Atypical E2F repressors E2F7 and E2F8:
Balancing E2F activity in normal and variant cell cycles
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University
By
Madhu Micheal Ouseph
Graduate Program in Biochemistry
The Ohio State University
2012
Dissertation Committee:
Gustavo W. Leone, Ph.D., Advisor
Stephen A. Osmani, Ph.D.
Mark R. Parthun, Ph.D.
Susan E. Cole, Ph.D.
Copyright
Madhu Micheal Ouseph
2012
Abstract
Coordinated activation and repression of E2F-responsive genes is believed to be pivotal for progression of normal cell cycle. Studies using lower organisms and mammalian cell culture systems support the idea that this balanced cumulative E2F activity in different phases of cell cycle is executed by canonical E2F activators (E2F1-
E2F3) and repressors (E2F4-E2F6). But, recent studies in mouse models with ablation of these factors have failed to provide substantial molecular and phenotypic evidence to support this notion. The evolutionarily ancient arm of E2F family consisting of newly identified atypical E2F repressors, E2F7 and E2F8, is known to be critical for mammalian embryonic development. Germline ablation of E2f7 and E2f8 leads to severe placental defects culminating in embryonic lethality by embryonic age 11.5. Remarkably, the concomitant loss of E2f3a normalized placental and fetal gene expression programs, corrected placental defects and fostered the survival of E2f7/E2f8 deficient embryos to birth. Using expression profiling and biochemical approaches, we show that atypical E2F repressors and canonical E2F activator E2F3a form key antagonistic regulators of G1-S transcriptional program during placental development. We thus provide the first in vivo evidence to show that balanced repression by E2F7 and E2F8 and activation by canonical
ii
E2F activators coordinate expression of E2F-resonsive genes in mitotic cell cycle during mammalian development.
Utilizing a combination of novel and established lineage-specific cre mice we also demonstrate that these two opposing arms of the E2F program, one driven by canonical transcription activation (E2F1, E2F2 and E2F3) and the other by atypical repression (E2F7 and E2F8), converge on the regulation physiological polyploidy in vivo.
Ablation of atypical repressors diminished ploidy in the trophoblast giant cells in the placenta, hepatocytes in the liver and megakaryocytes in bone marrow, whereas ablation of canonical activators in the trophoblast giant cells and hepatocytes augmented genome ploidy. In addition, the severe reduction of ploidy caused by E2f7/E2f8 deficiency could be rescued significantly by added loss of E2f1 and E2f3a. Taken together, the results presented within provide the first in vivo evidence for a direct role of E2Fs in regulating non-traditional cell cycles in mammals. Though polyploidy has been demonstrated to be essential for metazoan development and is widely believed to have conserved physiological functions in mammals, to our surprise, reduction of ploidy caused by loss of E2f7 and E2f8 had no apparent adverse impact on placental, hepatic or megakaryocyte physiology. In summary, our studies reveal novel functions of mammalian atypical E2Fs and provide significant insight into mechanism of cell cycle phase dependent regulation of E2F-responsive genes in normal mitotic cell cycle and variant cell cycles in mammals.
iii
Dedication
This work is dedicated to Valentina and William, for their unconditional love, support and tolerance during my graduate study period.
iv
Acknowledgments
I wish to thank my graduate study committee members Stephen Osmani, Mark
Parthun, Susan Cole and my advisor Gustavo Leone for the mentorship and support that was generously extended to me during the graduate study period. This study would not have been possible without help and guidance from former and current Leone lab members, who provided very productive collaborations throughout my graduate study.
Jing Li started the core components of the project and Hui-Zi Chen contributed significantly to the work with ideas and parallel study in liver. In addition, I wish to extend my sincere thanks to John Thompson (Chris) for help with in situ hybridizations and laser capture microscopy, Thierry Pécot for bioinformatic analysis of all the microarray, NanoString and RNASeq datasets, Prashant Trikha for help with megakaryocyte analyses, Shantibhusan Senapati, SooIn Bae, and Lindsey Kent for the great support during final phase of study, Hui Wang for guiding me through design and cloning of transgenic mouse models and Pamela Wenzel for generously providing E2F1-
3 null tissues used in this analysis. Undergraduate students Veda Chokshi, Grant
Comstock, Braxton Forte, Ian Green and Benjamin Briskin were of great help with several experiments and maintaining mouse colonies. Core facilities of Ohio State
University Medical Center including Nucleic Acid, Microarray, Laser capture
v microscopy, Analytical cytometry and Transgenic and embryonic stem cell core facility of Nationwide Children‟s Hospital were instrumental in successful completion of the projects. I also thank Maysoon Rawahneh (Lisa) and Julie Mofitt for help with histology,
Shantanu Singh and Sundaresan Raman for confocal image analysis, Xiaokui Mo and
Soledad Fernandez for statistical analysis, Anthony Trimboli and Liz Stranges for administrative and technical guidance and Sudarshana Sharma and Anil Singh for their friendship and guidance. Finally, I am grateful to my parents and former teachers for all the wisdom that they gave me and for laying the foundation of my life and career.
vi
Vita
May 1999 ………………………….. Bachelor of Medicine and Surgery (M.B.B.S.) Government Medical College, Kottayam Mahatma Gandhi University, India
June 2005 ………………………….. Doctor of Medicine (M.D.) Maulana Azad Medical College University Of Delhi, India
2005- 2006 ………………………… Senior Resident Department of Radiotherapy and Oncology Lok Nayak Hospital, New Delhi, India
2006- 2007 ………………………… Program Fellow Ohio State Biochemistry Program The Ohio State University, USA
2007- Present ……………………… Graduate Research Associate Comprehensive Cancer Center The Ohio State University, USA
vii
Publications
Ouseph MM, Green IF, Briskin BJ and Leone G. Generation and characterization of trophoblast giant cell specific Cre mice lines. Genesis (Manuscript in preparation).
Chen HZ*, Ouseph MM*, Li J, Pécot T, Liu B, Chokshi V, Byrne M, Duran C, Comstock G, Martin CK, Trikha P, Senapati S, Huang YW, Gandhi S, Wilson N, Thompson JC, Raman S, Singh S, Leone M, Machiraju R, Huang K, Wolgemuth DJ, Sicinski P, Huang T, Jin V and Leone G. Atypical E2Fs Link Mammalian Endocycle Control to Cancer. Nature Cell Biology (under revision) * Equal contribution first authors.
Ouseph MM*, Li J*, Chen HZ, Pécot T, Wenzel P, Thompson JC, Comstock G, Chokshi V, Byrne M, Forde B, Chong JL, Huang K, Machiraju R, de Bruin, A and Leone G. Atypical E2F Repressors and Activators Coordinate Placental Development. Developmental Cell (accepted Jan 2012) * Equal contribution first authors.
Moirangthem V and Ouseph MM. Atypical presentations of Amyotrophic Lateral Sclerosis: A case report. Journal of Neuropsychiatry and Clinical Neurosciences 2011; 23(3): 362-364.
Chong J, Wenzel P, Sáenz-Robles M, Nair V, Ferrey A, Hagan JP, Gomez YM, Sharma N, Chen HZ, Ouseph M, Wang SH, Trikha P, Culp B, Mezache L, Winton DJ, Sansom OJ, Chen D, Bremner R, Cantalupo PG, Robinson ML, Pipas JM, and Leone G. E2f1-3 switch from activators in progenitor cells to repressors in differentiating cells. Nature 2009; 462(7275): 930-4.
Abhishek A, Ouseph MM, Sharma P, Kamal V and Sharma M. Bulky scalp metastasis and superior sagittal sinus thrombosis from a cervical adenocarcinoma: an unusual case. Journal of Medical Imaging and Radiation Oncology 2008; 52(1): 91-4.
viii
Ouseph MM, Sharan GK, Sharma P, Rathi AK, Jain S, Singh K and Bahadur AK. Osteosarcoma with cutaneous metastases. A case report. Acta Cytologica 2007; 51(1): 102-6.
Rathi AK, Ouseph MM, Singh K and Bahadur AK. Radiation Therapy in Breast Cancer. Journal of International Medical Sciences Academy 2006; 19(1): 53- 58.
Sharma P, Jain S, Nigam S, Bahadur AK and Ouseph MM. Malignant fibrous histiocytoma of the chest wall masquerading as medullary breast carcinoma: a case report. Acta Cytologica 2006; 50(5): 577-80.
Ouseph MM, Moirangthem V., Rathi AK, Mohanta PK, Singh K and Bahadur AK. Palliative thoracic Radiotherapy in advanced Non-small Cell Lung Cancer patients with poor performance status: Results of a randomized control trial comparing single fraction of 10 Gy to a protracted regimen of 30 Gy/ 10 fractions. Educational Book of 26th Annual Conference of Association of Radiation Oncologist of India 2004: p50-69.
Fields of Study
Major Field: Biochemistry
ix
Table of Contents
Abstract ...... ii
Dedication ...... iv
Acknowledgments...... v
Vita ...... vii
List of Tables ...... xix
List of Figures ...... xx
Chapter 1: Cell cycle regulation and current CDK-RB-E2F paradigm ...... 1
1.1 E2F family of transcription factors ...... 1
1.2 Complex specific, synergistic and antagonistic in vivo roles of E2F family members
...... 5
1.3 E2F transcription factors and regulation of mitotic cell cycle ...... 9
1.4 Constitutive Derepression of E2F-regulated genes in stem cells ...... 12
1.5 E2F-Mediated activation and repression in differentiated cells ...... 14
1.6 Direct roles of E2Fs in DNA Replication Process ...... 20
1.7 E2Fs in DNA damage response and checkpoint activation ...... 24 x
1.8 Variant cell cycles and physiological polyploidy ...... 26
1.9 Regulation of atypical cell cycles and role of E2F transcription factors ...... 30
Chapter 2: Canonical E2F activators and atypical E2F repressors coordinate normal mitotic cell cycle progression during placental development ...... 43
2.1 Introduction ...... 43
2.2 Results ...... 46
2.2.1 Expression of E2f7 and E2f8 in murine placenta ...... 46
2.2.2 Loss of E2f7 and E2f8 causes differentiation and architectural defects in
placenta culminating in mid-gestational embryonic lethality ...... 47
2.2.3 Aberrant proliferation and cell death in trophoblast cells null for E2f7 and E2f8
...... 49
2.2.4 Derepression of S-G2-M genes underlie the cell cycle defects observed in
double mutant trophoblasts ...... 49
2.2.5 E2F3a regulates E2F-responsive genes in antagonistic way and rescues
molecular defects of E2f7/E2f8 null tissues ...... 51
2.2.6 Added loss of E2f3a rescues E2f7/E2f8 null phenotype ...... 53
2.2.7 E2F3a positively regulates expression of E2f7 and E2f8 ...... 53
2.3 Discussion ...... 54
2.4 Materials and methods ...... 57
2.4.1 Mouse Strains and Genotyping ...... 57 xi
2.4.2 Histology, Immunostaining and Quantification ...... 58
2.4.3 Quantitative RT-PCR (qRT-PCR) ...... 58
2.4.4 In situ Hybridization ...... 59
2.4.5 Affymetrix Microarray Analysis ...... 59
2.4.6 Chromatin Immunoprecipitation (ChIP) ...... 60
2.4.7 Statistical Analysis ...... 60
2.4.8 Laser Capture Microdissection ...... 61
Chapter 3: Generation and characterization of trophoblast giant cell specific cre mouse lines ...... 79
3.1 Introduction ...... 79
3.2 Results ...... 82
3.2.1 Targeting Proliferin and Placental Lactogen-1 gene loci ...... 82
3.2.2 PlfCre and Pl1Cre faithfully reproduce predicted expression pattern of natural
loci ...... 83
3.2.3 Loss of Plf and Pl1 has little consequence on murine embryonic development
...... 84
3.3 Discussion ...... 84
3.4 Materials and methods ...... 85
3.4.1 Mouse Strains and Genotyping ...... 85
xii
3.4.2 Histology and X-Gal Staining ...... 86
3.4.3 Southern Blot ...... 86
Chapter 4: Regulation of endocycle in trophoblast giant cells by atypical E2F repressors ..
...... 91
4.1 Introduction ...... 91
4.2 Results ...... 93
4.2.1 Expression of E2f7 and E2f8 in TGCs ...... 93
4.2.2 Loss of E2f7 and E2f8 leads to aberrant S-phase activity, activation of mitotic
programs and induces karyokinesis in TGCs ...... 94
4.2.3 Endocycle defects in E2f7/E2f8 null TGCs leads to significant reduction in
ploidy ...... 95
4.2.4 E2F7 and E2F8 synergistically promote TGC endocycles ...... 95
4.2.5 Cell autonomous effects of loss of E2f7 and E2f8 ...... 96
4.2.6 Defects in endocycle in trophoblast giant cells have no physiological
consequences in placentation and embryogenesis ...... 96
4.2.7 Endocycle defects in E2f7/E2f8 null giant cells is independent of p57 ...... 97
4.2.8 Deregulation of G2/M targets as the underlying mechanism behind endocycle
defects……………………………………………………………………………98
4.2.9 Similar molecular mechanism underlie regulation of endocycle and
acytokinetic mitosis in hepatocytes ...... 99
xiii
4.3 Discussion ...... 101
4.4 Materials and methods ...... 102
4.4.1 Mouse Strains and Genotyping ...... 102
4.4.2 Ploidy analysis using the Feulgen technique ...... 102
4.4.3 Ploidy analysis using flow cytometry ...... 103
4.4.4 RNASeq global gene expression profiling analyses ...... 103
4.4.5 NanoString gene expression analyses ...... 104
4.4.6 Confocal microscopy 3D reconstruction of TGC nuclei ...... 105
4.4.7 Chromatin immunoprecipitation (ChIP) Assays ...... 106
4.4.8 Immunohistochemistry (IHC) and Immunofluorescence (IF) ...... 107
4.4.9 BrdU incorporation assays ...... 107
4.4.10 IHC quantification and statistical analysis ...... 108
4.4.11 E2F binding site search ...... 108
4.4.12 Statistical Analysis ...... 108
Chapter 5: Restoration of G2/M block by ablation of Cyclin A1 and Cyclin A2 significantly rescues defects in E2F7 and E2F8 null trophoblast giant cells ...... 128
5.1 Introduction ...... 128
5.2 Results ...... 129
5.2.1 Deregulation of Cyclin A2 in TGCs deficient for E2f7 and E2f8 ...... 130
xiv
5.2.2. Loss of Cyclin A1/A2 has no effect on TGC endocycle ...... 130
5.2.3. Added loss of Cyclin A1/A2 rescues aberrant mitosis in E2f7/E2f8 deficient
TGCs ...... 130
5.2.4. Polyploidy defects in E2f7/E2f8 deficient TGCs is significantly ameliorated
by loss of Cyclin A1/A2 ...... 131
5.3 Discussion ...... 131
5.4 Materials and methods ...... 132
5.4.1 Mouse strains and genotyping ...... 132
5.4.2 Ploidy analysis using the Feulgen technique ...... 133
5.4.3 Immunohistochemistry (IHC) and Immunofluorescence (IF) ...... 133
5.4.4 BrdU incorporation assays...... 134
5.4.5 IHC quantification and statistical analysis ...... 134
5.4.6 Statistical Analysis ...... 134
Chapter 6: Canonical E2F repressors E2F4 and E2F5 has no significant role in regulation of endocycle in physiological conditions ...... 140
6.1 Introduction ...... 140
6.2 Results ...... 140
6.2.1 Effect of E2f7/E2f8 ablation in TGCs on canonical repressors ...... 141
6.2.2 Expression of canonical E2F repressors in TGCs ...... 141
xv
6.2.3 Loss of E2f4 and E2f5 have no significant effect in TGC biology ...... 141
6.3 Discussion ...... 141
6.4 Materials and methods ...... 142
6.4.1 Mouse Strains and Genotyping ...... 142
6.4.2 NanoString gene expression analyses ...... 142
6.4.3 Histology ...... 143
Chapter 7: Role of E2F activators in regulation of TGC endocycle ...... 147
7.1 Introduction ...... 147
7.2 Results ...... 148
7.2.1 Effect of loss of atypical repressors on expression of canonical activators .. 148
7.2.2 Expression of E2F activators in TGCs ...... 148
7.2.3. Loss of E2F activators promote TGC endocycle ...... 148
7.3 Discussion ...... 149
7.4 Materials and methods ...... 152
7.4.1 Mouse Strains and Genotyping ...... 152
7.4.2 Ploidy analysis using the Feulgen technique ...... 152
7.4.3 NanoString gene expression analyses ...... 153
7.4.4 Statistical Analysis ...... 154
Chapter 8: E2F1 and E2F3a functionally antagonizes atypical E2F repressors ...... 158
xvi
8.1 Introduction ...... 158
8.2 Results ...... 159
8.2.1 Added loss of E2f1 and E2f3a rescues defects in nuclear size and nuclear
morphology of E2f7/E2f8 null TGCs ...... 159
8.2.2 E2F1 and E2F3a dictate the molecular consequences in E2f7/E2f8 null TGCs
...... 159
8.2.3 E2F1 and E2F3a functionally antagonizes E2F7 and E2F8 in regulation of
endocycle in TGCs ...... 159
8.3 Discussion ...... 160
8.4 Materials and methods ...... 162
8.4.1 Mouse Strains and Genotyping ...... 162
8.4.2 Histology, Immunostaining and Quantification ...... 162
8.4.3 Ploidy analysis using the Feulgen technique ...... 163
8.4.4 E2F binding site search ...... 163
8.4.5 Statistical Analysis ...... 164
Chapter 9: Atypical E2Fs play significant role in regulation of endomitosis in megakaryocytes ...... 168
9.1 Introduction ...... 168
9.2 Results ...... 170
9.2.1. Loss of E2f7/E2f8 disrupts endomitosis in megakaryocytes ...... 170 xvii
9.2.2 Disruption of endomitosis by loss of E2f7/E2f8 is of little physiological
consequence ...... 171
9.2.3 Molecular mechanisms underlying the phenotype ...... 172
9.2.4 Loss of canonical activators also leads to similar phenotype ...... 172
9.3 Discussion ...... 172
9.4 Materials and methods ...... 174
9.4.1 Mouse Strains and Genotyping ...... 174
9.4.2 Ploidy analysis using flow Cytometry in megakaryocytes ...... 174
9.4.3 Platelet analysis ...... 174
9.4.4 pIpC injections for induction of Mx1-Cre ...... 175
Chapter 10:.Summary and perspectives ...... 181
10.1 Balancing E2F activity in mitosis and variant cell cycles ...... 181
10.2 Biological functions of polyploidy ...... 184
10. 3 E2f7/E2f8: Tumor suppressor or oncogene ...... 185
Bibliography ...... 188
xviii
List of Tables
Table 2.1 Potential direct targets of E2F7 and E2F8: based on presence of conserved canonical E2F binding sites ...... 76
Table 2.2 PCR genotyping primers used in the study ...... 77
Table 2.3 Quantitative Real-Time and ChIP PCR primers used in the study………..78
Table 4.1 PCR Genotyping Primers for Chapters 4 to 8 ...... 125
Table 4.2 ChIP PCR Primers ...... 127
Table 9.1 PCR genotyping primers ...... 180
xix
List of Figures
Figure 1.1 The canonical Cdk-Rb-E2F pathway ...... 39
Figure 1.2 The mammalian E2F gene family ...... 40
Figure 1.3 Regulation of E2F-responsive genes ...... 41
Figure 1.4 Chromatin modifications by E2F transcription factors ...... 42
Figure 2.1 Expression of E2f7 and E2f8 in mouse placenta ...... 62
Figure 2.2 Severe architectural defects in E2f7/E2f8 null placenta ...... 63
Figure 2.3 Differentiation defects in E2f7/E2f8 null placenta ...... 64
Figure 2.4 Defects in differentiation of major trophoblast lineages in E2f7/E2f8 null placenta ...... 65
Figure 2.5 Cell cycle defects in E2f7/E2f8 null placenta ...... 66
Figure 2.6 Deregulation of cell cycle genes in E2f7/E2f8 null placenta ...... 67
Figure 2.7 Validation of deregulated targets for direct E2F7/E2F8 binding ...... 68
Figure 2.8 E2F3a and E2F7/E2F8 regulates same subset of genes in antagonistic manner
...... 69
xx
Figure 2.9 Added loss of E2f3a rescues defects in expression of genes in E2f7/E2f8 null placentas and fetuses ...... 70
Figure 2.10 Added loss of E2f3a rescues cell cycle and structural defects in E2f7/E2f8 null placenta ...... 72
Figure 2.11 Added loss of E2f3a rescues mid-gestational lethality of E2f7/E2f8 null placenta ...... 73
Figure 2.12 E2F3a induces expression of E2f7 and E2f8 ...... 74
Figure 2.13 Current model of balanced E2F activity in regulation of E2F-responsive genes ...... 75
Figure 3.1 Generation of PlfCre and Pl1Cre ...... 87
Figure 3.2 Trophoblast giant cell specific Cre recombinase activity in PlfCre and Pl1Cre mice is noticeable as early as E4.5 ...... 88
Figure 3.3 TGC specific expression of cre recombinase in mature placentas ...... 89
Figure 3.4 PlfCre and Pl1Cre mice are viable and fertile even in homozygosity ...... 90
Figure 4.1 Expression of E2f7 and E2f8 in TGCs ...... 110
Figure 4.2 Loss of E2f7 and E2f8 leads to aberrant mitosis in TGCs ...... 111
Figure 4.3 Loss of E2f7 and E2f8 causes reregulation of cell cycle genes ...... 112
Figure 4.4 Loss of E2f7 and E2f8 leads to significant reduction of ploidy in TGCs ..... 113
Figure 4.5 Synergistic functions of E2f7 and E2f8 in regulation of endocycle in TGCs
...... 114 xxi
Figure 4.6 Cell autonomous functions of E2f7 and E2f8 in regulation of TGC endocycle
...... 115
Figure 4.7 Disruption of endocycle in TGCs causes no significant physiological consequences ...... 116
Figure 4.8 Confirmation of cell specific deletion of E2f7 and E2f8 ...... 117
Figure 4.9 Endocycle defects in E2f7 and E2f8 null TGCs are p57 independent ...... 118
Figure 4.10 Deregulation of key cell cycle genes as molecular basis for endocycle defects in E2f7/E2f8 null TGCs ...... 119
Figure 4.11 Validation of changes in transcriptome in E2f7/E2f8 null TGCs ...... 120
Figure 4.12 Derepression of G2/M targets in E2f7/E2f8 null TGCs ...... 121
Figure 4.13 Direct regulation of G2/M targets by E2F7 and E2F8 ...... 122
Figure 4.14 E2f7 and E2f8 regulate hepatocyte endocycles ...... 123
Figure 4.15 Similar molecular mechanisms underlie endocycle defects in E2f7/E2f8 null
TGCs and hepatocytes ...... 124
Figure 5.1 Transcriptional derepression of cyclin A2 in E2f7 and E2f8 null TGCs ...... 136
Figure 5.2 Added loss of cyclin A1/A2 rescues phenotypic defects in E2f7/E2f8 null
TGCs ...... 137
Figure 5.3 Added loss of cyclin A1/A2 partially rescues aberrant mitosis in E2f7/E2f8 null
TGCs ...... 138
Figure 5.4 Added loss of cyclin A1/A2 restores polyploidy in E2f7/E2f8 null TGCs ... 139 xxii
Figure 6.1 Loss of E2f7 and E2f8 leads to upregulation of E2f4 levels in TGCs ...... 144
Figure 6.2 Expression of canonical E2F activators in TGCs ...... 145
Figure 6.3Canonical E2F repressors E2F4 and E2F5 have no significant role in regulation of endocycle in TGCs ...... 146
Figure 7.1 Loss of E2f7 and E2f8 leads to upregulation of E2f3 levels in TGCs ...... 155
Figure 7.2 Expression of E2f activators in TGCs ...... 156
Figure 7.3 Loss of E2f activators promotes TGC endocycles ...... 157
Figure 8.1 Added loss of E2f1 or E2f3a rescues nuclear phenotypes in E2f7/E2f8 null
TGCs ...... 165
Figure 8.2 Ablation of E2f3a and E2f1 partially rescues cell cycle defects in E2f7/E2f8 null TGCs ...... 166
Figure 8.3 Ablation of E2f3a and E2f1 rescues ploidy defects in E2f7/E2f8 null TGCs
...... 167
Figure 9.1 E2f7/E2f8 deficient megakaryocytes have smaller nuclei ...... 176
Figure 9.2 Loss of E2f7 and E2f8 leads to significant decrease in ploidy in megakaryocytes ...... 177
Figure 9.3 E2f7/E2f8 null megakaryocytes generate sufficient platelets to maintain platelet count within normal limits
...... 178
xxiii
Figure 9.4 Loss of E2f1, E2f2 and E2f3 leads to significant decrease in ploidy in megakaryocytes ...... 179
xxiv
Chapter 1: Cell cycle regulation and current CDK-RB-E2F paradigm
Mammalian cells respond to various external mitotic stimuli by activation of signaling cascades that carry them through cell cycle to generate two genetically identical daughter cells.
