor progression30,185,186, CIN itself has also been shown to interact with the immune system22,30. The immune system may help eliminate CIN cells, as cells that have been subjected to high rates of CIN have been shown to be eliminated using co‐ culture experiments with CIN cells and a Natural Killer cell line 97. Additionally, highly aneuploid tumors have a higher chance of metastasizing115, and CIN itself may promote epithelial to mesenchymal transition (EMT) facilitating metastasis in vivo22,187,188. The CIN tracker model may therefore be used in future experiments to visualize CIN cells in immune‐competent mice, to gain a better understanding of the complex interactions between the immune system and CIN. Additionally, CIN tracker expression can be induced in a tissue‐specific fashion by combining the CIN tracker with tissue specific Cre expression, allowing for lineage tracing within tissues or even developing cancers18,50, and would allow for tissue specific evaluation of CIN rates, and the consequences of CIN. For instance, the CIN tracker can be used to better understand why some tissues can deal with CIN induction15,18 while others cannot4,15(See also chapter 6).
1
4
peptide sequences, and previous research179 has shown that the second position of a tri‐cistronic construct with 2A cleaving sequences has the lowest expression. Thus, while CenpB‐mCherry is genetically encoded in the cell, it may be expressed below detectable limits. Future work should reveal whether CenpB can be detected at later time points, or with higher resolution imaging or more sensitive detection. Gene expression in mESC without PolyA sites While developing a mouse model containing an inducible genetic transcript without a PolyA wasn’t intentional, it did lead to a potential new discovery about protein translation in embryonic stem cells. Several mESC clones had activatable fluorescent protein expression, while the mice and MEFs resulting from those mESC clones did not, even though they were genetically identical. This implies that mESCs may not need a polyA sequence in order to translate open reading frames. To our knowledge, this has not been previously reported, and may help to illuminate some of the protein expression changes that occur when a cell differentiates. The CIN tracker fluorescent mice The level of aneuploidy within an organism varies between tissues89,92, increases with age182 and can be dramatically altered with genetic modification14,159(also this thesis Chapter 3), and drug treatment in culturedcells 97,159. Additionally, three out of four cancers are aneuploid7,8, many
cancer cell lines have high rates of CIN137,183, and many cancers have been shown to have multiple aneuploid clones which allow for tumor selection and evolution9,27,183. Despite this, the level of chromosome mis‐segregation within these tumors has rarely been studied, due to the lack of available models to observe CIN in living organisms (See Chapter 2). While cancer cell lines can be used to study CIN in cancer, the frequency of chromosome mis‐ segregation events in cell lines does not necessarily represent the level of CIN in vivo (see also Chapter 2). The CIN tracker mouse model can be used to assess and better understand the rates and types of chromosome mis‐ segregation taking place in vivo within living cells in various tissues or tumors, at multiple time points and within a variety of genetic backgrounds without having to take cells ex vivo. There are several advantages to using the CIN tracker mouse model over other mouse models expressing fluorescently‐tagged H2B to monitor cell division, currently available. First, the CIN tracker does not only express H2B, but α‐tubulin as well, which labels the cell cytoplasm during interphase, and the spindle network during mitosis. This allows for better quantification of different forms of chromosome instability, such as multi‐ polar spindles, can possibly be used to study spindle orientation in mitosis, relevant for asymmetric cell division, and can be used to visualize cell size and shape. The CIN tracker mice express fluorescence expression in vivo, so the fluorescent markers do not need to be introduced ex vivo. Since a CIN phenotype is known to enhance tumor progression9,18,21,50,184, and some forms of CIN have even been reported to initiate tumorigenesis14, visualizing the true level and type of chromosome mis‐segregation rates in vivo is essential to better understanding the biology of cancer initiation and progression. Additionally, while low to intermediate levels of CIN have been reported to enhance tumorigenesis, high levels of CIN appear to inhibit tumor progression16,32,159,164,166, and may be a suitable approach to target CIN cells 159,164 (See chapter 3). Since the CIN tracker induces fluorescence through genetic switching, and does not rely on cell transplantation, there are no requirements to suppress the immune system in these mice. The immune system not only plays an essential, and frequently overlooked, role in tumor progression30,185,186, CIN itself has also been shown to interact with the immune system22,30. The immune system may help eliminate CIN cells, as cells that have been subjected to high rates of CIN have been shown to be eliminated using co‐ culture experiments with CIN cells and a Natural Killer cell line 97. Additionally, highly aneuploid tumors have a higher chance of metastasizing115, and CIN itself may promote epithelial to mesenchymal transition (EMT) facilitating metastasis in vivo22,187,188. The CIN tracker model may therefore be used in future experiments to visualize CIN cells in immune‐competent mice, to gain a better understanding of the complex interactions between the immune system and CIN. Additionally, CIN tracker expression can be induced in a tissue‐specific fashion by combining the CIN tracker with tissue specific Cre expression, allowing for lineage tracing within tissues or even developing cancers18,50, and would allow for tissue specific evaluation of CIN rates, and the consequences of CIN. For instance, the CIN tracker can be used to better understand why some tissues can deal with CIN induction15,18 while others cannot4,15(See also chapter 6).