Critical steps in these proliferative signaling cascades involves the activation of G1-specific cyclin-dependent kinases (Cdks), the phosphorylation of retinoblastoma (Rb) and Rb-related pocket proteins and the accumulation of E2F transcriptional activity (Figure 1.1). The execution of E2F-dependent transcription late in G1 phase is believed to be the final event in Cdk-mediated mitogenic signaling that commits cells to S phase entry, while subsequent waves of E2F- mediated repression are thought to coordinate the completion of remaining cell phase specific events and successful cell divisions. This classic paradigm of E2F-mediated gene activation and repression in the control of cell cycle progression is based almost exclusively on the analyses of invertebrates and overexpression strategies in mammalian cell culture systems and very little evidence is available to support that this balanced E2F activity is essential or in existence in vivo.
1.1 E2F family of transcription factors
E2Fs (Early region 2-associated factors) were identified as host factors that bind to viral gene promoters, regulate expression of viral oncogenes and promote proliferation in cells infected with DNA tumor viruses (Kovesdi et al., 1986). Since discovery of the founding E2F family member, E2f1 (Helin et al., 1992; Kaelin et al., 1992), over the past two decades eight
1 additional E2F family members have been identified in mammals (Figure 1.2, adapted from
Chen et al., 2009a). The observation that over-expression of E2F1 could trigger quiescent cells to enter cell cycle (Johnson et al., 1993) was followed by identification of many critical cell cycle components as E2F-regulated genes and drove further investigations into how regulation of these genes by E2Fs modulate cell cycle progression. These investigations revealed, in addition, upstream molecular events that regulate E2F activity as post-translational modifications of retinoblastoma (RB) and RB-related proteins by cyclin-Cdk complexes. Over the past years, the complexity of regulation of E2F targets and the cell cycle phase specific cumulative E2F activity has been a matter of elaborate study. More and more evidence have emerged to support roles of regulatory components in defining unique functions played by individual family members in different cell cycle phases, tissue and development contexts and their regulation through degradation and phosphorylation of DP (Campanero and Flemington, 1997; Krek et al., 1995).
Theses regulatory factors assure that sequential waves of E2F-dependent transcriptional activation and repression occurs in a well-coordinated manner to drive cell cycle in a timely manner through all four phases of the cell cycle: G1, S, G2 and M (Figure 1.3, adapted from Chen et al., 2009a).
Proteins related to the E2Fs are conserved across various species of plants and animals.
For example, in Drosophila melanogaster, there is only one activator E2F (dE2F1) and one repressor E2F (dE2F2), with no homologues for mammalian atypical E2F repressors identified yet (van den Heuvel and Dyson, 2008). In Caenorhabditis elegans three E2F-like proteins have been described, EFL-1, EFL-2 and EFL-3. Studies have shown mammalian RB-E2F like role of
EFL-1, DPL-1 (DP) and LIN-35 (Rb) in regulation of proliferation and apoptosis (Ceol and
Horvitz, 2001; Myers and Greenwald, 2005; Winn et al., 2011). EFL-1 is most closely related to
2 the mammalian repressor E2F4 while EFL-3 is considered as the homologs of mammalian
E2F7/E2F8, but the function of EFL-2 is unknown (Li et al., 2008b; Winn et al., 2011). In
Arabidopsis thaliana, two activators and four repressors have been identified with E2Fa/E2Fb capable of interacting with DP to bind DNA and activate E2F-responsive genes (Mariconti et al.,
2002; Sozzani et al., 2006). On the other hand, E2Fc has a truncated C-terminal transactivation domain and has been shown in vitro to down-regulate the expression of E2F targets making it similar to mammalian E2F6 (Kosugi and Ohashi, 2002). The “atypical” E2Fd-f are considered the plant homologues of mammalian E2F7/E2F8 with documented functions in regulation of cell growth and endoreplication (Lammens et al., 2008a; Lammens et al., 2009).
By positive or negative regulation of its target genes, mammalian E2Fs have been shown to regulate processes as diverse as DNA replication, DNA repair, mitosis, apoptosis, and differentiation. The mammalian E2F family consists of nine related proteins (DeGregori and
Johnson, 2006a) that based on sequence conservation and structure-function studies have been historically classified into transcription activators and repressors. Canonical E2F activators, consisting of E2F1, E2F2, E2F3a and E2F3b, have transactivation domains and are known to associate with co-activator proteins to robustly induce RNA polymerase II-dependent target gene expression (Danielian et al., 2008; Trimarchi and Lees, 2002). E2F repressors fall into two subclasses, with E2F4, E2F5 and E2F6 in one subclass (canonical) and E2F7 and E2F8 in the other (atypical) based on the number of DNA binding domains. E2F1-E2F5 are thought to influence recruitment of transcriptional machinery and chromatin remodeling by binding the Rb pocket proteins., while E2F6, E2F7, and E2F8 are found to repress transcription without direct interaction with pocket proteins (Figure 1.5, adapted from Chen et al., 2009a). While E2F4-6 mediated repression is responsive to Cdk signaling and involves the recruitment of histone
3 deacetylases (HDACs), polycomb group proteins as well as Mga and Max to E2F-target promoters (Attwooll et al., 2004), the mechanisms of how E2F7/E2F8 mediate repression is essentially unknown.
Canonical E2F repressors E2f4 and E2f5 are constitutively expressed throughout the cell cycle and are loaded onto E2F target genes primarily in G0, whereas the activator subclass of
E2Fs is regulated in a cell-cycle dependent manner by transcription, post-transcriptional regulation by miRNAs, post-translational modifications, binding to cofactors, and proteolysis
(Sylvestre et al., 2007; Takahashi et al., 2000; Welch et al., 2007). During G0, canonical repressor E2Fs form complex with histone deacetylases (HDACs), chromatin-remodeling proteins such as BRM/BRG-1, and histone methyl transferases such as SUV39H1 (DeGregori,
2006; Johnson and Degregori, 2006). On cell-cycle re-entry, E2F4/p130 repressor complexes and
HDACs are replaced on promoters by E2F1-E2F3 and activator E2Fs typically form complexes with histone acetyl transferases (HATs) such as p300/CBP, P/CAF, and Tip60 (Johnson and
Degregori, 2006). The loading of E2F1-E2F3 and HATs promotes histone acetylation, open chromatin structure, and increased E2F target gene expression (Frolov and Dyson, 2004). Thus
E2F repressors and activators can function sequentially to coordinate the expression of genes required for S phase and progression through mitosis by binding to same promoters in a cell cycle-dependent manner.
The question of how the E2F-responsive genes are turned off as cells move from one phase to another is not clear yet, though existing evidence suggests that this is driven mostly through decrease in binding of E2F activators to target genes owing to their degradation and DP phosphorylation dependent decline in affinity and direct repression by E2F repressors. Three major candidates which potentially could have significant role in this context are E2F6-8, as they
4 are expressed in late S-G2 phase of cell cycle. Published results from germline ablation studies of E2f6 have failed to substantiate presence of such a critical role for this E2F factor (Pohlers et al., 2005). Unlike other E2F family members, E2F7 and E2F8 are unique as they associate with
DNA independent of dimerization with DP1/DP2 proteins. Instead they utilize two tandem
DNA-binding domains to recognize and bind target DNA sequences. These two atypical E2Fs also lack amino acid sequences typically used to physically interact with Rb related proteins, and thus may function outside the canonical Cdk-Rb-E2F pathway. Though considered highly likely, whether or not E2F7 and or E2F8 will play the critical role of downregulating E2F-responsive genes as the cells progress through cell cycle is a question for which no conclusive in vivo evidence exists currently.
1.2 Complex specific, synergistic and antagonistic in vivo roles of E2F family members
Several compounding factors have been deterrent to analysis of mammalian E2F family members in vivo. These include the sheer number of family members, difference in functional domains and expression pattern among family members, cell type specific unique functions, redundancy among family members and promiscuity in binding to DNA elements. The analyses of mice deficient for various E2F family members have shown a spectrum of tissue specific phenotypes that are inconsistent with the rigid view of E2Fs as universal factors ubiquitously required to coordinate cell cycle dependent gene expression programs (Chen et al., 2009; Chong et al., 2009b; Cloud et al., 2002; Danielian et al., 2008; Field et al., 1996; Humbert et al., 2000;
Kinross et al., 2006; Li et al., 2003; Li et al., 2008a; Lindeman et al., 1998; Murga et al., 2001;
Pohlers et al., 2005; Rempel et al., 2000; Tsai et al., 2008; Yamasaki et al., 1996). These studies suggest that E2F family members either have unique functions or perform similar functions in a
5 tissue-specific manner. It is also possible that the ablation of individual family members is insufficient to expose how their combined activities might be coordinated in vivo.
E2Fs of the same subclass are known to be able to functionally compensate for loss of similar family members, leading to the notion that there is a good deal of functional redundancy among the E2F family members. Although there is functional overlap, E2Fs have shown to play very different roles in mediating processes such as differentiation, apoptosis, and DNA-damage response. A good deal of work has been done to understand the role of E2Fs in vivo, especially in the context of Rb loss, in an effort to identify redundancy and specificity of each of the E2F family members. For example, proliferative and apoptotic phenotypes of Rb deficiency were partially rescued by the genetic inactivation of E2f1, E2f3a or E2f3b, demonstrating significant functional redundancy for E2F1, E2F3a and E2F3b at protein level during development which also extended the viability of Rb knockout embryos to late stages of embryogenesis (Chong et al.,
2009a; Tsai et al., 1998). The high degree of functional redundancy among member of activator and repressor subclasses has been further revealed by the systematic analyses of mice harboring single and compound E2f mutations (DeGregori and Johnson, 2006a; Li et al., 2008b).
On the other hand, E2F knockout mice have also provided evidence for non-redundant tissue specific functions. For example, E2f1-/- mice are predisposed to testicular atrophy; E2f2-/- mice have mild proliferation defects in hematopoietic lineages; E2f4-/- mice are runted and have hematopoietic, craniofacial, and intestinal defects; and E2f5-/- mice suffer from hydrocephalus.
Some of these studies have been designed such that specificity of E2F function is analyzed in an
Rb-/- background, in part due to the fact that the phenotypes of E2F knockout mice present later in adulthood (DeGregori, 2002). As an example, ablation of E2f1, E2f2, or E2f3 in Rb-/- embryos
6 abrogates unscheduled proliferation in the lens and retina, whereas only E2f1 or E2f3 deletion suppresses ectopic apoptosis (Saavedra et al., 2002; Ziebold et al., 2001).
Specificity of E2F functions is currently believed to be achieved through regulation of
E2F activity via binding with different affinities to the pocket proteins, various cofactors and specificity to promoter sequences (Croxton et al., 2002; Giangrande et al., 2007; Hauck et al.,
2002; Schlisio et al., 2002; Trimarchi and Lees, 2002). Immunoprecipitation assays in vitro have shown that E2Fs associate with different chromatin remodeling enzymes such as histone deacetylases or histone methyltransferases to modulate the transcriptional landscape (Blais and
Dynlacht, 2007). Given that changes in gene expression may reflect both direct and indirect functions of transcription factors, several laboratories have extensively utilized ChIP-based technologies to identify direct E2F-regulated targets (Takahashi et al., 2000; Wells et al., 2000;
Jin et al., 2006; Xu et al., 2007). These approaches have identified many novel E2F-responsive genes, consistent with numerous reports that have exposed context-dependent functions in differentiation and development for different E2Fs in vivo (McClellan and Slack, 2007). But to add to the complexity, it is now clear that at least in some developmental contexts, E2F activators can also function to repress gene expression, though the molecular basis for such plasticity is not completely understood (Chen et al., 2009; Chong et al., 2009b; Trikha et al.,
2011; Wenzel et al., 2011).
One of the unresolved questions in E2F biology is the non-correlation of phenotypic and molecular antagonism seen between E2F activators and repressors. Early studies in Drosophila melanogaster, which contain just one E2F activator and one repressor (dE2F1 and dE2F2, respectively), appeared to validate the relatively straightforward model of E2F function as an activity that either promoted or inhibited progression of dividing cell through the cell cycle. Loss
7 of function mutations in dE2f1 has been shown to reduce expression of classic E2F-regulated genes, DNA synthesis and larval survival (Duronio et al., 1995; Frolov et al., 2001). As anticipated, loss of dE2f2 reversed the defects in proliferation in dE2f1 mutant larvae and significantly delayed lethality (Frolov et al., 2001). These results suggest that the net effect of
E2Fs on cell proliferation appeared to depend on the balanced activation and repression of a core set of target genes. However, subsequent gene expression analysis in S2 cells lacking dE2f1, dE2f2 or dDP revealed surprisingly little overlap in genes regulated by dE2f1 and dE2f2
(Dimova et al., 2003). Consistent with this observation, genetic mosaic screens found that the cell cycle defects induced by loss of dE2F1 could be rescued without functionally inactivating dE2f2 and suppressing dE2F2-mediated repression (Ambrus et al., 2007). Thus, instead of regulating a common set of target genes, dE2F1 and dE2F2 in this context seems to be contributing independently to the cumulative proliferative potential of a cell.
Ability of E2F family members to regulate the expression of one another is another factor that further adds complexity to identifying direct consequences of loss of different E2F family members in mammals. It has been known for long that most of the E2f family members had E2F consensus binding sequences in their promoters. A clear example of this kind of feedback regulation was demonstrated recently by our laboratory using molecular analysis of mouse embryos doubly deficient for E2f7 and E2f8. Consistent with chromatin immunoprecipitation
(ChIP) experiments that demonstrated loading of E2F7 and E2F8 to the E2f1 promoter (Di
Stefano et al., 2003; Christensen et al., 2005), ablation of E2f7/E2f8 in the mouse embryo resulted in derepression of E2f1, ectopic levels of E2F1 protein and E2F1-dependent apoptosis.
On the other hand, the presence of multiple putative E2F binding sites in the promoter regions of
E2f7 and E2f8 indicates that their expression may be activated by E2F1. Thus E2F1 function,
8 peaking at the G1/S transition, robustly transactivates the expression of E2f7 and E2f8 in addition to genes essential for S phase progression. As a result, the level of E2F7 and E2F8 proteins accumulate and reach maximal levels in late S phase or near S/G2, which coincides with the timely cessation of DNA replication that in part is due to direct repression of E2f1 (Moon and
Dyson, 2008). The study discussed in here shows that such a feed forward regulatory loop exists for E2F3a also, though we have not observed direct repression of E2f3a by E2F7 and E2F8.
Recent work from the Farnham laboratory has added another level of complexity to this by demonstrating that E2F1 could be recruited to a number of gene promoters that do not contain consensus E2F binding sites (Bieda et al., 2006, Cao et al., 2011). The promiscuity in E2F- chromatin associations could be attributed to either the co-recruitment of E2Fs by other transcription factors such as NF-Kβ (Lim et al., 2007), MYC (Leung et al., 2008) and CEBPα
(Iakova et al., 2003), or by components of the transcriptional machinery, but underlying physiology is still unclear.
1.3 E2F transcription factors and regulation of mitotic cell cycle
Mitotic cell cycle in mammals is tightly regulated by a cascade of signaling and downstream responsive components that includes growth factors, cyclins, protein kinases, proteases, transcription factors, chromatin remodeling factors, DNA repair machinery to ensure proper cellular responses to external signaling and fidelity in duplication of the genome (Morgan
1997; Sherr and Roberts 1999, Sherr and Roberts 2004; Machida et al. 2005). On mitogenic stimulation, cyclin D accumulates within the nucleus and activates cdk4 and cdk6 cyclin- dependent kinases (cdks). Among the key targets of these cdks is the retinoblastoma (RB) family of pocket proteins. Phosphorylation of pocket proteins in complex with E2F transcription factors
9 triggers the activation of genes required for the G1/S transition. Before the cell can begin the process of duplicating the genome, numerous sites on the DNA are occupied by a group of proteins called the origin recognition complex. Replication initiation at multiple sites facilitates genome duplication to be completed quickly with the process converging finally at junctions between replicative forks along the length of the DNA. As an initial step, in G1 Cdc6 and Cdt1 join the ORC family of proteins and together recruit the minichromosome maintenance helicase
(Mcm2-7) (Mendez and Stillman, 2003; Takahashi et al., 2005).
The formation of the pre-RC (ORC, Cdc6, Cdt1, and Mcm2-7 complex) is referred to as licensing and is thought to limit DNA replication to a single round per replication cycle (Blow and Laskey 1988). In fact, there are several key mechanisms that ensure that the cell replicates its genome only once, including destabilization of the pre-RC components by cdk phosphorylation, degradation of Cdt1 by the proteasome, and inhibition of the loading of Mcm2-7 helicase onto the pre-RC by geminin (Blow and Dutta 2005). DNA replication licensing and the cell cycle are closely linked via E2F-mediated transcriptional upregulation of cyclins and the APC-inhibiting factor Emi1, which promotes the accumulation of cyclin proteins (Hsu et al. 2003). In late G1 and S phase, cyclin A and cyclin E levels increase and form active cdk2-cyclin A and cdk2- cyclin E complexes. One important substrate of cdk2-cyclin A during late G1 is the APC subunit cdh1, which upon phosphorylation inhibits APC-mediated proteolysis of mitotic cyclins and allows the accumulation of cyclins necessary for entry into S phase and later for completion of cell division (Hsu et al 2002; Lukas et al 1999; Kramer et al. 2000). Increased cdk activity aids not only in the maintenance of APC inhibition but also in the recruitment of DNA polymerases and replicative machinery to origins of replication. Finally, as DNA replication completes and
M phase begins, cytoplasmic cyclin B1 translocates into the nucleus and binds Cdc2 kinase
10
(Cdk1), thereby driving the cell through mitosis and pushing cells to divide (Nurse 1990; Pines and Hunter 1991).
Though this model suggests that most of the components of this well-orchestrated machinery will be indispensable for cell cycle progression, over past years this paradigm has been changing with results from studies utilizing transgenic mouse models. One could attribute this to compensation by different family members or by a related family member. But nevertheless, work done over the past few years highlight the complexity and redundancy of this process in mammals compared to lower organisms. In light of the apparent importance of cyclins in the progression of the cell cycle, it is surprising that all of the known cyclins have been found to be dispensable for normal mitotic cycling (Sherr and Roberts, 2004; Santamaria and Ortega,
2006). Even targeted deletion of multiple cyclins simultaneously to minimize compensation by members with overlapping or redundant functions has failed to result in a definitive cell cycle arrest in vivo. Many of the knockout mice generated with single or multiple gene deletions are either viable into adulthood or die between mid- to late-gestation (embryonic day 10.5 – 16.5).
Cyclin A2 is perhaps the only exception, in that its loss causes early embryonic lethality after implantation around E5.5 (Murphy et al., 1997). This was explained later that maternal pool of cyclin A2 exists in the blastocyst that allows embryos to reach early stages of post-implantation
(Murphy et al. 1997). Even cyclin D, which is the first cyclin to respond to mitogenic signals, is dispensable. Simultaneous knockout of cyclins D1, D2, and D3 leads to defects in fetal hematopoiesis but embryos survive to E16.5 (Kozar et al., 2004). So, although ablation of the cyclin genes eventually leads to defects associated with impaired cellular proliferation either late in embryogenesis or during adulthood, the cyclins are not critically required for progression of the cell cycle. The cdks that associate with these cyclins have also been targeted for deletion in
11 mice singly and in combination. Cdk2, cdk4, and cdk6 knockout mice are all viable, although cdk2 and cdk4 deletion causes sterility in males and females due to defects in meiosis (Rane et al., 1999; Tsutsui et al., 1999; Berthet et al., 2003; Ortega et al., 2003; Malumbres et al., 2004).
Double deletion of cdk2 and cdk6 are viable (Malumbres et al., 2004). In contrast, double deletion of cdk4 and cdk6 results in embryonic lethality between approximately E14.5 to E18.5 with defects in fetal hematopoiesis (Malumbres et al., 2004). Naturally, due to the possibility of functional redundancy and compensation among the cdks, it could be that triple deletion of cdk2, cdk4, and cdk6 would result in dramatically different cell cycle kinetics or complete failure to respond to mitogenic signals. One of the biggest drawbacks in the conclusions drawn out of these embryonic developmental studies is that the differences in mitotic cycles in pluripotent cells and differentiated cells are not considered, where redundancy and compensation for loss of these cell cycle genes could be completely different.
1.4 Constitutive Derepression of E2F-regulated genes in stem cells
The mitotic cell cycle in stem cells is characterized by truncated gap phases, elevated levels of cdk activity that lack periodicity, and constitutively high expression of many replicative and cell cycle regulatory genes (Stead et al., 2002; Fujii-Yamamoto et al., 2005). p53 protein is abundant, and, yet, embryonic stem cells can efficiently activate checkpoints and repair DNA independent of p53 (Sabapathy et al., 1997; Aladjem et al., 1998; Prost et al., 1998). On release from mitotic arrest, stem cells begin DNA replication in two hours whereas fibroblasts require ten to fifteen hours to initiate DNA synthesis (Savatier et al., 1994). This accelerated entry into S phase and rapid progression through M phase could be explained by the constitutively high levels of cyclins, which do not oscillate as they would during normal somatic cell division (Stead
12 et al., 2002). Cell cycle periodicity is found only for Cdc2-cyclin B activity, although only a moderate fluctuation is apparent (Stead et al., 2002). Interestingly, cyclin D-associated kinase activity is present in stem cells but is undetectable in the early embryos prior to gastrulation
(Savatier et al., 1996; Wianny et al., 1998; Faast et al., 2004). Similarly, during early development of Xenopus, Danio, and Drosophila, levels of cyclins A and E do not dramatically fluctuate until cells have begun to differentiate into the three germ layers (Rempel et al. 1995), consistent with the idea that cdk regulation may provide an additional level of control over the
G1/S transition in differentiated cells. Hence, it could be that the transition from G1 to S phase is regulated by a minimal mechanism in stem cells and, as cells commit to specific lineages, more sophisticated mechanisms that include cdk-mediated checkpoints are implemented to control S phase entry (Burdon et al. 2002).
Since E2f1-/-; E2f2-/-; E2f3-/- embryos do not die until after E9.5, it would be expected that embryonic stem cells do not require the activator E2Fs for entry into S phase and completion of the cell cycle. In fact, E2f1-/-; E2f2-/-; E2f3-/- ES cells have been found to proliferate and express
E2F target genes at normal levels (G Leone, unpublished data). The fact that E2f7-/-; E2f8-/- embryos also can survive until embryonic age 11.5 suggests that E2F7 and E2F8 might also have significantly different functions in stem cells and differentiated cells (Li et al, 2008). Chromatin in pluripotent cells is dramatically different from that of differentiated cells, in that histones and
DNA contain less methylation and acetylation. It could be that the open state of the chromatin in
ES cells permits constitutive expression of E2F target genes and the transactivation by E2Fs is probably not required. E2F targets are constitutively active in ES cells perhaps not because Rb is hyperphosphorylated but because DNA is de-repressed to be acted upon by other transcription
13 factors. Thus, the open chromatin state of pluripotent cells allows E2F targets to be active in the absence of the E2F activators.
The association of Rb with chromatin remodeling factors such as HP1, histone deacetylases, methyltransferases, components of the SWI/SNF remodeling complex, and
Polycomb group proteins suggests that Rb may be important for the establishment of localized closed DNA conformations, and perhaps also heterochromatic regions, during the differentiation process (Dunaief et al. 1994; Brehm et al. 1998; Luo et al. 1998; Magnaghi-Jaulin et al. 1998;
Robertson et al. 2000; Strobeck et al. 2000; Zhang et al. 2000; Dahiya et al. 2001; Nielsen et al.
2001; Vandel et al. 2001). The role of Rb in ES cells could be to coordinate expression of E2F targets in a cell cycle dependent manner not via repression of E2F transactivation activity but by remodeling of chromatin during commitment to the differentiation process. E2F repressors may cooperate with pocket proteins to coordinate developmentally regulated transcriptional programs by recruitment of chromatin remodeling factors. Only after cells have committed to a specific lineage would E2F activators impose cell cycle-dependent gene activation. Whereas E2F repressors are important for establishing target gene expression patterns early in the differentiation process, fine tuning the cycling of S phase genes or inducing checkpoint readiness during DNA repair or replication, is currently unknown.
1.5 E2F-Mediated activation and repression in differentiated cells
Once differentiated the complexity of cell cycle increases to accommodate expression of unique cell specific developmentally-regulated genes. Gap phases increase and cell-cycle dependent genes adopt cyclic expression patterns (Stead et al. 2002; Faast et al. 2004; White et al. 2005). Cell cycle periodicity of cyclins is also more pronounced, likely due to changes in
14 proteolysis and transcription of E2F target genes (White et al. 2005). Previously hyperphosphorylated Rb pocket proteins acquire the ability to bind E2Fs, possibly via dephosphorylation caused by the down-regulation of cdk6-cyclin D3 (Savatier et al. 1994;
Savatier et al. 1996; Stead et al. 2002; Faast et al. 2004). The mammalian genome undergoes dynamic remodeling of chromatin structure, presumably to direct transcription of genes essential for embryonic patterning, cell commitment, and proliferation. Methylation and acetylation patterns change both globally throughout the genome and surrounding specific developmentally- regulated genes (Kafri et al. 1992; Brandeis et al. 1994; Macleod et al. 1994; Lee et al. 2004).
Cytosine methylation, a common characteristic of heterochromatin, increases globally and at specific CpG motifs associated with genes that are silenced as cells commit to specific lineages, such as Oct4 (Gidekel and Bergman 2002; Lee et al. 2003). Methylation of histones associated with heterochromatic (H3-Lys9) regions increases and methylation in euchromatic (H3-Lys4) areas decreases (Lee et al. 2003). Global histone acetylation declines rapidly (Lee et al. 2003). In fact, deacetylase activity has been shown to be required for differentiation of ES cells in vitro, supporting that there is a critical role for histone acetylation state, and more generally for chromatin structure, in the regulation of genes important for development (Lee et al. 2003).
Modifications of nuclear histone proteins unfold or compact chromatin to facilitate or impede, respectively, the loading and activation of the transcriptional machinery. E2F family members contribute to the organization of both euchromatin and heterochromatin. During cell proliferation when RB is hyperphosphorylated, E2F activators recruit the basal transcription factor TFIID and other co-activators such as histone acetyltransferases (HATs) p300/CBP, GCN5 and Tip60 to specific gene promoters (Taubert et al., 2004; Nicolas et al., 2003; Lang et al., 2001), leading to an open chromatin configuration and transcription initiation. In quiescent or differentiated cells,
15 pocket protein-bound E2F4/E2F5 have been found to associate with a variety of co-repressors such as histone methyltransferases (HMTs) and deacetylases (HDACs), DNA methyltransferase
(DNMT1) and C-terminal binding protein (CtBP), leading to chromatin compaction and transcription inhibition (Tyagi et al., 2007; Ogawa et al., 2002; Rayman et al., 2002; Ferreira et al., 1998). The more recently identified repressors E2F6-8 mediate repression of E2F-responsive genes independent of binding to pocket proteins. While E2F6, through interaction with the
Polycomb complex (Trimarchi et al., 2001), still requires dimerization with DP to function in transcription repression, E2F7/E2F8 are distinctly unique in that they form homo (E2F7-E2F7,
E2F8-E2F8) or heterodimers (E2F7-E2F8) to repress transcription of cell cycle-related genes (Li et al., 2008; Di Stefano et al., 2003; Christensen et al., 2005). The co-repressors that associate with E2F7/E2F8 in vivo are currently unknown and remain to be identified (Figure 1.5).
Coincident with these global and localized changes in chromatin structure is the activation of the Rb pocket protein family (White et al. 2005). Prior to differentiation, the Rb pocket proteins are constitutively hyperphosphorylated and unable to bind to E2F (Savatier et al.
1994; Savatier et al. 1996). This prevents pocket protein-mediated inhibition of E2F transactivation and recruitment of chromatin remodeling factors to E2F target genes in pluripotent cells. The increase in E2F-binding activity of Rb, p107, and p130 in response to differentiation signals may reflect a requirement for the pocket proteins in organizing chromatin structure and imposing specialized gene expression patterns during development. This activation of the pocket proteins during the differentiation process could also explain why E2F targets that were elevated in a cell-cycle independent manner in undifferentiated ES cells begin to adopt cell cycle periodicity upon commitment to specific cell lineages. The activation of Rb, p107, and p130 could, in fact, be a key event that coordinates cell cycle periodicities and activation of
16 genes that control cell fate decisions, and the pocket proteins may mediate these processes through remodeling of chromatin in localized and global regions of DNA and by direct transcriptional control of E2F targets such as cyclins A and E.
The Drosophila genome has provided the Rb-E2F field with a powerful reagent to study the effects of loss of E2F activator and repressor functions in differentiated cells. Loss of the activator, dE2f1, causes slow growth and abnormal larval development; however, concomitant loss of the repressor, dE2f2, substantially suppresses the defects caused by loss of dE2f1 (Frolov et al. 2001). Further, while proliferation is blocked in eye and wing imaginal discs of dE2f1 mutant flies, dE2f1; dE2f2 double mutants display relatively normal levels of DNA synthesis in the eye discs (Frolov et al. 2001). Ablation of E2F-mediated repression by use of a mutant that does not bind RBF (Drosophila Rb homolog) but can transactivate target genes leads to severe defects in endoreplication when dE2f2 is absent (Weng et al. 2003). Interestingly, all tissues of the developing fly that do not rely on endocycle are normal. Hence, in Drosophila, E2F is not required for normal proliferation; however, these results show that the net effect of having both repressor and activator functions leads to a tighter control over the cell cycle.
In mammalian cells, the existence of multiple E2F family members that can compensate for one another has complicated the genetic analysis of E2F function. With the exception of
E2F3, single deletion of activator or repressor E2Fs does not impair cellular proliferation in mouse embryonic fibroblasts (MEFs) (Humbert et al., 2000; DeGregori at al., 2002). Analysis of the effect of loss of E2F dimerization partner (DP) proteins, however, could provide a simpler system to study E2F function. Both the activator E2Fs (E2F1, E2F2, and E2F3a) and the repressor E2Fs (E2F3b, E2F4, E2F5, and E2F6) need to dimerize with DP1 or DP2 to bind
DNA. Hence, with the exception of E2F7 and E2F8, ablation of DP1 and DP2 would
17 functionally eliminate all E2F activity. In concordance with studies done in Drosophila, it would be expected that concurrent ablation of repressor and activator E2Fs in mammalian cells would impair the ability of cells to exit the cell cycle but may not block proliferation (Attwooll et al.
2004). In fact, Dp1-/- embryonic cells do not have defects in proliferation; however, embryos die during embryogenesis between E10.5-12.5 due to placental abnormalities, including impaired proliferation in trophoblasts and reduced DNA synthesis and nuclear size in giant cells (Kohn et al. 2003). No clear defects in differentiation were observed. An important consideration, however, is that there is likely to be compensation or redundancy of function between DP1 and
DP2 and, until the DP proteins are knocked out in combination, it will be difficult to know what
E2F function mediates the lethality of Dp1-deficiency. The caveat to these types of experiments that utilize DP ablation, of course, is that it is difficult to tease apart functional contributions from the E2F repressor and activator subclasses. Concomitant knockout of three or more genes in mice can be tedious; however, combinatorial means of deleting the E2F family members may be the most direct way to discriminate between the activator and repressor functions of the E2F family members.
Despite obvious challenges associated with redundancy of function, efforts have been made to knock out mouse E2Fs in combination. Deletion of all repressor E2Fs is not terribly practical, as there are five E2Fs with clear repressor activity. E2F4 and E2F5 have been deleted in combination (Gaubatz et al. 2000), but as of yet no one has attempted the formidable task of deleting those concomitant with E2F6, E2F7, and E2F8. Ablation of E2F activator function has been more feasible, and the effect on cell cycle progression is dramatic. Wu et al. (2001) found that conditional deletion of E2f3 on an E2f1-/-; E2f2-/- background resulted in cell cycle arrest of
MEFs, demonstrating that the activator subclass of E2Fs is essential for transactivation of targets
18 that drive entry into S phase and progression through the cell cycle. In the absence of E2F1,
E2F2, and E2F3, fibroblasts arrest in all stages of the cell cycle and adopt senescent-like morphologies (Wu et al. 2001; Sharma et al. 2006; Timmers et al. 2007). Only with loss of p53 can MEFs deficient for E2F1, E2F2 and E2F3 enter S phase and complete the cell cycle
(Timmers et al. 2007). E2f1-/-; E2f2-/-; E2f3-/- mice do not survive to birth, as would be expected from experiments done in MEFs (Wu et al. 2001). What is unexpected, however, is that triply- deleted embryos can develop to at least E9.5 (G Leone, unpublished data), which may correspond closely to the time of death of the Dp1 mutant (E10.5-12.5) (Kohn et al. 2003).
Taken together with the time of lethality of the Dp1 mutant, this suggests that cells in the developing embryo and extra-embryonic lineages can proliferate in the absence of activator E2Fs and, possibly, that inactivation of repressor E2Fs would not substantially extend the life of the
E2f1-/-; E2f2-/-; E2f3-/- embryos.
Although E2F1, E2F2, and E2F3 can be activators of transcription, they do not always promote gene expression. Analysis of E2F target expression shows that some targets are elevated in E2f1-/-; E2f2-/-; E2f3-/- MEFs (cyclin E) whereas others are down-regulated (Dhfr, Mcm3, Tk,
Cdc6, Polα, cyclin A), suggesting that some of the targets are actively repressed and some are transactivated by E2F1-3 (Wu et al. 2001). It is possible that E2F1-3 function primarily as anti- repressors and they occupy E2F sites on promoters in order to prevent the binding of E2F3b-8
(Rowland and Bernards 2006). Differences in binding to promoters and recruitment of pocket proteins could account for why activator E2Fs appear to have both activator and repressor activities (Chen et al. 2002; Dahiya et al. 2001; Luo et al. 1998; Zhang et al 2000; Ait-Si-Ali et al. 2004; Balciunaite et al. 2005). These results were further validated in vivo by gene ablation
19 studies in mouse intestine, lens, hematopoietic system and retina (Chen et al., 2009; Chong et al.,
2009b; Trikha et al., 2011; Wenzel et al., 2011).
Although E2f3-/- MEFs exhibit inhibited growth, the defects in proliferation are the result of relaxed p19ARF repression (Humbert et al. 2000; Aslanian et al. 2004). p19ARF is normally repressed by E2F3, and concomitant ablation of the two genes relieves the cell cycle block
(Aslanian et al. 2004). Not only does this demonstrate repressor function in an activator but also it shows that E2F3 is not strictly required for proliferation. Similar conclusions can be drawn from studies that have utilized a mutant of E2F that lacks the carboxyl terminus. Introduction of this dominant-negative form of E2F competes at E2F-responsive promoters with endogenous activator and repressor E2Fs. Interestingly, cells expressing this dominant-negative mutant proliferate and are resistant to exit from the cell cycle. Moreover, most E2F target genes are expressed, suggesting that target repression is an important function of E2F and that genes can be turned on even when transactivation is attenuated (Krek et al. 1995; Bargou et al. 1996; Zhang et al. 1999; Rowland et al. 2002; Gonzalo et al. 2005; Maehara et al. 2005). Taken together, it seems that our classical view of E2F activators and repressors must be refined to account for their dual functionality as positive and negative regulators of gene expression and that activation of some E2F target genes could be managed by de-repression.