Importantly, mice can be imaged multiple times, allowing researchers to observe the same tissues over time, particularly when combining imaging micro‐tattooing to return to the exact same imaging position as before. Multiple imaging experiments per mouse does not only lower the number of mice necessary per experiment, but it also allows researchers to observe CIN rates in tumors or tissues before and after interventions, for instance to better understand the mechanism and/or side effects of specific drugs or the effect of genetic interventions. For example, one could setup an experiment to quantify the rate of chromosome mis‐segregation in Mad2 knockout tumors18,38(also see Chapter 6) before and after treating the tumor with SKI606, a drug that we find to increase mis‐segregation rate of Mad2‐knockdown cell in vitro (see Chapter 3), to study the interaction between CIN and the drug in vivo. While the CIN tracker was created to monitor CIN in vivo, it can also be used for other experiments. For example, one could induce mosaic switching within a single tissue, combined with the ability to image the same mouse multiple times, which would allow researchers to track cell movement and (stem) cell clusters over time. Depending on resolution that can be acquired from the imaging setup, the CIN tracker could also be used to study the structure and function of the tubulin cytoskeleton and the spindle network in different tissues and how this tubulin fibers are affected by drugs or mutations in vivo. In conclusion, this novel CIN tracker mouse model will allow us to observe chromosome segregation, chromosome mis‐segregation and the fate of the emerging cells and tissues, allowing the field to better understand the role of CIN in cancer initiation and progression, cell death, and aging. Material and Methods Cloning The tri‐fluorescent vector was made using Gibson Assembly cloning kit (New England Biolabs Inc.) to insert multiple fluorescent sections, and the connecting T2A sequences within a single cloning step. The T2A sequence was taken from Liu et al. 2017179. Vectors and cell lines are available upon request. Cell culture
RPE1 cells (ATCC) were cultured at 37°C, in 5% CO2 and 18% O2, with DMEM
media supplemented with 10% Fetal Bovine Serum and 100 Units/ml Penicillin and 100ug/ml Streptomycin. NIH 3T3 MEFs and Primary MEFs were cultured at 5% CO2 and 5% O2, in DMEM media supplemented with
FBS and Pen/Strep as above, 1% MEM NEAA (Cat. Num. 11140050) (Gibco, Thermo Scientific), and 0.1% 5.5x10^‐2 β‐Mercapto‐Ethanol (Cat. Num. 21985023) (Gibco, Thermo Scientific). Primary MEFs were kept between 30% and 80% confluency. MEF harvesting Male mice with the desired genotype were setup in a timed mating with female Black6 mice. Mothers were terminated on day 13.5 of their pregnancies. The E13.5 embryos were harvested dissected in a laminar flow cabinet. For this, the embryonic sac, internal organs and the head were removed, and tissue was homogenized with a scalpel, then digested with Trypsin for 15 minutes. Cells were resuspended in MEF media. The resulting MEF cell lines were genotyped for presence of the desired genotype and negative MEFs were discarded. Imaging Live Mice and Cre switching All animal experiments were performed in accordance with Dutch law and approved by the University of Groningen Medical Center Committees on Animal Care. Mice carrying the CIN tracker were bred with mice expressing ubiquitous Cre‐ERT2. The CIN tracker construct was activated via topical application of 4‐hydroxy‐Tamoxifen onto shaved mouse skin. A total of 100 μl of 5 mg/ml 4‐hydroxy‐tamoxifen dissolved in pure ethanol was applied to mice three times, with a 48 hour rest period between applications. Live, anaesthetized, mice were imaged with a constant supply of isofluorane (1.5‐2.5%) and 8% oxygen. The imaged mouse and all microscope equipment was kept at 37°C. Before imaging, mice were subcutaneously injected with sterile PBS equal to 1% of their body weight to keep them hydrated throughout the experiment. Mice were imaged on a Zeiss LSM 780 2‐photon microscope (Axio Observer.Z1, Zeiss International) for a maximum of 3 hours under anesthesia. Imaging cell lines