1.6 Direct roles of E2Fs in DNA Replication Process
There is substantial evidence to support a direct role for E2Fs in regulation of DNA replication. In vivo analysis of genes in the Rb-E2F pathway suggests that deregulation of E2F activity leads to DNA replication defects. For example, DP1 knockout impairs proliferation and
DNA replication in the extraembryonic lineages (Kohn et al. 2003). Specifically, trophoblast
20 giant cells fail to efficiently endoreplicate, and cell numbers in the ectoplacental cone and chorion are greatly reduced. Also consistent with this phenotype, knockout of cyclins E1 and E2, important downstream targets of E2F, causes disruption of hematopoietic and extraembryonic endocycles, leading to substantially decreased amounts of DNA in megakaryocytes and trophoblast giant cells (Geng et al. 2003; Parisi et al. 2003). It seems then that cyclins E1 and E2 are dispensable for mitotic cycles in the embryo but are required for endoreplication. Disruption of this pathway in Drosophila leads to a similar outcome. Hypomorphic alleles of Dp and dE2f1 decrease chorion gene amplification whereas loss of RBf causes abnormally high levels of endoreplication (Royzman et al. 1999; Weng et al. 2003). Double deletion of RBf and dE2F1 rescues endocycle defects of both (Weng et al. 2003). Impaired DNA replication is likely due to disruption of replicative complexes resulting from changes in transcriptional regulation or physical associations. This suggests that a key role of Rb-E2Fs is either to control transcriptional activation of genes required for DNA synthesis or to regulate recruitment of DNA replication factors to origins of replication by association with E2F.
Factors that regulate gene expression have been postulated to also influence temporal and spatial activities of DNA replication initiation. In fact, it could be that E2F plays not only a transcriptional role in the control of DNA replication but also a direct physical role. Numerous studies show that Rb responds to anti-proliferative and developmental cues by transcriptional down-regulation of S-phase specific genes (Korenjak and Brehm 2005). Rb is also posited to suppress DNA replication initiation through transcriptional regulation of factors involved in localization and recruitment of the ORC, pre-RC licensing subunits, and replication factors
(Sever-Chroneos et al. 2001; Hsu et al. 2003). In addition, direct physical interactions between pocket proteins and replicative machinery, such as Mcm7, DNA polymerase α, and replication
21 factor C, could modulate origin firing and replicative processing (Takemura et al. 1997; Sterner et al. 1998; Pennaneach et al. 2001; Gladden et al. 2003). It is further possible that pocket proteins, in association with histone deacetylase, could locally remodel chromatin such that accessibility of DNA origins is reduced or blocked. Specifically, Rb-dependent acetylation of nucleosomes at origins could prevent the loading of replication machinery (Brehm et al. 1998;
Luo et al. 1998; Magnaghi-Jaulin et al. 1998; Lai et al. 2001; Aggarwal and Calvi 2004).
There are several lines of evidence that support the notion that Rb and E2F are physically present at DNA replication origins. The pattern of colocalization of RB pocket proteins during
G1/S with replicative machinery suggests that late G1 regulatory events, such as transcription, may be physically linked to the initiation of DNA synthesis. During G1 and early S phase, chromatin-associated protein complexes are apparent near the nucleous that contain replication proteins and members of the RB pocket protein family (Kennedy et al. 2000). Moreover, Rb, p107, and p130 are all able to bind Mcm7, an essential replication protein and subunit of the pre-
RC (Sterner et al. 1998). Rb, in particular, can be found in complex with Mcm7 in human cells, and binding of Rb to Mcm7 inhibits DNA replication in Xenopus egg extracts (Sterner et al.
1998). The localization of pocket proteins such as p130 and Rb to perinucleolar sites correlates well with factors that are required for replicative processing, such as p150 (CAF-1) and PCNA, during G1 phase and with foci actively synthesizing DNA during early S phase, as shown by
BrdU incorporation (Kennedy et al. 2000). As cells progress through S phase, the pocket proteins dissociate from these foci as sites of active replication distribute to regions throughout the nucleus (Kennedy et al. 2000).
In Drosophila there is strong evidence to support the model that Rb and E2F are present at DNA origins during S phase specifically to regulate DNA replication initiation (Royzman et
22 al. 1999; Bosco et al. 2001; Beall et al. 2002; Korenjak et al. 2004). Additional support for this model comes from studies that show an interaction between dE2F1 and the DmORC (Drosophila melanogaster origin recognition complex) that is independent of DNA binding (Bosco et al.
2001). The dE2F1-Rbf complex could be bound near the DmORC at the replication foci, and through interactions with the DmORC, could regulate origin firing. In fact, Drosophila E2F mutants unable to bind DNA cause mislocalization of ORC2 but not Mcm protein during amplification of chorion gene clusters (Royzman et al. 1999). These and other E2F or RBF mutants fail to properly initiate replication, causing aberrant levels of DNA amplification in specific regions of DNA in ovarian follicle cells (Royzman et al. 1999; Bosco et al. 2001).
Although binding of dE2F1 to the ORC is not a prerequisite for replication, loss of DNA binding ability of dE2F1 could allow dE2F2 to bind sites normally occupied by dE2F1 thereby antagonizing DmORC binding (Royzman et al. 1999). Two distinct forms of an RB-E2F repressor complex were purified from Drosophila embryo extracts that contain dRB, dE2F2, and dMyb-interacting proteins (dREAM). Mip130/TWIT, Mip120, dMyb, CAF1p55, dE2F2, dDP,
Mip40, and either RBF1 or RBF2. Interestingly, some components of the complex are also found in a dMyb complex that regulates gene amplification in follicle cells (Beall et al. 2002). Some of the properties of the subunits that comprise the dREAM complex also hint at its possible role in
DNA replication. Specifically, depletion of dMyb by RNAi had no detectable effect on transcription, lending support to the idea that the Myb component of the dREAM complex may be primarily for control of DNA replication (Korenjak et al. 2004). A comparable complex has been isolated in human Burkitt lymphoma Raji cells, but precisely which E2F and Rb proteins are in the mammalian complex are still unknown (Maser et al. 2001).
23
1.7 E2Fs in DNA damage response and checkpoint activation
DNA repair is thought to be closely linked to the process of DNA replication. Not only is
Rb required for proper control of entry into S phase, it is required for DNA damage-induced cell cycle arrest at the G1/S phase checkpoint (Harrington et al. 1998). Rb-/- MEFs do not arrest in response to DNA damage, suggesting that the Rb-E2F pathway is a downstream target of DNA damage-sensing mechanisms. The mechanism for inhibition of S phase by Rb was suggested to be E2F-dependent transcriptional down-regulation of replication machinery, leading to disruption of replication complexes on chromatin, cessation of DNA synthesis, and cell cycle arrest (Angus et al. 2004). On the other hand, it could be that Rb physically resides at sites of
DNA replication ready to activate repair signals if damage occurs during DNA synthesis. In fact,
E2F1 is stabilized by the DNA-damage-response kinases ATM and ATR (Lin et al. 2001) and has been found to associate with the MRE11 recombination/repair complex in response to DNA damage (Maser et al. 2001). It could be that stabilized E2F1 recruits Rb and the Mre11 complex to origins of DNA replication to monitor replication initiation and suppress the firing of origins if
DNA is damaged.
E2Fs also regulates DNA damage response at the G2/M phase checkpoint (Ishida et al.
2001; Young et al. 2003; Eguchi et al. 2007). Ectopic overexpression of E2F1, E2F2 and E2F3 has been found by microarray analysis to induce many genes critical for the G2/M transition, chromatin assembly, chromosome segregation and cytokinesis (Ishida et al. 2001; Müller et al.
2001; Young et al. 2003). Rb is thought to protect cells from genomic instability through its regulation of Mad2, a mitotic spindle checkpoint gene. It is postulated that inactivation of Rb contributes to aberrant Mad2 expression, thereby leading to aneuploidy in human cancers
(Hernando et al. 2004). Additional evidence for a protective role of Rb-E2Fs in the G2/M
24 checkpoint comes from studies utilizing siRNA to knock down human RB in the presence of
DNA damaging agents. RB depletion allowed cells to bypass the G2/M checkpoint and led to deregulation of ECT2, a Rho-interacting proto-oncogene required for cell division at telophase
(Tatsumoto et al. 1999; Eguchi et al. 2007). Concurrent knock down of both RB and ECT2 restored arrest during cytokinesis (Eguchi et al. 2007), suggesting that RB-mediated transcriptional regulation of ECT2 is required for proper G2/M arrest in human cells.
Activation of the p53 pathway by loss of E2Fs further links E2F to DNA damage response and checkpoint activation. Concomitant loss of E2f1, E2f2, and E2f3 causes down- regulation of E2F target genes and arrest in all stages of the cell cycle (Wu et al. 2001), but it also results in the upregulation of p53 transcriptional activity. p53 target genes such as p21, killer/dr5, cd95/fas, pidd, and noxa are strongly induced, pointing to the possibility that E2F1-3 act upstream of p53 to induce cell cycle arrest (Sharma et al. 2006; Timmers et al. 2007). Despite the rapid upregulation of p53 protein and activity, proliferation does not immediately stop, suggesting that activator E2Fs are not necessary for S-phase gene activation or cell cycle progression but that they are important for the p53 response. After a 4-day latency period, E2F targets begin to dramatically decrease, during which time cells begin to arrest during all phases of the cell cycle and presumably accrue damage to DNA. Importantly, deletion of p53 in E2f1-/-;
E2f2-/-; E2f3-/- fibroblasts restores expression of E2F target genes and allows cells to proliferate relatively normally. That loss of p53 can bypass the requirement for the activator subclass of
E2Fs suggests that the function of E2F is not simply to activate genes required for entry into
G1/S but also is to sense DNA damage and signal to downstream effectors of DNA repair. Just as p53-deficiency allows cells to progress through the cell cycle, loss of p21, a key target of p53, restores the ability of E2f1-/-; E2f2-/-; E2f3-/- fibroblasts to transition through G1/S (Sharma et
25 al. 2006). Interestingly, cells do not progress through mitosis and ultimately fail to proliferate.
This suggests that, in addition to p21, other p53 targets are important during the cellular response to loss of E2F. The upregulation of p53 in response to loss of E2F activators also provides evidence for a signaling mechanism between the p16INK4A/Rb/E2F and the p19ARF/p21CIP1/p53 pathways that is critical for p53-mediated cell cycle arrest, and likely DNA repair. Atypical E2F repressors also have been recently shown to have significant roles in DNA damage repair (Ref).
Whether or not the role of E2Fs in regulation of G2-M check points are shared with their interesting functions in variant cell cycles that lead to polyploidy is not clear yet, but seems highly probable.
1.8 Variant cell cycles and physiological polyploidy
Most of the metazoan cells utilize mitotic cycles for their cell division, limiting genomic
DNA duplication to once per cell cycle. While vast majority of these cells stay diploid (except being tetraploid in transition through S/G2/M), variations in genomic DNA content, has been reported frequently (Lee et al., 2009; Ullah et al., 2009a). These include developmentally programmed and environmental stress induced physiological polyploidy and gene amplification as well as pathological aberrations in cell cycle regulation leading to aneuploidy, mitotic failure or rereplication (Awad et al., 2000; Blow and Hodgson, 2002; Calvi et al., 1998; Calvi and
Spradling, 1999; Claycomb et al., 2004; Claycomb and Orr-Weaver, 2005; Sun et al., 2008;
Ullah et al., 2009a; Ullah et al., 2009b; Zhong et al., 2003). The mechanisms that lead to physiological polyploidy is of three basic forms: cell fusion (skeletal muscle and osteoclasts), developmentally programmed and stress induced endoreplication (trophoblast giant cells and megakaryocytes) and acytokinetic mitosis hepatocytes (Awad et al., 2000; Blow and Hodgson,
26
2002; Calvi et al., 1998; Calvi and Spradling, 1999; Claycomb et al., 2004; Claycomb and Orr-
Weaver, 2005; Storchova and Pellman, 2004; Sun et al., 2008; Ullah et al., 2009a; Ullah et al.,
2009b; Zhong et al., 2003).
Endoreplication is a physiological state resulting from multiple rounds of DNA replication without associated karyokinesis or cytokinesis and is usually a consequence of endocycle or endomitosis, endocycle by far being commonest in nature (Cookson et al., 2006;
Cross, 2000; Edgar and Orr-Weaver, 2001; Grafi and Larkins, 1995; Ivanov et al., 2003; Jakoby and Schnittger, 2004; Kondorosi and Kondorosi, 2004; Lazzerini Denchi et al., 2006; Lee et al.,
2009; Leiva-Neto et al., 2004; Lilly and Spradling, 1996; Lozano et al., 2006; Maines et al.,
2004; Meckert et al., 2005; Yeung and Meinke, 1993). Polyploidy of whole organisms is frequently found in plants, invertebrates and lower vertebrates but is rare in higher vertebrates, where polyploidy of a limited subset of cells is achieved in most developmental contexts by the endocycle or in pathological contexts as part of an adaptive response to stress and disease
(Storchova and Pellman, 2004). In mice, trophoblast giant cells (TGCs), hepatocytes, megakaryocytes and pericardiocytes become endopolyploid during development. Polyploidy achieved through developmentally programmed endocycles in mice is restricted to TGCs of the placenta and hepatocytes of the liver. The megakaryocyte also achieves polyploidy but by means of undergoing „abortive‟ mitosis or endomitosis (Ullah et al., 2009). Polyploidy induces under conditions of stress in mice, seen for example in cardiomyocytes and vascular smooth muscles in a hypertensive setting (Vliegen et al., 1995; Hixon et al., 2000) and in hepatocytes that have sustained oxidative damage (Gorla et al., 2001; Gupta 2000), has been hypothesized to occur through cell fusion. It has been proposed that polyploidy is utilized by different species to provide nutrients and proteins for developing egg and embryo, for rapid growth in energy
27 deficient conditions, to maintain cell shape and tissue architecture and to maintain tissue homeostasis after injury or during phases of stress; and the disruption of this is has been shown indirectly in few contexts as not combatable with life (Cookson et al., 2006; Edgar and Orr-
Weaver, 2001; Grafi and Larkins, 1995; Ivanov et al., 2003; Jakoby and Schnittger, 2004;
Kondorosi and Kondorosi, 2004; Lazzerini Denchi et al., 2006; Leiva-Neto et al., 2004; Lilly and
Spradling, 1996; Lozano et al., 2006; Maines et al., 2004; Meckert et al., 2005; Yeung and
Meinke, 1993).
TGCs of mouse placenta are unique in the usage of developmentally controlled endocycles to achieve endopolyploidy. Once differentiated from trophoectoderm, TGCs undergo switch from mitosis to endocycle and reach ploidy levels as high as 1000N in nearly 15 days, making this most suitable system to study mammalian endopolyploidy, especially endocycle. In rodents, TGCs are believed to use endocycle to support proper implantation and placentation.
Even though there is no evidence to suggest that these cells supply protein products to the growing embryo, it is generally thought that massive polyploidy in this context is to assist in generating the energy and materials required for aggressive invasion into maternal tissues (Cross,
2000). In support to this hypothesis, significant reduction of ploidy in these cells is associated, though not proven to be the cause, with embryonic lethality (Garcia-Higuera et al., 2008; Geng et al., 2003; Parisi et al., 2003). Existing data from Drosophila melanogaster and plants including
Arabidopsis thaliana supports the current consensus that endocycle is a variant of mitotic cell cycle, where regulation of normal mitotic DNA replication machinery is modified to permit repeated cycles of G1-S phases while bypassing G2-M phases and associated checkpoints. On the other hand, recent reports of normal DNA replication components like Orc1 being
28 dispensable for endocycle suggest that this could be far from reality (Asano, 2009; Park and
Asano, 2008).
The developmentally programmed polyploidization of hepatocytes begins at the time of weaning in rodents, when animals are approximately three weeks of age (Barbason et al., 1974;
Dallman et al., 1974). This process continues throughout normal growth and aging such that up to 90% of hepatocytes in adult mice may be polyploid (Gerhard et al., 1971; Saeter et al., 1988).
Hepatocytes achieve polyploidy through endocycles, which generate mononuclear cells with high DNA content. Interestingly, hepatocytes also undergo acytokinetic mitosis, in which S and
M phases are actually completed but cytosolic partioning does not occur (Ullah et al., 2009), resulting in the appearance of bi- or multinucleated cells. A mature differentiated hepatocyte could, therefore, have single or multiple polyploid nuclei. The molecular mechanisms leading to the inhibition of cytokinesis in acytokinetic mitosis are thought to be dependent on modifications of processes involved in actin cytoskeleton assembly and microtubule reorganization (Margall-
Ducos et al., 2007), but remain in general much less understood than mechanisms that control endocycles.
Megakaryocytes are platelet precursor cells which are characterized by expression of
CD41, CD61, CD42 (glycoprotein Ib) and glycoprotein V (Hodohara et al., 2000; Roth et al.,
1996). Platelets are formed by fragmentation of pseudopodial projections (proplatelets) from mature megakaryocyte cell membranes (Patel et al., 2005). This process involves massive reorganization of megakaryocyte cytoskeleton leading to generation of a new compartment complementing cytoplasmic components including organelles, granules and soluble macromolecules (Italiano et al., 1999; Kaushansky, 2008). It is believed that each mature megakaryocte will generate close to 3000 platelets and once the life span is over, the residual
29 nuclear material undergoes macrophage mediated phagocytosis (Radley and Haller, 1983;
Stenberg and Levin, 1989). Indirect evidence suggest that size of megakaryocytes could have an impact on the amount of platelets that it could generate, as seen in patients with GATA1,
Gylcoprotein Ib/IX complex and Wiskott Aldrich syndrome protein mutations (Geddis and
Kaushansky, 2004). Current consensus is that the number of platelets in circulation is dependent on total number of megakaryocytes and size of individual megakaryocytes (Hoffman, 1989). The increase in size of megakaryocytes is conclusively shown to be associated with an increase in ploidy and a reciprocal increase in mRNA and protein content (Hancock et al., 1993). The process of polyploidy starts with terminal differentiation of megakaryocytes and is associated with a process called endomitosis where nuclear DNA content and cytoplasmic volume increases in the absence of karyokinesis and cytokinesis leading to formation of multilobular nucleus
(Odell and Jackson, 1968).
1.9 Regulation of atypical cell cycles and role of E2F transcription factors
Though very little is known about how mammalian cells switch from mitosis to endocycle, studies from Drosophila have given elaborate information of the important regulators including the role of Notch signaling pathway in initiation of endocycle in both nurse and follicle cells. These experiments have not provided answer to question of how differentiation and endocycle are linked in this context, but have proven that disruption of notch signaling hampers the ability of these cells to terminally differentiate and endocycle (Deng et al., 2001; Lopez-
Schier and St Johnston, 2001). Overexpression of Delta activates notch-signaling pathway and leads to premature initiation of endocycles (Jordan et al., 2006). Hedgehog pathway, on the other hand, antagonizes Notch pathway and promotes a proliferative phenotype (Zhang and Kalderon,
30
2000). Notch signaling executes this through induction of genes involved in destruction of mitotic regulators (e.g.-fzr/Cdh1), and repression of genes involved in mitosis like dacapo
(dapp21/p27) and stringcdc25 (Cebolla et al., 1999; Deng et al., 2001; Garcia-Higuera et al.,
2008; Schaeffer et al., 2004; Shcherbata et al., 2004). In mammals, Notch signaling has been implicated in regulation of endocycle through studies involving downstream effectors like Mash-
2 and upstream components like F-box protein Fbw7 (Nakayama et al., 1997; Tetzlaff et al.,
2004). It is not known whether E2Fs play significant role in regulation of this switch process.
Mitotic DNA replication machinery has inbuilt control mechanisms for limiting DNA replication to once per cell cycle and thereby to prevent polyploidy and rereplication.
Prereplication complex (PreRC) assembly happens only in the absence of cyclin-dependent kinase (Cdk) activity, which normally occurs from mitotic anaphase to late G1-phase through effects of APCCdc20 complex and Wee1/Myt1 protein kinases (Pesin and Orr-Weaver, 2008).
This prerequisite forms the primary mechanism by which cells prevent DNA rereplication during a single S-phase (DePamphilis et al., 2006). Cdks phosphorylates Orc1, Cdc6 and Cdt1 and target them for ubiquitin mediated degradation by SCFSkp2 ubiquitin ligase. Moreover, nonphosphorylated Cdt1 is a substrate of CDRDdb1 ubiquitin ligase. The fact that these two ubiquitin ligases are present only during S and G2-phases provide a second mechanism by which eukaryotes prevent MCM helicase loading on to chromatin during S and G2 phases. A third mechanism in form of geminin-mediated inhibition of Cdt1 is active during S, G2 and early M phases of cell cycle (DePamphilis et al., 2006; Zhu and Depamphilis, 2009). Cdk specific inhibitor p27Kip1 mediated delay in G1 to S phase by inhibition of Cdk2/Cyclin E activity until preRC assembly is complete forms a fourth regulatory step (Lengronne and Schwob, 2002;
Tanaka and Diffley, 2002).
31
In endocycle, as with the normal mitotic cycle, the origin control mechanisms ensure that the origins are not rereplicated (Edgar and Orr-Weaver, 2001). Progression of endocycle is maintained through cyclic phases of activation and inactivation of Cdks, as in case of mitotic cycles, regulating preRC assembly. However, mitotic cyclins are downregulated once cells have committed to endocycle (Narbonne-Reveau et al., 2008; Zielke et al., 2008). Instead, cyclin E fluctuates in a cyclic manner to provide an intervening G-phase (Follette et al., 1998; Weiss et al., 1998). Cyclic fluctuations of Cyclin E during endocycle, in addition, dictate the oscillations in APCCdh1 activity through phosphorylation of Cdh1, which in turn regulates geminin levels in a reverse fashion thereby contributing further to regulation of preRC formation by constraining
Cdt1 (Zielke et al., 2008). Thus, cyclic changes in Cyclin E are essential and form the key driving force for maintenance of endocycle through its regulation of preRC formation and dissociation (Arias and Walter, 2007; Coverley et al., 2002; Geng et al., 2003). This is unlike mitotic cycles in embryonic stem cells, where constitutive expression of cyclin E is permissive for DNA replication (Jackson et al., 1995). Finally, through a kinase independent mechanism, cyclin E has been shown to stimulate the association of minichromosome maintenance (Mcm) proteins, which form the essential replicative DNA helicase, with chromosomes to facilitate
DNA synthesis. Consistent with this model; knockout of cyclin E1 and 2, though dispensable for mitotic cycles, causes disruption of endocycles leading to substantial decrease in ploidy levels in megakaryocytes and TGCs (Geng et al., 2003; Parisi et al., 2003).
Two members of Cip/Kip family, p21Cip1 and p57Kip2 are upregulated in mammalian cells undergoing endocycle. Through its inhibitory effects on Cdk1 activity at end of S-phase, p57Kip2 leads to accumulation of cells in G2-phase where reformation of preRC is promoted consequential to geminin degradation by APCCdh1 (Ullah et al., 2008; Ullah et al., 2009b). This
32 is exemplified by the fact that loss of p57 leads to failure of TGCs to endocycle, even when the differentiation as such is not affected, a role which is not compensated by p21 (Susaki et al.,
2009; Ullah et al., 2008). Since p57 can target both Cdk1 and Cdk2, maintenance of endocycle requires p57 to be regulated in a cyclic manner to be present only in G-phase. Cyclin E, not being a target of APCCdh1 does this job. Thus, the p21 and p57 levels oscillate, peaking in G- phase and disappearing in S-phase. The role of p21 in the context of endocycle seems to be limited to inhibition of apoptosis by suppression of Chk1, a role which can be carried over by p57 also (Ullah et al., 2008).
Thus the inbuilt mechanisms that prevent rereplication in mitotic cell cycle are functional in endocycle (DePamphilis et al., 2006; Edgar and Orr-Weaver, 2001; Lengronne and Schwob,
2002; Pesin and Orr-Weaver, 2008; Tanaka and Diffley, 2002; Zhu and Depamphilis, 2009), but at the same time the reentry of cells from early G2 phase back to G1/S phase is facilitated over completion of remaining phases of cell cycle by upregulation of p21Cip1 and p57Kip2 and cyclic oscillation of Cyclin E levels along with persistent downregulation of mitotic cyclins
(Arias and Walter, 2007; Coverley et al., 2002; Follette et al., 1998; Geng et al., 2003;
Narbonne-Reveau et al., 2008; Susaki et al., 2009; Ullah et al., 2008; Ullah et al., 2009b; Weiss et al., 1998; Zielke et al., 2008).
E2F-mediated transcriptional regulation of important cell cycle components like cyclins and the anaphase promoting complex (APC) inhibiting factor Emi1 closely orchestrate DNA replication and cell cycle progression through feedback mechanisms for maintenance of cell cycle periodicities (Boudolf et al., 2004; DeGregori and Johnson, 2006b; Dimova and Dyson,
2005; Duronio et al., 1998; Duronio and O'Farrell, 1995; Duronio et al., 1995; Margottin-Goguet et al., 2003; Polager and Ginsberg, 2008; van den Heuvel and Dyson, 2008). According to this
33 model, the cyclic changes in Cyclin E level necessary for endocycle are maintained predominantly by reciprocal changes in activator E2Fs, which also form negative feedback loop with Cyclin E (Duronio et al., 1998; Duronio et al., 1996; Duronio and O'Farrell, 1995; Duronio et al., 1995; Sauer et al., 1995; Zielke et al., 2008). In support of this hypothesis, DNA synthesis and endocycle progression is affected in E2f1 knockout Drosophila and Dp1 knockout mice
(Duronio et al., 1998; Duronio et al., 1996; Duronio and O'Farrell, 1995; Duronio et al., 1995;
Kohn et al., 2003; Royzman et al., 1997). In addition, dE2F1 is also involved in negative feedback regulation of Cyclin E, as Cyclin E can downregulate dE2F1 in this context (Duronio et al., 1995; Sauer et al., 1995). As expected from the fluctuation of the products regulated, dE2F1 levels also fluctuate cyclically peaking in G phase (Zielke et al., 2008). This model suggest that dE2F1 accumulation during G1 leads to Cyclin E transcription and entry of cells into S phase, after which dE2F1 is down regulated transcriptionally by Cyclin E and actively degraded by
Cul4Cdt1 E3 ubiquitin ligase (mediated by PCNA binding to dE2F1 PIP box region) (Arias and
Walter, 2006; Havens and Walter, 2009; Higa et al., 2006; Hu and Xiong, 2006; Senga et al.,
2006; Shibutani et al., 2008). This model, though not proven in mammals, has been validated by mutational disruption of PIP box in dE2f1, where endocycles and not mitotic cycles are affected
(Lee et al., 2009; Shibutani et al., 2008). Even though current evidence suggests that E2Fs might play similar roles in mammalian endocycle also, it is not known how the regulatory mechanism involved are modified to permit endocycle (del Pozo et al., 2006; Duronio et al., 1998; Duronio et al., 1996; Duronio and O'Farrell, 1995; Duronio et al., 1995; Frolov et al., 2001; Kohn et al.,
2003; Mayhew et al., 2005; Park et al., 2005; Royzman et al., 1997).
Current evidence from different models predicts important roles of E2F repressors in regulation of mammalian endocycle. It is known that loss of DP1 (binding partner of E2F1-6)
34 leads to reduced ploidy in murine TGCs (Kohn et al., 2003). In addition, loss of retinoblastoma protein leads to increase in ploidy in plants, flies and mammals, though it is controversial whether this is purely through derepression of E2F activators (Du, 2000; Du and Dyson, 1999;
Du et al., 1996; Mayhew et al., 2005; Park et al., 2005; Royzman et al., 1997; Weng et al., 2003).
Increase in ploidy has been reported consequential to loss of E2FC/DPB in Arabidopsis and dE2F2 in Drosophila (del Pozo et al., 2006; Frolov et al., 2001). One of the atypical E2Fs in plants, E2FE/DEL1, has been shown to have inhibitory effect on endocycle onset by controlling expression of CCS52A2 (a Cdh1 homolog), a regulation that is likely to be preserved in other species including mammals also (Boudolf et al., 2009; Lammens et al., 2008b; Lammens et al.,
2009). It is still not clear whether E2F repressors play a similar significant role in regulation of endocycle in mammals.
As DNA content of cells is a strictly regulated process alterations in ploidy and associated accumulation of DNA damage are not usually tolerated by cells and result in activation of cell cycle and DNA damage checkpoints and apoptotic response; necessitating silencing of these pathways in endocycling cells thereby increasing sensitivity to genotoxic damage (Klapholz et al., 2009; Mehrotra et al., 2008; Storchova and Pellman, 2004; Ullah et al.,
2008; Ullah et al., 2009a). Once cells alter the cell cycle checkpoints to achieve polyploidy, the chances of accumulation of DNA damage becomes high with every round of genomic replication. This becomes relevant as polyploid cells can revert to their diploid status through a process of genome reduction cell division called depolyploidization. Polyploid cancer cells are known to depolyploidize with associated activation of meiosis specific genes though it is not clear whether this contributes in a causal way to clonal diversification and evolution of cancer
(Erenpreisa et al., 2005a; Erenpreisa et al., 2005b; Kalejs et al., 2006; Prieur-Carrillo et al., 2003;
35
Puig et al., 2008; Storchova and Pellman, 2004). The accumulation of DNA damage might complicate this by affecting the ability of cells to undergo genome reduction to become diploid, by centrosomal amplification or alteration of key cell cycle regulators. As an example, polyploid cells with supernumerary centrosomes in absence of p53, p21 or pocket proteins are known to undergo multipolar mitosis ending up with aneuploidy (Andreassen et al., 1996; Storchova and
Pellman, 2004).
Given the importance of endopolyploidy in physiological and pathological contexts across different species including mammals, it is essential for us to have a clearer understanding of the initiation, regulation and maintenance of the processes like endocycle leading to it. Due to lack of good in vivo systems, mechanisms involved in initiation and maintenance of mammalian endocycle are relatively uncharacterized yet. Based on the existing data related to endocycles from Drosophila melanogaster and plants including Arabidopsis thaliana the consensus is that endocycle is a variant of mitotic cell cycle, where regulation of normal mitotic DNA replication machinery is modified to permit repeated cycles of S-phase bypassing G2/M phases, associated checkpoints, karyokinesis and cytokinesis. However, recent data suggest that endocycle machinery components might have unique players quite different from its mitotic counterpart.
Plant endocycling cells preferentially express Orc1 and the prereplication complex (preRC) is more stable in these cells (Bermejo et al., 2002; Castellano et al., 2001; Diaz-Trivino et al.,
2005). At the same time another recent report suggest that, in Drosophila, Orc1 is dispensable for endocycle, but not for mitosis (Asano, 2009; Park and Asano, 2008). These studies also point to fundamental differences in endocycle regulation among different species.
Endomitosis in megakaryocytes and the transcription factors involved in commitment of megakaryocytic progenitors to megakaryocytes is reasonably well characterized. For example,
36 megakaryocyte specific elimination of GATA1 leads to severe thrombocytopenia due to dysmegakaryopoiesis (Shivdasani et al., 1997). Other studies have shown that Ets transcription factor Fli-1 is essential for megakaryopoeisis and mutations in this transcription factor has been known to be associated with congenital thrombocytopenia in humans (Doubeikovski et al., 1997;
Hart et al., 2000). Identification and cloning of c-Mpl and Tpo (ligand for c-Mpl) has advanced the field of megakaryocyte biology significantly by exposing the biological pathways involved in differentiation of these cells (de Sauvage et al., 1994; Kuter et al., 1994; Lok et al., 1994; Methia et al., 1993; Sohma et al., 1994; Solar et al., 1998; Wendling et al., 1994).
The process of polyploidy starts with terminal differentiation of megakaryocytes and is associated with a process called endomitosis where nuclear DNA content and cytoplasmic volume increases in the absence of karyokinesis and cytokinesis (Odell and Jackson, 1968). The process of endomitosis is characterized by multiple rounds of DNA synthesis with intervening
Gap phase, synthesis of Cyclin D3 isoform, decrease in CDK1 kinase activity, low levels of
Cyclin B1, centrosomal and centriolar amplification, DNA condensation and breakdown and reorganization of nuclear membrane (Datta et al., 1996; Garcia and Cales, 1996; Gu et al., 1993;
Handeli and Weintraub, 1992; Odell et al., 1968; Vitrat et al., 1998; Vitrat et al., 1996; Zhang et al., 1996). Together these observations suggest that endomitosis as a variant or abortive cell cycle where the cells go through successive rounds of G1, S, G2 and part of M (till anaphase A) leading to cells with ploidy content reported as high as 128N (Nagata et al., 1997; Ravid et al.,
2002; Roy et al., 2001). Several studies have shown that continually proliferating megakaryocytes (induced by overexpression of T antigen c-myc or E2F1), even when expressing differentiation markers, can have low ploidy presumably by failure to switch from mitosis to endomitosis (Guy et al., 1996; Kubota et al., 1996; Ravid et al., 1993). At what point does the
37 megakaryocytes get committed to endomitosis is still not clear, though current evidence from analysis of ploidy of these cells in different species suggests that majority of the are in 16N, while a small proportion goes further to achieve ploidy levels as high as 128N (Jackson et al.,
1997). A switch from cyclin D1 to cyclin D3 happens, but is of unknown significance as transgenic models overexpressing either of it leads to increase in ploidy (Sun et al., 2001;
Zimmet et al., 1997). In addition, augmented levels of cyclin A and cyclin E is also noted in these cells (Datta et al., 1998; Garcia et al., 2000). The levels of cyclin B1 is low as compared to cyclin A and has been demonstrated to be due to increased destruction by Anaphase Promoting
Complex resulting in high Cyclin B1/DNA ratio in megakaryocytes of high ploidy (Vitrat et al.,
1998; Zhang et al., 1998). One of the key observations as to how these cells go through abortive mitosis is through the observation that AIM-1 (one of the Aurora kinases which is believed to be essential for progression through mitosis) is transcriptional downregulated in both human and murine megakaryocytes and also in response to TPO, though the cause and effect of this is still not proven (Kawasaki et al., 2001; Zhang et al., 2001).
In summary, the experiments discussed in this study are aimed at addressing the lacunae in knowledge on E2F biology discussed above. We have tried to address two important questions here. First, we have tried to dissect out the mechanisms underlying the regulation of E2F- responsive genes and to provide in vivo evidence to support the existence of a unique cell phase dependent function of atypical repressors in modulating response of these genes to activation by canonical E2F activators. Secondly, we have attempted to identify role of canonical E2F activators and atypical repressors in regulation of polyploidy and address whether polyploidy in mammals is of any physiological significance as compared plants and flies.
38
Figure 1.1 The canonical Cdk-Rb-E2F pathway: Hyperphosphorylation of RB and related pocket proteins by Cyclin/Cdk activity causes release of RB from E2F activators, thereby removing transcriptional repression of E2F-responsive genes. GF; growth factors, CoA; coactivators.
39
Figure 1.2 The mammalian E2F gene family: E2F3 isoforms that are transcribed from distinct promoters, while E2F7 isoforms are produced by alternative splicing of the primary transcript
(Di Stefano et al., 2003). DBD, DNA binding domain; LZ, leucine zipper; MB, marked box; RB, pocket binding; NLS, nuclear localization sequence; CycA, cyclin A binding; NES, nuclear export sequence; DP, dimerization partner protein.
40
Figure 1.3 Regulation of E2F-responsive genes: In quiescent cells, the ubiquitously expressed
E2F4/E2F5 associate with pocket proteins and other co-repressors to repress genes that promote entry into the cell cycle. Upon mitogenic stimulation, the sequential phosphorylation of RB mediated by cyclin D-CDK and cyclin E-CDK complexes results in release of E2F4/E2F5 from
RB-E2F repressor complexes. The inactivation of RB also leads to the maximal accumulation of newly synthesized free E2F1-3 on target sequences late in G1. Together, these events initiate a transcriptional program that drives cells into S phase. The G1/S-specific transcriptome declines upon completion of S phase by degradation of E2F1-3 and by the action of repressors E2F6-8, independent of RB and RB-related proteins.
41
Figure 1.4 Chromatin modifications by E2F transcription factors: During active cell proliferation when RB is hyperphosphorylated, E2F activators recruit the basal transcription factor TFIID and other co-activators such as histone acetyltransferases (HATs) p300/CBP, GCN5 and Tip60 to target genes leading to an open chromatin configuration and transcription initiation. On the other hand, in quiescent or differentiated cells, pocket protein-bound E2F4/E2F5 have been found to associate with a variety of co-repressors such as histone methyltransferases (HMTs) and deacetylases (HDACs), DNA methyltransferase (DNMT1) and C-terminal binding protein
(CtBP), leading to chromatin compaction and transcription inhibition. The more recently identified repressors E2F6-8 mediate repression of E2F-responsive genes independent of binding to pocket proteins. While E2F6, through interaction with the Polycomb complex, function in transcription repression, co-repressors that associate with E2F7/E2F8 in vivo are currently unknown.
42
Chapter 2: Canonical E2F activators and atypical E2F repressors coordinate normal mitotic cell
cycle progression during placental development
2.1 Introduction
Upon external growth stimuli, the primary response in most cells is activation of signal cascades that ends up in initiation and orchestration of cell cycle resulting in two genetically identical daughter cells. One of the critical steps in this pro-proliferative pathway is the activation of G1-specific cyclin-dependent kinases (Cdks) which further leads to phosphorylation of retinoblastoma (Rb) and Rb-related pocket proteins and resultant the accumulation of E2F transcriptional activity (Frolov and Dyson, 2004). The availability of Rb unbound E2F activators and consequential transcription of E2F dependent genes late in G1 phase is generally considered to be the event in mitogenic signaling that commits cells to S phase entry. A subsequent wave of
E2F repressor activity, long believed to be through canonical E2F repressors, along with phosphorylation of DP is thought to coordinate turning down the expression of these genes towards the end of S-phase and helps completion of remaining phase-specific events and successful cell divisions (Krek et al., 1995). This classic paradigm of E2F-mediated activation and repression of responsive genes in the control of cell cycle progression is based almost exclusively on outcome of studies spanning two decades in invertebrates and in mammalian cell culture systems (using overexpression strategies) (Dimova and Dyson, 2005; Frolov et al., 2001).
However, the spectrum of tissue specific phenotypes observed in knock-out mice deficient for
43 various E2F family members is inconsistent with the rigid view of E2Fs as universal factors required to coordinate cell cycle dependent gene expression programs and instead suggest that different E2F family members either have different functions or they perform similar functions in a tissue specific manner (Chen et al., 2009; Chong et al., 2009b; Cloud et al., 2002; Danielian et al., 2008; Field et al., 1996; Humbert et al., 2000; Kinross et al., 2006; Li et al., 2003; Li et al.,
2008a; Lindeman et al., 1998; Murga et al., 2001; Pohlers et al., 2005; Rempel et al., 2000; Tsai et al., 2008; Yamasaki et al., 1996). It is also possible that the ablation of individual family members in these experiments were insufficient to expose how their combined activities might be coordinated in vivo.
The E2F family consists of nine closely related proteins (DeGregori and Johnson, 2006a) which are classified into transcription activators and repressors based on their sequence conservation and structure-function studies. Canonical E2F activators, E2F1, E2F2, E2F3a and
E2F3b, have transactivation domains and are known to associate with co-activator proteins and induce RNA polymerase II-dependent gene expression (Danielian et al., 2008; Trimarchi and
Lees, 2002). Recent evidence show that at least in some developmental contexts, canonical E2F activators can also function to repress gene expression, but the molecular basis for such plasticity is not completely understood (Chen et al., 2009; Chong et al., 2009b; Trikha et al., 2011; Wenzel et al., 2011). E2F repressors fall into two subclasses, with E2F4, E2F5 and E2F6 in one subclass
(canonical) and E2F7 and E2F8 in the other (atypical) based on their requirement for DP1/DP2 for DNA binding. Current in vitro evidence suggest that E2F4-6 mediated repression is responsive to Cdk signaling and involves the recruitment of histone deacetylases (HDACs), polycomb group proteins as well as Mga and Max to E2F-target promoters (Attwooll et al.,
2004), with E2F4/E2F5 having predominant functionality in G0 and E2F6 in S-G2 (Gaubatz et
44 al., 2000; Giangrande et al., 2004). The molecular basis of E2F7/E2F8 mediated repression is essentially unknown but appears to be independent of Cdk mediated phosphorylation of pocket proteins and dimerization with DP proteins.
Placenta had been historically used by mouse geneticists to understand in vivo function s of different genes. First organ to form in mammals, structural components of placenta are critical in transfer of nutrients, oxygen and waste products between fetus and mother. In addition, placenta is believed to be important in production of hormones necessary for maintenance of pregnancy and provides first defense against allogeneic rejection of fetus. Due to the importance of proper placentation and the high sensitivity of this tissue to genetic alterations, a large number of knock-out mouse models have shown defects in placentation and „believed to be‟ consequential mid-gestational lethality (Cross et al., 2003; Rossant and Cross, 2001). Previous work showed that embryos lacking E2f7 and E2f8 exhibit widespread apoptosis consequential to defects in placentation and die by E11.5 (Li et al., 2008a). Since very little is known about the in vivo functions of E2F7 and E2F8, we decided to utilize placental phenotype as a guide to improve our understanding of molecular functions of these two atypical E2F factors. Focusing on defects in the three main trophoblast derived lineages, labyrinthine trophoblast, spongiotrophoblasts and trophoblast giant cells (Rossant and Cross, 2001), and using mouse genetic, biochemical and bioinformatic approaches, we identified two antagonistic arms of the
E2F program, one driven by E2F7/E2F8 and the second by E2F3a, that coordinate the G1-S transcriptional output necessary for balancing cell proliferation and differentiation in the placenta. Ablation of the repressive E2F7/E2F8 arm resulted in ectopic proliferation of trophoblast progenitor cells and disrupted placental architecture and function, resulting in embryonic death by E11.5. Remarkably many of the phenotypes observed in E2f7-/-;E2f8-/-
45 embryos, including their early lethality, was ameliorated by the concomitant ablation of the
E2f3a activator. These findings provide first in vivo evidence to support E2F activator/repressor model of cell cycle regulation and suggest that canonical and atypical E2F pathways antagonistically coordinate the control of transcriptional programs essential for mammalian cell proliferation and development.