1
4
Importantly, mice can be imaged multiple times, allowing researchers to observe the same tissues over time, particularly when combining imaging micro‐tattooing to return to the exact same imaging position as before. Multiple imaging experiments per mouse does not only lower the number of mice necessary per experiment, but it also allows researchers to observe CIN rates in tumors or tissues before and after interventions, for instance to better understand the mechanism and/or side effects of specific drugs or the effect of genetic interventions. For example, one could setup an experiment to quantify the rate of chromosome mis‐segregation in Mad2 knockout tumors18,38(also see Chapter 6) before and after treating the tumor with SKI606, a drug that we find to increase mis‐segregation rate of Mad2‐knockdown cell in vitro (see Chapter 3), to study the interaction between CIN and the drug in vivo. While the CIN tracker was created to monitor CIN in vivo, it can also be used for other experiments. For example, one could induce mosaic switching within a single tissue, combined with the ability to image the same mouse multiple times, which would allow researchers to track cell movement and (stem) cell clusters over time. Depending on resolution that can be acquired from the imaging setup, the CIN tracker could also be used to study the structure and function of the tubulin cytoskeleton and the spindle network in different tissues and how this tubulin fibers are affected by drugs or mutations in vivo. In conclusion, this novel CIN tracker mouse model will allow us to observe chromosome segregation, chromosome mis‐segregation and the fate of the emerging cells and tissues, allowing the field to better understand the role of CIN in cancer initiation and progression, cell death, and aging. Material and Methods Cloning The tri‐fluorescent vector was made using Gibson Assembly cloning kit (New England Biolabs Inc.) to insert multiple fluorescent sections, and the connecting T2A sequences within a single cloning step. The T2A sequence was taken from Liu et al. 2017179. Vectors and cell lines are available upon request. Cell cultureRPE1 cells (ATCC) were cultured at 37°C, in 5% CO2 and 18% O2, with DMEM
media supplemented with 10% Fetal Bovine Serum and 100 Units/ml Penicillin and 100ug/ml Streptomycin. NIH 3T3 MEFs and Primary MEFs were cultured at 5% CO2 and 5% O2, in DMEM media supplemented with
FBS and Pen/Strep as above, 1% MEM NEAA (Cat. Num. 11140050) (Gibco, Thermo Scientific), and 0.1% 5.5x10^‐2 β‐Mercapto‐Ethanol (Cat. Num. 21985023) (Gibco, Thermo Scientific). Primary MEFs were kept between 30% and 80% confluency. MEF harvesting Male mice with the desired genotype were setup in a timed mating with female Black6 mice. Mothers were terminated on day 13.5 of their pregnancies. The E13.5 embryos were harvested dissected in a laminar flow cabinet. For this, the embryonic sac, internal organs and the head were removed, and tissue was homogenized with a scalpel, then digested with Trypsin for 15 minutes. Cells were resuspended in MEF media. The resulting MEF cell lines were genotyped for presence of the desired genotype and negative MEFs were discarded. Imaging Live Mice and Cre switching All animal experiments were performed in accordance with Dutch law and approved by the University of Groningen Medical Center Committees on Animal Care. Mice carrying the CIN tracker were bred with mice expressing ubiquitous Cre‐ERT2. The CIN tracker construct was activated via topical application of 4‐hydroxy‐Tamoxifen onto shaved mouse skin. A total of 100 μl of 5 mg/ml 4‐hydroxy‐tamoxifen dissolved in pure ethanol was applied to mice three times, with a 48 hour rest period between applications. Live, anaesthetized, mice were imaged with a constant supply of isofluorane (1.5‐2.5%) and 8% oxygen. The imaged mouse and all microscope equipment was kept at 37°C. Before imaging, mice were subcutaneously injected with sterile PBS equal to 1% of their body weight to keep them hydrated throughout the experiment. Mice were imaged on a Zeiss LSM 780 2‐photon microscope (Axio Observer.Z1, Zeiss International) for a maximum of 3 hours under anesthesia. Imaging cell lines