2.2 Results
2.2.1 Expression of E2f7 and E2f8 in murine placenta
Previous work from our lab using gene knockout approaches in mice has shown that
E2F7 and E2F8 are essential for embryonic development and loss of both leads to mid- gestational lethality (Li et al., 2008a). One of the commonly seen defect in knock-out mouse models with mid-gestational lethality is placental defects (Rossant and Cross, 2001). Placenta being a good system to study cellular processes and molecular pathways regulated by genes, we decided to analyze placentas from these embryos. Expression analysis demonstrated that E2f7 and E2f8 mRNA levels are relatively high in placental tissues, with peak expression at E10.5, surprisingly corresponding to the time of lethality of these embryos (Figure 2.1A). Interestingly, a second wave of placental E2f8 expression corresponding to expected time of the proliferation of glycogen trophoblast cells at E15.5 was also noticeable, though physiological significance of this observation is not known (Coan et al., 2006). In concordance with the observed mRNA levels, immunohistochemistry (IHC) on paraffin embedded placental sections of embryonic age
10.5 showed strong expression of E2F7 and E2F8 proteins in all the three major trophoblast lineages, labyrinth trophoblasts (LT), spongiotrophoblasts (ST) and trophoblast giant cells (TGC)
46
(Figure 2.1B, generated in collaboration with Jing Li), in comparison to germline deleted control samples. The observation that E2F7 and E2F8 proteins were very highly expressed in close to one-third of cells but undetectable in a large number of cells likely reflects their cell cycle dependent expression, predominantly in G2-M phases (de Bruin et al., 2003; Di Stefano et al.,
2003; Maiti et al., 2005b).
2.2.2 Loss of E2f7 and E2f8 causes differentiation and architectural defects in placenta culminating in mid-gestational embryonic lethality
Having observed the high expression of E2F7 and E2F8 in placentas, we went ahead and analyzed placentas from E2f7-/-;E2f8-/- embryos. Histological examination of hematoxylin and eosin stained double mutant placental sections revealed severely compromised tissue architecture
(small placenta phenotype) at embryonic age 10.5 (Figure 2.2, generated independently by Jing
Li). While wild type placentas have a well-organized labyrinth of trophoblast cells with the vasculature arranged as a network of maternal sinusoids juxtaposed to fetal-derived blood vessels, E2f7-/-;E2f8-/- placentas were overtly small and had abnormally large clusters of densely packed trophoblast cells that failed to effectively vascularize, pointing to defects in ability of these cells to invade into the maternal decidua. The vascular network in labyrinth was poorly formed and maternal sinusoids were rarely observable adjacent to fetal blood vessels suggesting major defects in placental function. It is interesting to note that these defects were comparable to that is observed in germline deletion of Rb (Wenzel, 2006), and very different from the placental phenotype of E2f1-E2f3 null embryos where the defect is limited to the junctional zone of placenta (spongiotrophoblasts and giant cells) (G Leone, unpublished data).
47
Given the fact that E2Fs play significant role in inducing terminal differentiation of cells by coordinating their cell cycle exit, one of the important questions was whether these defects are consequence of defects in differentiation of trophoblast lineages (Trimarchi and Lees, 2002;
Wang et al., 1998). To address this, we analyzed expression of various lineage markers using quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) and found that that double mutant placentas had normal expression of trophoblast stem (TS) cell-specific markers
(Eomes), while mRNA levels of spongiotrophoblast and giant cell specific markers (Tpbp and
Proliferin; Figure 2.3) were consistently reduced. We further validated these observations in tissue context using in situ hybridization (done through collaboration with John Thompson) for
Tpbp and Proliferin and immunohistochemistry for Esx1 (labyrinthine trophoblast specific) and
PL-1 (trophoblast giant cell specific) (Figure 2.4).
To identify whether the placental defects are the cause of observed midgestational lethality, we interbred E2f7loxp/loxp;E2f8loxp/loxp and Cyp19-cre mice (work done in collaboration with Jing Li), which express cre in all trophoblast cells, including in undifferentiated trophoblast progenitor cells as early as E6.5 (Wenzel and Leone, 2007). Strikingly, these intercrosses failed to yield any live Cyp19-cre;E2f7loxp/loxp;E2f8loxp/loxp embryos past E10.5. Gross and histological examination of live E10.5 Cyp19-cre;E2f7loxp/loxp;E2f8loxp/loxp embryos revealed a similar collapse of the labyrinth-like placental architecture along with vascular dilation, hemorrhage and growth retardation in associated fetuses (with intact E2f7 and E2f8) as observed in E2f7-/-;E2f8-/- embryos (Li et al., 2008a). This data suggests that the disruption of E2f7 and E2f8 in trophoblast progenitor cells and consequential placental defects are most likely the defining event causing mid-gestation lethality of E2f7-/-;E2f8-/- embryos.
48
2.2.3 Aberrant proliferation and cell death in trophoblast cells null for E2f7 and E2f8.
Results from germline deleted mouse models of E2F family members have shown that defects in terminal differentiation is usually associated with a failure to achieve cell cycle exit, pointing to similar defects in E2f7-/-;E2f8-/- trophoblast cells also (Trimarchi and Lees, 2002). We analyzed S-phase entry in these cells using BrdU uptake as a surrogate marker and found that uptake was elevated in double mutant spongiotrophoblasts and giant cells compared to their wild type controls (Figure 2.5). There was no discernible difference in BrdU uptake in labyrinthine trophoblasts probably because of the fact that this subset of cells is highly proliferative even in wild type samples. We then analyzed the number of cells in mitosis by immunohistochemistry using anti-phopho-histone3 antibodies and observed that mutant giant cells and spongiotrophoblasts have higher number of cells in mitosis (Figure 2.5). This increase in proliferation was in contradiction to the thinning of spongiotrophoblasts layer observed in figure
2.4. We hypothesized that this could be due to associated increase in cell death, a phenomenon that we had observed in double mutant fetuses (Li, 2008). To test this, we analyzed programmed cell death in this population. As expected, an increased number of apoptotic spongiotrophoblasts were observed in double mutant placentas (Figure 2.5). These evidences suggest that the architectural compromise and defects in differentiation observed in E2f7-/-;E2f8-/- might be due to aberrant proliferation and failure of these cells to exit cell cycle.
2.2.4 Derepression of S-G2-M genes underlie the cell cycle defects observed in double mutant trophoblasts
Given the well characterized known roles of E2F7 and E2F8 in transcriptional repression, we reasoned that changes in global gene expression profile of double mutant samples in
49 comparison to wild type littermate controls could give clues into the molecular mechanisms underlying the phenotype that we observed (Christensen et al., 2005; de Bruin et al., 2003; Di
Stefano et al., 2003; Li et al., 2008a; Maiti et al., 2005a). We focused on the genes that were found to be upregulated in global gene expression profile (Affymetrix Mouse Genome 430 2.0) of placentas lacking E2f7 and E2f8 (from germline deleted and Cyp19-Cre crosses, to avoid indirect effects due to feedback from affected fetal tissues). The venn diagram shown in figure
2.6 depicts the overlap of genes upregulated in E2f7-/-;E2f8-/- and Cyp19- cre;E2f7loxp/loxp;E2f8loxp/loxp placentas (p>0.05 and >2-fold). Using TFsearch engine, we analyzed the promoters of 49 genes derepressed in E2f7/E2f8-deficient placentas and found that only 16
(Table 2.1) contain consensus E2F binding sites that are conserved between mouse and human.
Close to 40% of these 16 potential targets encode proteins known to be involved in regulation of cell cycle, predominantly G1-S (Bioinformatic analysis done in collaboration with Thierry
Pécot).
We decided to validate whether the genes identified through expression analysis are direct targets of E2F7 and E2F8. Due to the lack of available ChIP-grade antibodies against endogenous mouse E2F7 and E2F8 proteins we overexpressed flag-tagged versions of E2F7 and
E2F8 in human embryonic kidney cells (HEK 293) and performed ChIP assays using anti-flag antibodies (done in collaboration with Jing Li). These assays showed that anti-flag, but not control IgG antibodies, could co-immunoprecipitate promoter sequences of most of the 16 genes with E2F binding elements, but could not co-immunoprecipitate irrelevant sequences lacking
E2F binding sites (downstream (ds) extronic sequences of E2f1 and Tubulin (Tub)) (Figure 2.7).
In addition to IgG, we also used mutant versions of E2F7 and E2F8 lacking DNA-binding capacity as a control to confirm the specificity of these assays. From these results we conclude
50 that many of the placental E2F target genes containing E2F binding sites in their promoters identified here by expression profiling represent true targets of E2F7 and E2F8. Whether the remaining targets which lacked consensus E2F binding sites may be directly regulated by
E2F7/E2F8, as it would appear for E2F1 (Cao et al., 2011; Rabinovich et al., 2008), remains to be determined.
2.2.5 E2F3a regulates E2F-responsive genes in antagonistic way and rescues molecular defects of E2f7/E2f8 null tissues
We then considered potential role of E2F activators in driving the expression of E2F- responsive genes in the absence of repression by E2F7 and E2F8. The observation that E2F target genes are upregulated in placental tissues lacking E2f7 and E2f8 suggested a simple mechanism involving transcriptional repression. However, the possibility that other E2Fs could be regulating the same target genes, particularly in the absence of E2F7 and E2F8 function remained. Indeed, recent gene knockout studies of E2F family members in mice have revealed incredible cooperation, redundancy and cross-regulation among E2F family members (Chen et al., 2009; Chong et al., 2009b; Li et al., 2008a; Trikha et al., 2011; Tsai et al., 2008). Previous work from our lab has shown that added ablation of E2f1 though could mitigate the apoptotic phenotype observed in E2f7/E2f8 null fetuses, but failed to rescue the lethality phenotype (Li et al., 2008a). Since we knew that the placental defects were underlying cause of embryonic lethality, we hypothesized that E2F1 may not be the key modulator of the molecular phenotype in placentas in absence of E2F7 and E2F8.
Given the established role for E2F3a in extra-embryonic tissues in context of loss of function of RB (Chong et al., 2009a), we decided to explore the functional relationship between
51
E2F7/E2F8 and E2F3a in placental development. Quantitative RT-PCR and Western blot assays showed that the levels of E2f3a mRNA and its protein product were not altered in E2f7-/-E2f8-/- placental tissues (unpublished observation, Jing Li), suggesting that E2f3a is unlikely to be a direct target of E2F7/E2F8-mediated repression. Moreover, ChIP assays demonstrated significant occupancy of E2F3a on E2F7/E2F8 target promoters, which was not affected by the absence of E2f7 and E2f8 (unpublished observation, Jing Li). Based on these results, we entertained the possibility that E2F3a and E2F7/E2F8 may act in parallel to regulate the expression of same target promoters, with the abundance of each of which according to cell cycle phase dictating the level of expression of the target genes. To confirm this, we analyzed gene expression profiles in E2f3a-/- animals. Remarkably, greater than ninety percent of genes that were differentially expressed in E2f7/E2f8-deficient tissues were also differentially expressed in
E2f3a-deficient tissues, but in the opposite direction (Figure 2.8, bioinformatic analysis done in collaboration with Thierry Pécot), suggesting that E2F3a and E2F7/E2F8 activate and repress, respectively, an overlapping set of targets.
To test the hypothesis that E2F3a and E2F7/E2F8 represent two antagonistic arms of the same E2F program, we interbred E2f7+/-;E2f8+/- and E2f3a-/- animals and analyzed global gene expression profile in triply mutant tissues. Surprisingly, expression profiling showed that a significant number of differentially expressed genes shared between E2f7/E2f8 doubly deficient and E2f3a deficient E10.5 placentas and fetuses, including most E2F target genes, were expressed to normal or near-normal levels in triple deficient embryos (Figure 2.8). These results are further demonstrated in the water fall plot (Figure 2.9). Added loss of E2f3a normalized the expression of 285 gene targets in placentas, with some targets being completely rescued (Class
I), others partially rescued (Class II) and a small subset not rescued at all (Class III). We verified
52 these findings using quantitative RT-PCR for representative genes from the complete or partial rescued group (Figure 2.9). These results suggest that the high levels of E2F target genes observed in E2f7/E2f8 doubly deficient cells is due to the unopposed E2F3a-mediated activation in the context of lack of repression by E2F7 and E2F8.
2.2.6 Added loss of E2f3a rescues E2f7/E2f8 null phenotype
Since the transcriptome of E2f7-/-;E2f8-/-;E2f3a-/- null tissues showed dramatic rescue over E2f7-/-;E2f8-/- and E2f3a-/- tissues, we were interested in seeing whether this molecular rescue will be translated into rescue of phenotype. As one would predict, E2f7-/-;E2f8-/-;E2f3a-/- placentas appeared normal with well-organized labyrinthine architecture and vasculature (Figure
2.10). Added loss of E2f3a rescued cell cycle defects including ectopic DNA replication that were noticed in double mutant placentas. These placentas were apparently functional since associated fetuses, though smaller in size, could be recovered alive even until birth (Figure 2.11, embryonic lethality data generated independently by Jing Li). Together, these findings suggest that E2F3a is a key modulator that antagonizes E2F7 and E2F8 mediated repression through activation of the same transcriptional program, which is critical for placental development.
2.2.7 E2F3a positively regulates expression of E2f7 and E2f8
E2f7 and E2f8 are known to have E2F binding sites and could be regulated by E2F1 both in vitro and in vivo (de Bruin et al., 2003; Maiti et al., 2005a). We tested whether this feed forward loop exists for E2F3a also. We analyzed E2f7 and E2f8 mRNA levels in E2f3a null placentas by qRT-PCR and found significant downregulation of these messages (Figure 2.12).
We could not validate whether this upregulation happens at the protein level also, due to lack of
53 specific antibodies against E2F7 and E2F8. These results suggest that E2F3a mediated accumulation of E2F7/E2F8 proteins later in S-phase coordinates the timely activation and repression of E2F targets during cell cycle.
2.3 Discussion
The atypical repressors E2F7 and E2F8 form the most evolutionary ancient arm of the
E2F family of transcription factors. The classic repressors E2F4-6, unlike E2F7 and E2F8, associate with pocket proteins and dimerization proteins (DP1/DP2) and are widely viewed as the major E2F repressive arm that drives cell cycle exit and differentiation, though in vivo evidence to support this is lacking. In this study we demonstrate that the atypical repressors play a critical role in balancing cumulative E2F activity in late S-phase by functionally antagonizing canonical E2F activators. The transcriptional program that is coordinated by these two opposing arms of the E2F family, the activating E2Fs and the two atypical E2Fs, is essential for repressing a broad spectrum of E2F targets that control trophoblast proliferation, placental development and embryonic viability. This study also highlight an unanticipated feed-forward loop between these two arms which ensures orderly cell cycle progression.
The placenta is one physiological context in mammals that has provided great insight into the functions of the Cdk-Rb-E2F pathway in mammals (Geng et al., 2003; Kohn et al., 2003;
Kohn et al., 2004; Kozar et al., 2004; Malumbres et al., 2004; Parisi et al., 2003; Wenzel et al.,
2007). For example, genetic ablation of p57KIP, p21CIP1, Cyclin E1/E2, Rb and DP1, has been shown to result in extra-embryonic defects that are attributable or demonstrated to contribute to the lethality of mutant embryos (Geng et al., 2003; Kohn et al., 2004; Wenzel et al., 2007).
Previous work from our lab demonstrated that the targeted disruption of E2f7 and E2f8 in mice
54 leads to lethality by embryonic age 11.5. Cconsistent with their high expression in placental tissues, we now show that double mutant embryos exhibit severe placental abnormalities. It would thus appear that two parallel E2F regulatory pathways, one dependent (E2F1-3) and the other independent (E2F7-8) on the canonical Cdk-Rb-E2F axis, are critically important in differentiation and cell cycle exit of major extra-embryonic trophoblast lineages.
Published work in flies has demonstrated antagonistic roles for dE2f1 and dE2f2 in the control of cell proliferation and larval development (Ambrus et al., 2007; Cayirlioglu et al.,
2003; Frolov et al., 2001; Frolov et al., 2003; Weng et al., 2003). While it was shown that in some cases dE2F1 and dE2F2 regulate the same target genes (including many cell cycle regulatory genes) through activation and repression, respectively, overall two fly E2Fs regulate distinct sets of targets whose functions converge, in ways not completely understood, on key developmental processes (Cayirlioglu et al., 2003; Dimova et al., 2003; Stevaux et al., 2005). It is generally viewed that mammalian activators (E2F1, E2F2 and E2F3) and classic repressors
(E2F4, E2F5 and E2F6) also work together in a similar manner to orchestrate the expression of
E2F target genes during the cell cycle. However, surprisingly little evidence in vivo exists to support the notion that these two subsets of mammalian E2Fs regulate the same set of target genes or even function in the same tissues. Mice deficient for each E2F activator or canonical repressor display distinct tissue specific phenotypes that are surprisingly devoid of major cell cycle alterations (Chen et al., 2009; Chong et al., 2009b; Cloud et al., 2002; Danielian et al.,
2008; Field et al., 1996; Humbert et al., 2000; Kinross et al., 2006; Li et al., 2003; Lindeman et al., 1998; Murga et al., 2001; Pohlers et al., 2005; Rempel et al., 2000; Yamasaki et al., 1996).
While redundant functions among family members may explain the absence of clear-cut cell cycle related phenotypes in these mice, it is most likely that the expected outcomes were overly
55 considered in the context of previous results obtained from cell culture systems. We show here that a single activator (E2F3a) and the two atypical repressors (E2F7/E2F8) are utilized in vivo to co-regulate cell proliferation. Using systematic analyses of global gene expression profiles and biochemical approaches, we demonstrate that E2F7 and E2F8 directly repress a subset of G1-S genes that control S-phase entry in trophoblast lineages. Surprisingly, microarray gene expression analysis of E2f3a deficient placentas revealed a significant decrease in most of these same target genes. This observation is remarkable given the divergent nature of their DNA binding domains, with E2F3a utilizing inter-molecular interactions with its partner protein DP
(Zheng et al., 1999) and E2F7/E2F8 utilizing both inter- and intra-molecular interactions between its two DNA binding domains (Di Stefano et al., 2003; Logan et al., 2005; Maiti et al.,
2005a). The fact that expression of target genes in E2f7, E2f8 and E2f3a triple deficient placentas was partially or completely restored to wild type levels, and even more surprising that these mutant embryos now survived to term, suggests that these three E2Fs coordinately regulate the expression of genes that are physiologically meaningful for embryonic development.
Studies performed in multiple cell lines have shown that E2F3a protein levels peak at
G1/S, followed by a precipitous drop in early-mid S phase (Ishida et al., 2001; Leone et al.,
1998). In contrast, levels of E2F7 and E2F8 begin to increase in mid-late S phase, peak in G2 and decline as cells approach M and the next G1 phase (de Bruin et al., 2003; Maiti et al., 2005a).
The cell cycle dependent accumulation of E2F7 and E2F8 proteins in cell lines is consistent with our IHC of placentas showing that ~30% of trophoblast nuclei stain strongly positive for these two proteins at any given time point. The sequential accumulation of E2F3a followed by
E2F7/E2F8 proteins during the cell cycle is likely no coincidence, since we show here that loss of E2f3a leads to a decrease in E2f7 and E2f8 expression. These observations support a
56 mechanism where the loading of E2F3a on target promoters at G1/S leads to their activation, including that of E2f7 and E2f8. As atypical E2F protein levels increase in mid-to-late S phase and E2F3a protein is targeted for degradation (Leone et al., 1998), G1/S targets become efficiently repressed by E2F7 and E2F8, leading to their decline by G2/M (Figure 2.13). Thus, the E2F3a-dependent activation of E2f7 and E2f8 may be viewed as a mechanism to ensure that rhythmic waves of E2F-dependent activation and repression drive cell cycle dependent gene expression. Whether E2F3a directly activates E2f7 and E2f8 promoters is not yet clear, but certainly is an attractive possibility.
We propose a new model for how E2F targets are regulated, which contrary to current belief does not generally involve classical E2F repressors, but rather involves repression by the most ancient and ironically termed „atypical‟ E2Fs. In this view, cell cycle dependent gene expression requires the balanced and timely interplay between Rb-regulated E2F activators and
Rb-independent atypical repressors.
2.4 Materials and methods
2.4.1 Mouse Strains and Genotyping
All protocols involving mice were approved by the Institutional Animal Care and Use
Committee at The Ohio State University. Transgenic mice used for this study were maintained in a mixed 129SvEv; C57BL/6; FVB background. Allele-specific (E2f7/8, Rosa26, PlfCre, and
Pl1Cre) and transgene specific primers (Sox2-cre and Cyp19-cre) were used for PCR genotyping
(Hayashi et al., 2002; Li et al., 2008a; Soriano, 1999; Wenzel and Leone, 2007) (Table2. 2).
57
2.4.2 Histology, Immunostaining and Quantification
Standard protocols were used for preparation of 5μm thick paraffin embedded sections of placentas and for hematoxylin and eosin staining. For immunohistochemistry, primary antibodies against E2F7 (ab56022, Abcam), E2F8 (custom-made polyclonal antibody raised against a peptide representing amino acids 576-595 of murine E2F8, Quality Controlled Biochemicals),
Esx1 (sc-133566, Santa Cruz), P-H3 (06-570, Millipore), BrdU (MO-0744, DAKO), TTF-1 (sc-
13040, Santa Cruz), CC10 (sc-25555, Santa Cruz) and PL-1 (a gift from Dr. F. Talamantes) were used. Pregnant mice at 10.5 days postcoitum were given intraperitoneal injections of BrdU (100
µg/grams of body weight) 30 min prior to harvesting. Detection of primary antibodies was done using species specific biotinylated secondary antibodies along with Vectastain Elite ABC reagent
(Vector labs) and DAB peroxidase substrate kit (Vector labs) or fluorophore (Alexa Fluor,
Invitrogen) conjugated secondary antibodies. Nuclear counterstaining was done using either hematoxylin or DAPI. Apoptotic cells were detected using TUNEL (S7101, Millipore) assays, performed according to the manufacturer‟s protocol. Images of immunostained sections were captured using Eclipse 50i (Nikon) and Axioskop 40 (Zeiss) microscopes and positive cells were quantified using Metamorph Imaging 6.1 software. Three sections per sample and at least three different samples for each genotype were analyzed. Data is reported as the average ± SD of percentage of positive cells.
2.4.3 Quantitative RT-PCR (qRT-PCR)
Total RNA was extracted using Qiagen RNA miniprep columns with in-column DNase treatment according to the manufacturer‟s protocol. Reverse transcription of total RNA was performed using Superscript III reverse transcriptase (Invitrogen) and RNAse Inhibitor (Roche)
58 as described by the manufacturer. Quantitative PCR was performed using SYBR Green reaction mix (BioRad) and the BioRad iCycler PCR machine. All reactions were performed in triplicate and relative amounts of cDNA were normalized to Gapdh. Data is reported as the average ± SD fold induction.
2.4.4 In situ Hybridization
In situ hybridization was performed on E9.5 (Proliferin) and E10.5 (Tpbp) placenta sections using a previously reported protocol (Christensen et al., 2002) modified
(deparaffinization in xylene and Proteinase K digestion) for paraffin-embedded sections.
Plasmids for Proliferin and Tpbp (gifts from Dr. J. Rossant) were linearized with HindIII and
XbaI, respectively, to generate templates for riboprobe synthesis. Hybridizations were performed with 1x107 dpm/slide of radiolabeled probes generated by in vitro transcription with either T7
(Proliferin) or T3 (Tpbp) RNA polymerase (Roche) using both 35S-CTP and 35S-UTP.
Autoradiography emulsion NTB (Kodak) was applied to the slides and the emulsion was exposed for 1 day (Proliferin) or 3 days (Tpbp) before being developed.
2.4.5 Affymetrix Microarray Analysis
Total RNA was isolated using Qiagen RNA miniprep columns according to the manufacturer‟s protocol. Global gene expression analyses were performed on Affymetrix Mouse
Genome 430 2.0 arrays at the Ohio State University Comprehensive Cancer Center
(www.osuccc.osu.edu/microarray/). Expression values were adjusted by quantile normalization and log2 transformation with RMAExpress and were analyzed using BRB-ArrayTools 3.7.0
(http://linus.nci.nih.gov/BRB-ArrayTools.html). Class comparison was used to select genes
59 differentially expressed at a significance level of p<0.05. Probes with a >2-fold misexpression in
E2f7-/-;E2f8-/- when compared to E2f7+/+;E2f8+/+ were used and the average relative expression level of each genetic group was used to generate heatmaps. Clustering and scatter plot analyses were performed by using functions of BRB Array Tools. Promoter sequences of each gene (-1000bp to +300bp relative to transcriptional start site) were obtained from UCSC genome browser (http://genome.ucsc.edu/). TFSearch (http://www.cbrc.jp/research/db/TFSEARCH.html) aided in the identification of genes containing E2F consensus binding sites (Heinemeyer et al.,
1998).
2.4.6 Chromatin Immunoprecipitation (ChIP)
The EZ CHIP assay kit (Upstate Biotech) was used as described by the manufacturer.
Primary antibodies used were anti-flag (M2, Sigma) and normal mouse IgG (Oncogene) antibodies [E2F7 and E2F8 ChIP] or anti-E2F3a (N-20, Santa Cruz), anti-E2F3 (C-18, Santa
Cruz) and normal rabbit IgG (Oncogene) [E2F3a ChIP]. Quantitive PCR was performed on de- crosslinked and column purified (Qiaquick, Qiagen) DNA fractions using the Bio-Rad iCycler system with primers specific for the indicated promoter regions. Reactions were performed in triplicate and normalized using the threshold cycle number for the 1% of total input sample. Data is reported as the average ± SD fold change.
2.4.7 Statistical Analysis
Pairwise comparisons of quantifications from histology and immunohistochemistry samples were evaluated by two-tailed Student‟s T-test. Statistical analysis of viability of embryos harvested from timed pregnancies was done by Fisher‟s exact test.
60
2.4.8 Laser Capture Microdissection
TGCs and STs were isolated from hematoxylin and eosin stained 8µm sections of OCT embedded frozen placental tissues using Laser Capture Microdissection (PALM MicroLaser system, Laser Capture Molecular Core at The Ohio State University Medical Center, https://lcm.osu.edu/) and DNA was extracted using QIAamp DNA micro Kit (Qiagen) according to manufacturer‟s protocol.
61
Figure 2.1 Expression of E2f7 and E2f8 in mouse placenta: (A) Quantitative RT-PCR analysis of
E2f7 (grey) and E2f8 (black) expression in wild type placentas at different stages of embryonic development. (B) E2F7 (top) and E2F8 (bottom) protein expression as identified by immunohistochemistry (IHC) in E10.5 placental sections with the indicated genotypes. Scale bars, 100 µm.
62
Figure 2.2 Severe architectural defects in E2f7/E2f8 null placenta: Hematoxylin and Eosin
(H&E) of E10.5 placental sections of indicated genotypes. Bottom panels are high magnification views of representative boxed areas in top panels. Scale bars, 100 µm. Yellow dotted line demarks junctional zone from decidua. De., Decidua; La., Labyrinth.
63
Figure 2.3 Differentiation defects in E2f7/E2f8 null placenta: Quantitative RT-PCR analysis for trophoblast lineage markers in E10.5 E2f7+/+E2f8+/+ (black) and E2f7-/-E2f8-/- (red) placentas.
Data shown as average ± SD of triplicate after normalization with Gapdh
64
Figure 2.4 Defects in differentiation of major trophoblast lineages in E2f7/E2f8 null placenta:
Qualitative analysis of differentiation markers in major trophoblast lineages in placentas. Top left panel, representative IHC analysis of labyrinthine trophoblast specific Esx1 (E10.5); Top right panel, immunofluorescent detection of trophoblast giant cell specific Placental Lactogen 1
(E10.5). Lower two panels, RNA in situ hybridization analysis of spongiotrophoblast specific
Tpbp on the left (E10.5) and giant cell specific Proliferin (E9.5) on the right. Yellow dotted line demarks junctional zone from decidua. De., Decidua; La., Labyrinth.
65
Figure 2.5 Cell cycle defects in E2f7/E2f8 null placenta: (A) Representative immunohistochemistry images for BrdU, TUNEL and P-H3 stainings (B) quantification of positive trophoblast cells in E10.5 E2f7+/+E2f8+/+ (black) and E2f7-/-E2f8-/- (red) placentas (** p<0.01). TGC, trophoblast giant cells; ST, spongiotrophoblasts; LT, labyrinthine trophoblasts.
66
Figure 2.6 Deregulation of cell cycle genes in E2f7/E2f8 null placenta: (A) Venn diagram depicting the overlap in genes significantly upregulated (>2-fold, p<0.05) in E2f7-/-E2f8-/- and
Cyp19 E2f7f/fE2f8f/f placentas (left panel). Schematic diagram of two types of representative promoters are depicted on the right (-1000bp to +300bp relative to transcriptional start site) of the 49 genes shared between two placental sets. 16 of 49 promoters have at least one E2F binding site conserved between mice and humans. (B) Gene ontology of the 49 overlapping genes in panel A.
67
Figure 2.7 Validation of deregulated targets for direct E2F7/E2F8 binding: Chromatin immunoprecipitation (ChIP) assays confirming promoter occupancy by E2F7 and E2F8 in a subset of the 16 potential E2F targets from E. HEK293 cells overexpressing either flag-tagged versions of wild type E2F7 and E2F8 (wt), or DNA binding mutant E2F7 and E2F8 (m) were used. Quantitative RT-PCR (normalized to 1% of input) was performed using primers specific to the E2F binding sites in target gene promoters as well as to irrelevant sequences in the tubulin promoter (Tub) and in the E2f1 downstream coding region (E2f1ds). Data shown as average ±
SD of triplicates.
68
Figure 2.8 E2F3a and E2F7/E2F8 regulates same subset of genes in antagonistic manner: Left panel, Heat maps of genes with >2-fold misexpression in E2f7-/-;E2f8-/- (7-/-8-/-), E2f3a-/- (3a-/-), or
E2f7-/-;E2f8-/-;E2f3a-/- (7-/-8-/-3a-/-) placentas relative to E2f7+/+;E2f8+/+ (7+/+8+/+) counterparts
(p<0.05). Right panel, Heat map of microarray probes that showed a >2-fold misexpression in
E2f7-/-;E2f8-/-E2f3a-/- fetuses relative to E2f7-/-;E2f8-/- and E2f7+/+;E2f8+/+ counterparts (p<0.05).
Results of the indicated groups are presented as an average of relative expression level. n, number of samples analyzed.
69
Figure 2.9 Added loss of E2f3a rescues defects in expression of genes in E2f7/E2f8 null placentas and fetuses: (A) Graphical representation of differences in fold-changes of genes between class comparisons E2f7-/-;E2f8-/-E2f3a-/- versus E2f7+/+;E2f8+/+ and E2f7-/-;E2f8-/- versus 70
Figure 2.9 continued..
E2f7+/+;E2f8+/+ placentas. Cut-offs used are as follows: complete rescue, more than 75% change in opposite direction; partial rescue, between 25-75% change in opposite direction and no rescue, less than 25% change in opposite direction or change in same direction. Genes with and without at least one conserved E2F binding site, between mice and humans, is represented numerically in bars. (B) Waterfall plot representing degree of gene expression rescue in E2f7-/-;E2f8-/-;E2f3a-/- placentas for 285 genes (>2-fold, p<0.05) that are deregulated in E2f7-/-;E2f8-/- placentas. Data shown as percentage change in average expression between E2f7-/-;E2f8-/-;E2f3a-/- and E2f7-/-
;E2f8-/- placentas relative to E2f7+/+;E2f8+/+ placentas. Three classes of targets are noted, with
Class I representing genes that are rescued by more than 75% relative to wild type levels, Class
II representing genes that are rescued by 25-75%, and Class III representing genes that are rescued by less than 25%. (C) Quantitative RT-PCR confirmation of expression changes in a subset of genes in D Values are normalized to Gapdh levels and wild type control sample with lowest expression was normalized to 1.
71
Figure 2.10 Added loss of E2f3a rescues cell cycle and structural defects in E2f7/E2f8 null placenta: (A) Quantification of BrdU-positive cells in trophoblast giant cells (TGC) and spongiotrophoblast cell (ST) lineages with the indicated genotypes. (* p<0.02, ** p<0.003). (B)
Representative low and high magnification H&E images of E10.5 placental tissues with the indicated genotypes. Scale bars, 100 μm. De., Decidua; La., Labyrinth. Yellow dotted line demarks junctional zone from decidua.
72
Figure 2.11 Added loss of E2f3a rescues mid-gestational lethality of E2f7/E2f8 null placenta: (A)
Alive 1 day old E2f7-/-;E2f8-/-;E2f3a-/- pup and littermate control. (B) Genotypic analysis of embryos derived from intercrosses of E2f7+/-;E2f8+/-;E2f3a+/- and/or E2f7+/-;E2f8-/-;E2f3a+/- mice at the indicated stages of development.
73
Figure 2.12 E2F3a induces expression of E2f7 and E2f8: Relative expression of E2f7 and E2f8 in
E2f3a+/+ and E2f3a-/- E10.5 whole placentas by quantitative RT-PCR assays. Values are normalized to Gapdh levels and wild type control sample with lowest expression was normalized to 1.
74
Figure 2.13 Current model of balanced E2F activity in regulation of E2F-responsive genes: In
G1-S, hyperphosphorylation of RB releases E2F activators to transactivate E2F target genes.
E2F1 and E2F3a at the same time induce expression of E2F7 and E2f8. In G2-M, E2F activator function decreases due to degradation of E2F activators, decreased affinity of E2F activators due to phosphorylation of DP1/DP2. At the same time, accumulated E2F7 and E2F8 actively repress
E2F targets.
75
Table 2.1 Potential direct targets of E2F7 and E2F8: based on presence of conserved canonical
E2F binding sites
Gene symbol Gene symbol Gene symbol Gene symbol
2810417H13Rik Uhrf1 (ICBP90) Mcm4 (Paf) Mcm6 Cdc6 Mcm5 Gins2 E2f1 2600005O03Rik (Dscc1) Mcm10 Hells Dtl (Cdt2) Chek1 Fbxl20 Zfp367 Phf19
76
Table 2.2 PCR genotyping primers used in the study
Expected genotyping Gene Primer sequences fragment size 5‟ GACCTTGCTGAGATTAGATC Cyp19-Cre Transgene: 545 bp 5‟ GAGAGAGAAGCATGTTTAGCTGGCC 5‟CACAGCTAAGCCTGGGTAGG Wild type: 186bp Plfcre/+ 5‟TTAATCAGTCTTCTTCATTCCTGA Knockin: 251bp 5‟CGGTTATTCAACTTGCACCA 5‟ AGCTCTCCTTGGGTTCAGGT Wild type: 177bp Pl1cre/+ 5‟ CATGGATGAATGGCAACGTA Knockin: 262bp 5‟ATCCCTGAACATGTCCATCAG 5‟ TCCAGTGACAGTCTTGATCCTTAAT Tpbp-Cre Transgene: 200bp 5‟ AAATTTTGGTGTACGGTCAGTAAAT 5‟ CTCCAGACCCCCGATTATTT Wild type: 318 bp E2f3a ko 5‟ TCCAGTGCACTACTCCCTCC Knockout: 237 bp 5‟ GCTAGCAGTGCCCTTTTGTC 5‟AGGCAGCACACTTGACACG Wild type: 300bp E2f7flox 5‟ACTTTTGGGACAGAGGTAGGA Knockout: 400bp E2f7ko 5‟CCAAGATGAAGGCCGAGATGCTAC Conditional: 340bp 5‟TAAAAAGCTTTGCGGTCGTT Wild type: 192bp E2f8flox 5‟AAGCCAACCTCGATGAATTG Knockout: 500bp E2f8ko 5‟CTCGCATCATCGTCTGCTAA Conditional: 230bp 5‟GCGAAGAGTTTGTCCTCAACC Wild type: 550bp RosaLoxP/+ 5‟GGAGCGGGAGAAATGGATAT Transgene: 260bp 5‟AAAGTCGCTCTGAGTTGTTAT
77
Table 2.3: Quantitative Real-Time and ChIP PCR primers used in the study
Gene qRT-PCR primer sequence forward qRT-PCR primer sequence reverse Eomes GTGACAGAGACGGTGTGGAGG AGAGGAGGCCGTTGGTCTGTGG Tpbp CAGGTACTTGAGACATGACTC GGCAGAGATTTCTTAGACAATG Proliferin GTGCAATGAGGAATGGTCG CATTCTGAAGCATGGTGCTC Eme1 GGAAATCAGGAATGGCAAAT TAAAGCTGCAGATCCACCAG Ccnb1 ATGGTGCATTTTGCTCCTTC CTTTGTGAGGCCACAGTTCA Cdkn2a GAACTCGAGGAGAGCCATCT GGGTACGACCGAAAGAGTTC Cdc6 AGTTCTGTGCCCGCAAAGTG AGCAGCAAAGAGCAAACCAGG Mcm10 GCAAAGTCCAAGCACACAGA GTCGTGCCAGTGGAGGTTAT E2f7 GCCAAGCAGGAAACAGAAGA ACCGTGCCAACCATACTGAT E2f8 GAGAAATCCCAGCCGAGTC CATAAATCCGCCGACGTT Espl1 AGGTAACTGCAGGCGTTTCT GAGAGGAGGTTGGACGAGAG Gapdh TTTGATGTTAGTGGGGTCTCGC CGGTGTGAACGGATTTGGC E2f3a GCCTCTACACCACGCCACAAG TCGCCCAGTTCCAGCCTTC Gene ChIP primer sequence forward ChIP primer sequence reverse Mcm4 CCACCACCTCCCGTCCTTAA AATCACAGCGGCGCTCGTAC Mcm5 ATTGTTCCGCACACAAAATG CAAACCAGGGGAGACAAGAA Ung CGGGCCGCTTGGCGCCAAT GGGCGTGTCGCTTCCTGGC Zfp367 GAAATGTAACGCGGGAAAAG GCCCTCCGCTCTTTGTACTC E2f1ds CGCCCAGACGCCACTTCATC TTCATTCCCTCACTCATTCAACAA Tub ATGGAGGGATGAATGGTTATGC CTTTTTGGGTCTGGCTTCTTTCAC
78
Chapter 3: Generation and characterization of trophoblast giant cell specific cre mouse lines
3.1 Introduction
Trophoblast giant cells are large cells found in the junctional zone of placenta of developing mammals. They are the first cells to differentiate during embryogenesis and are derived early in development from the outermost layer of the blastocyst called the trophoectoderm (Rawn and Cross, 2008). The trophoectoderm layer later develops into the placenta proper. Early in development, the trophoectoderm can be divided into two cell layers: the mural cell layers which eventually form primary giant cells and the polar layer which forms the trophoblast lineage including, among other cell types, the secondary giant cells. Trophoblast giant cells are unique in the fact that upon differentiation they undergo switch from mitosis to endocycle and continue replicating genomic DNA content to reach ploidy levels as high as
1000N in nearly 15 days, which is the highest level of ploidy that one could find in mammalian cells (Cross, 2005; Lee et al., 2009). The endocycle is characterized by alternating DNA synthesis (S) and Gap (G) phases without an intervening mitosis or cytokinesis. The increase in genome ploidy is usually accompanied by a corresponding increase in nuclear size, with nuclear size as high as 100um reported in literature (Varmuza et al., 1988). The fact that TGCs achieve this feat in a short period of time makes it the most suitable cell type for studying mammalian endocycle.
79
The conserved utilization of endocycle in organisms ranging from the unicellular protists to eukaryotes as complex as metazoans suggests an evolutionary significance to this process.
Indeed, many biological advantages of endocycles have been proposed, including an augmented response to strenuous cellular metabolic demands (Lee et al., 2009); buffering against environmental, metabolic-derived or replication-associated DNA damage (Otto and Whitton,
2000); and dampened sensitivity to apoptotic stimuli such as telomeric attrition and increased p53-p21 checkpoint function (Lilly and Spradling, 1996; Lazzerini Denchi et al., 2006; Mehrotra et al., 2008). The trophoblast giant cells themselves also produce key hormones and other unique proteins essential to the development of the fetus and placenta including but not limited to placental lactogens, matrix metalloproteinases, and growth factors such as proliferin (Lilly M,
2005 Oncogene). The primary trophoblast giant cells as part of the mural cell layer of the trophoectoderm play a major role in the proper implantation and later development of the placenta and embryo; they are therefore presumed to be essential to its development. On the other hand, it is observed that the higher the organism is in the evolutionary hierarchy, the lesser is the incidence of polyploidy. This clearly brings up a question whether this physiological process is just an evolutionary vestige for higher organisms.
Germline ablation of a large subset of cell cycle related genes including DP1 and Cyclin
E have been observed to have defects in size and or ploidy of trophoblast giant cells (Rossant and Cross, 2001). These defects have been frequently considered as cause of mid-gestational lethality observed in these cases. Given the fact the differentiation and maintenance of giant cells are dependent on the communication between inner cell mass and trophoblast stem cells through
Fgf4 (Fgfr2 being receptor) and Nodal (ActR being receptor) and that there is cross talk between these cells and adjacent layers in placenta itself, it is not clear whether most of the observed
80 findings are cell autonomous (Regenstreif and Rossant, 1989; Sferruzzi-Perri et al., 2009). Also, in the absence of a cell specific deletion strategy one cannot attribute defects in these cells to lethality.
Most of the information available regarding development and maturation of trophoblast giant cells originate from in vitro studies using Rcho1 cell lines (rat choriocarcinoma cell line) or are byproducts of studies in flies and plants. Recent work from the Asano and Leone labs has identified key differences in the mechanisms that execute mitotic DNA replication and endoreduplication, generating a lot of enthusiasm into understanding the molecular mechanisms underlying the regulation of endocycle (Ref). Studies in plants have shown that the inactivation of the newest repressive arm of the E2F family of transcription factors (DEL1 in Arabidopsis, considered to be homologues of E2F7 and E2F8 in mice) delays the timing of when cells switch from mitotic DNA replication to endoreduplication (Lammens et al., 2008b). It is not clear whether these mechanisms are conserved in mammals, though our results suggest a role of E2F7 and E2F8 in maintenance of endocycle in trophoblast giant cells (see chapter 2). Currently any attempt to utilize trophoblast giant cells as a model system to study mammalian endocycle is limited by lack of tools for manipulating the genes under question in cell specific manner.
Previous experiments have shown that the expression of Proliferin and Placental
Lactogen-1 is limited to the trophoblast giant cells within developing mice placenta, with no perceivable expression within the embryo proper (Rawn and Cross, 2008; Rinkenberger et al.,
1997; Yamaguchi et al., 1994). Both of these genes are encoded on chromosome 13and belong to prolactin gene family. These genes are well characterized and known to be expressed in trophoblast giant cell specific manner; with Proliferin being expressed in all subsets of trophoblast giant cells including secondary giant cells while expression of Placental Lactogen-1
81 is limited to parietal (primary) layer of trophoblast giant cells (Rawn and Cross, 2008; Simmons et al., 2007; Simmons et al., 2008). Proliferin subfamily consists of four members (PLF-I, PLF-2,
PLF-3 and PLF-4, with 91% coding similarity) while Placental Lactogen subfamily has four members (PL-Iα, PL-Iβ, PL-Iγ and PL-II, with 98% coding similarity) (Wiemers et al., 2003).
We were interested in developing genetic tools to help us in analyzing the cell cycle defects that were observed in E2f7/E2f8 null trophoblast giant cells and generated mouse lines cre recombinase knocked into Proliferin (Plf) and Placental Lactogen-1 (Pl1) loci downstream of the respective gene promoters. Though this approach would produce a germline ablation of corresponding genes, we expected this not to be of significant consequence as one would expect compensation from other family members. In this chapter we show evidence to demonstrate that cre recombinase, when driven by Plf and Pl1 promoters, is expressed in special and temporal pattern expected for the natural loci of these genes; making these knock-in mouse models valuable for studying endocycle and biology of trophoblast giant cells.
3.2 Results
3.2.1 Targeting Proliferin and Placental Lactogen-1 gene loci
To generate a genetic tool for studying endocycle in trophoblast giant cells, we utilized two genes, Proliferin (Plf) and Placental Lactogen 1 (Pl1), which are known to be expressed only in these cells (Simmons et al., 2007). RPCI-22 (129S6/SvEvTac) Mouse BAC Library (BPRC,
Children‟s Hospital Oakland Research Institute) was screened to identify BAC clones containing these two genes. Gene targeting construct was generated to include close to 4kb short arm and
8kb long arm using standard cloning and recombineering protocols (Figure 3.1) (Liu et al.,
82
2003). Frt sequence flanked Neomycin selection (NeoR) cassette was used to facilitate positive selection while thymidine kinase (TK) cassette was utilized for gancyclovir based negative selection. Linearized targeting vector was used for gene targeting using TC1 embryonic stem cells (Nationwide Children‟s Hospital Research Core Facility). Positive clones were identified by southern blot screening as shown in figure 3.1 and were injected into C57BL/6 blastocyst for generating chimeras. Chimeric mice were bred with Black Swiss wild type mice to acquire germline transmission (Figure 3.1, lower panels). These founders were further bred with Flpe transgenic mice (B6.Cg-Tg (ACTFLPe) FLPe: JAX Strain # 005703) to remove neomycin selection cassette and were back bred with Black Swiss wild type mice to remove Flpe transgene and to generate final experimental animals.
3.2.2 PlfCre and Pl1Cre faithfully reproduce predicted expression pattern of natural loci
To analyze the expression pattern of cre recombinase, we bred both PlfCre and Pl1Cre mice with Rosaβgeo26 (Rosa26) reporter mice to generate embryos heterozygous for reporter allele with or without Cre (Soriano, 1999). X-Gal staining was done in whole mount tissues and frozen sections to identify beta-galactosidase activity. As shown figure 3.2, we observed that cell specific expression of cre recombinase was noticeable as early as embryonic age 4.5 in both mouse lines. By embryonic age 10.5, majority of trophoblast giant cells were X-Gal positive, suggesting that active recombinase protein was expressed in these cells (Figure 3.3) cells were was limited to TGCs in both cases. Also, it was noticed that there was no noticeable X-Gal activity in other layers of placenta or associated fetuses. These results suggest that both PlfCre and
Pl1Cre knock-in mouse lines express cre recombinase in trophoblast giant cell specific manner.
83
3.2.3 Loss of Plf and Pl1 has little consequence on murine embryonic development
Since both PlfCre and Pl1Cre were generated by knocking in cre open reading frame into native Plf and Pl1 loci, we were worried whether an expected germline ablation of Plf and Pl1 would compound our ability to use these animal models. To analyze consequences of knock-in, we tried to breed these mice to homozygosity and observed that pups which are homozygous for knock-in allele were born with expected mendalian ratio (Figure 3.4). In addition, the litter size and longevity of these mice, even in homozygosity were comparable to wild type controls. In summary, these results suggest that germline deletion of Plf and Pl1 induced by knock-in has no significant consequence in fertility and longevity of PlfCre and Pl1Cre mice.
3.3 Discussion
Mid-gestational lethality and placental phenotypes are a commonly observed consequence of genetic studies involving germline ablation of genes (Rossant and Cross, 2001).
A large subset of these studies show defects in size or ploidy of trophoblast giant cells (Rossant and Cross, 2001). Though cause and effect is not proven, these defects are frequently considered as cause of lethality in these studies. This hypothesis is based on studies in plants and flies, where defects in endocycle have been shown to be detrimental for normal physiological function of the cells.
Trophoblast giant cells, the first cells to differentiate during embryogenesis, are considered unique owing to its ability to undergo mitosis to endocycle switch and replicate their genomic DNA content as high as 1000N in nearly 15 days (Cross, 2005; Lee et al., 2009). The functions of trophoblast giant cells like production of key hormones and proteins essential for successful placentation and development of the fetus and for preparation of mother for lactation
84 are considered biological advantages of having a polyploidy genome. Owing to the rate of polyploidization and the hypothesized critical functions of trophoblast giant cells, they are considered as the most suitable model system for studying mammalian endocycle. The lack of proper genetic tool for cell specific manipulation of genes in trophoblast giant cells has been the major deterrent in utilization of this system for studying mammalian endocycle.
We have generated two new knock-in mouse models, where cre recombinase is driven by trophoblast giant cell specific promoters Plf and Pl1. We demonstrate that both mouse models drive expression of cre in a trophoblast giant cell specific manner with recombinase activity noticeable as early as embryonic age 4.5. Moreover, these mouse models did not show any discernible consequences of knock-in induced ablation of native loci. In summary, we have generated new genetic tools that will help us in manipulating genes in trophoblast giant cell specific manner. These tools will be valuable as demonstrated in subsequent chapters of this manuscript for furthering our understanding of biology and functions of these cells and the process of endocycle.
3.4 Materials and methods
3.4.1 Mouse Strains and Genotyping
All protocols involving mice were approved by the Institutional Animal Care and Use
Committee at The Ohio State University. Transgenic mice used for this study were maintained in a mixed 129SvEv; C57BL/6; FVB background. Allele-specific (Rosa26, PlfCre, and Pl1Cre) were used for PCR genotyping (Hayashi et al., 2002; Li et al., 2008b; Soriano, 1999; Wenzel and
Leone, 2007) (Table2. 2).
85
3.4.2 Histology and X-Gal Staining
Whole-mount samples or 10µm frozen sections were fixed in 2% formaldehyde and 0.2% glutaraldehyde in PBS for 1.5 hours, and then washed 3 times for 5 minutes each in PBS at room temperature. Staining was carried out in 1mg/ml X-gal, 5mM potassium ferrocyanide, and 5mM potassium ferricyanide in lacZ wash buffer (0.01% deoxycholate, 0.02% NP-40 and 2mM
MgCl2) at 37°C overnight. Nuclear Fast Red counterstaining was used to visualize nuclei post- sectioning.
3.4.3 Southern Blot
Tail-genomic DNA was purified by standard methods and 3 g of phenol/chloroform extracted DNA was digested in a 25 L volume using EcoR1 (New England Biolabs) and separated on a 0.8% agarose gel. The DNA fragments were denatured in the gel for 30min in a
0.5N sodium hydroxide/1.5mM sodium chloride solution and then transferred overnight to a
Hybond-N+ membrane (Amersham) by capillary action using the same solution. The southern probe DNA was gel purified, extracted, quantified and labeled with P32 using the Rediprime system (Amersham). The probe was hybridized to the membrane overnight and washed 2 times for 15min with high-salt washing solution (250mM sodium phosphate/1mM EDTA 2% SDS) and then 2 times for 15min with low-salt washing solution (50mM sodium phosphate/1mM
EDTA 1% SDS). Hybridization bands were identified by exposing membranes to phosphoimager screens overnight.
86
Figure 3.1 Generation of PlfCre and Pl1Cre: (A-B) schematic diagram showing gene targeting and southern screening strategy. (C-D) Southern blot verification of germline transmitted PlfCre and
Pl1Cre mice
87
Figure 3.2 Trophoblast giant cell specific Cre recombinase activity in PlfCre and Pl1Cre mice is noticeable as early as E4.5: Representative images of X-Gal stained embryos showing Cre recombinase activity in Plfcre/+ and Pl1cre/+ mice heterozygous for Rosa26loxp reporter allele.
Right panel images for each genotype are higher magnification of corresponding left images.
88
Figure 3.3 TGC specific expression of cre recombinase in mature placentas: Representative images of X-Gal stained embryos and placentas showing specificity of Cre recombinase activity in PlfCre and Pl1Cre mice in mid-gestational (E10.5) placentas
89
Figure 3.4 PlfCre and Pl1Cre mice are viable and fertile even in homozygosity: Top panel shows representative images of PCR genotyping gels for indicated genotypes. Lower panel shows genetic tables.
90
Chapter 4: Regulation of endocycle in trophoblast giant cells by atypical E2F repressors
4.1 Introduction
Normal mitotic cell division cycle results in the generation of genetically identical diploid progenies. However, a number of variant cell cycles exist in nature. Endocycle is remarkable in this context because it is one of the most commonly seen variant cell cycle and is used by a wide range of organisms in diverse environmental contexts. Endocycle is characterized by alternating
DNA synthesis (S) and Gap (G) phases in the absence of intervening mitosis (M), karyokinesis and cytokinesis, though it is unclear to what extend the Gap phase is comparable to G1 and G2 of a mitotic cell cycle. The molecular machinery utilized in mitotic and variant cell cycles among species are presumed to be similar, though some recent evidence suggests that there could be significant differences between mitotic and endocycle DNA replication machinery.
The physiological advantages of a cell being polyploid as compared to being diploid are a much debated topic. The proposed advantage, based on studies in Arabidopsis thaliana and
Drosophila, include an augmented response to strenuous cellular metabolic demands by dedicating cell machineries to metabolic and anabolic processes, ability to produce higher copies of cell specific transcripts, dampened sensitivity to apoptotic and checkpoint stimuli, buffering against deleterious mutations that may cause cancer and advantages in evolution by depolyploidization. Endocycles in mammals are mainly restricted to trophoblast giant cells
(TGCs) in the placenta and hepatocytes in the liver. Out of the polyploid mammalian cells, TGCs
91 in the murine placenta are remarkable as they are highly polyploid, with copy numbers reported in literature as high as 1000 copies of the genome arranged in a polytene configuration (Varmuza et al., 1988). As with lower organisms, endocycles in mammalian tissues are believed to be important in maintaining normal organ physiology. On the other hand, another hypothesis is that the endocycles in mammals is a vestigial process from the evolution and has little role in the wellbeing of the animal. But surprisingly, the process is not well studied in mammals and so little evidence exists to support either of the hypotheses.
Mammalian atypical E2Fs have known homologues in Arabidopsis thaliana (DEL1-3 or
E2Fd-f) and in Caenorhabditis elegans (EFL-3; Winn et al., 2011), though so far none has been identified in Drosophila. Quite a lot of work has been done in Arabidopsis on the role of these
E2Fs in regulation of endocycle and the current evidence through overexpression and knock- down (RNAi) experiments support their role in the regulation of the timing of endocycle onset.
For example, plant cells overexpressing DEL1 do not increase their DNA content and instead continue mitotic cell division even past the expected developmental age when endocycle is supposed to be initiated. On the other hand, those lacking or with decreased expression of DEL1 end up with precocious and enhanced endoreplication (Vlieghe et al., 2005). Further studies have shown that the underlying mechanism involve direct repression of CCS52A (Cdh1 in mammals), which is required for activation of the APC/C complex to promote ubiquitin-mediated degradation of mitotic cyclins (Lammens et al., 2008). Finally in C. elegans, the newly identified
EFL-3, like E2F7 and E2F8 (Li et al., 2008), was reported to be involved in the control of apoptosis during worm development. It would be highly interesting to determine whether EFL3 would also function in the control of endocycles in the worm, which contain highly polyploid tissues such as the hypodermal syncytium (Lozano et al., 2006). Studies done in RchoI cell lines
92 show that downregulation of E2F activators are essential component of endocycle (Ref).
However, it is not known whether E2F repressors play any significant role in regulation of endocycle in mammals.
In this chapter, we give first in vivo evidence to support that the atypical E2Fs E2F7 and
E2F8 plays significant role in regulation of mammalian endocycle. Using novel genetic tools in mice we show that loss of atypical E2F repressors causes disruption of endocycle in trophoblast giant cells and hepatocytes and activated mitotic machinery. In addition, we demonstrate that this function is cell autonomous and that these functions appear to involve the repression of genes involved in mitosis, karyokinesis and cytokinesis. Surprisingly, placentas and livers with severely restricted polyploidization retained sufficient physiological function to carry them through apparently normal development, questioning the hypothesized functions of polyploidy in mammals.
4.2 Results
4.2.1. Expression of E2f7 and E2f8 in TGCs
Since it is known that downregulation of E2F activator activity is essential for mitosis to endocycle switch, we hypothesized that a corresponding upregulation of E2F repressors should be happening at the time of this switch. To evaluate this we examined mRNA levels of E2Ff7 and E2f8 using NanoString assays in TGCs of different age collected by laser capture microscopy (Figure 4.1). We observed consistent high expression of E2f7 and E2f8 in TGCs, suggesting a role of these proteins in the process of endocycle. Moreover, immunohistochemistry
93
(IHC) showed robust staining of E2F7 and E2F8 proteins in a portion of TGCs, consistent with their established cell cycle dependent expression in late S and G2 phases.
4.2.2 Loss of E2f7 and E2f8 leads to aberrant S-phase activity, activation of mitotic programs and induces karyokinesis in TGCs
Given their redundant functions during embryonic development, we decided to examine placentas in embryos deficient for both E2f7 and E2f8 at E10.5, one day prior to their embryonic lethality (Li et al, 2008). Visual inspection revealed a surprising number of E2f7-/-;E2f8-/- TGCs at various stages of mitosis (Figure 4.2A), implying an interruption of normal endocycles and switch in the endocycle program to that of mitosis. 3D reconstruction of serial stacks of images generated by confocal microscopy of E2f7-/-;E2f8-/- placentas showed that close to 40% of
TGCs contained at least two closely apposed nuclei (Figure 4.2D, done in collaboration with
Shantanu Singh), suggesting that these cells have undergone karyokinesis. These observations were confirmed by transmission electron microscopy (Figure 4.2C, done independently by Jing
Li). This is quite unusual for TGCs as these cells are terminally differentiated and have switched from mitosis to endocycle.
Having observed a large proportion of giant cells with binucleation, we went ahead and looked for molecular evidence for activation of mitotic program in these cells.
Immunofluorescence microscopy using antibodies against the TGC-specific marker placental lactogen-1 (PL-1), and the mitotic marker phospho-H3 (P-H3), identified numerous E2f7-/-
;E2f8-/- TGCs in metaphase and anaphase (Figure 4.2). In addition, the combined inactivation of
E2f7/E2f8 led to an increase in the levels of G2/M (cyclin A2) and M phase-specific (cyclin B1) cyclins (Figure 4.3). In addition, the combined inactivation of E2f7 and E2f8 also led to ectopic
94
BrdU incorporation in TGCs (Chapter 2). These results suggest that in the absence of E2F7 and
E2F8, trophoblast giant cells switch back from endocycle to mitosis. Whether or not these cells go through cytokinesis is not clear to us.
4.2.3 Endocycle defects in E2f7/E2f8 null TGCs leads to significant reduction in ploidy
With the observations that the endocycle is disrupted in these cells, one would anticipate a corresponding decrease in ploidy levels in these cells. We went ahead and analyzed the consequences of this on ploidy of TGCs. Using Feulgen analysis, we found that indeed, the ploidy of E2f7-/-;E2f8-/- TGCs (both nuclei were considered together in case of binucleated cells) was dramatically reduced and never exceeded 64n (Figure 4.4), whereas wild type TGCs with genomes >1000n could be readily detected. This is of relevance because the polyploidy of
TGCs is attributed to important functions in placental development and maturation.
4.2.4 E2F7 and E2F8 synergistically promote TGC endocycles
E2F7 and E2F8 function has been shown in the past to be able to function as either homodimers or heterodimers, giving them potential for high level of redundancy. In fact this level of redundancy in functions was demonstrated in vivo also (Li et al., 2008). We were interested in seeing whether their ability to support mammalian endocycles was also redundant.
We analyzed placentas lacking E2f7, E2f8, or one allele of each (E2f7+/-;E2f8+/-). These placentas exhibited TGCs with an intermediate reduction in ploidy and a proportionate increase in G2/M and M phase related events, indicating a gene dosage effect in their function (Figure
4.5)
95
4.2.5 Cell autonomous effects of loss of E2f7 and E2f8
Because the expression of multiple trophoblast cell lineage markers was altered in placentas globally deleted for E2f7 and E2f8 (TGCs, labyrinth trophoblasts and spongiotrophoblasts; chapter 1), we could not formally rule out a general disruption of differentiation as a possible cause for the endocycle defects observed in mutant TGCs.
Moreover, it is known that paracrine effects from other placental cells influence trophoblast giant cells. To analyze the extent to which the endocycle defects were cell autonomous, we used homologous recombination to develop a novel knockin mouse model where the cre open reading frame was inserted directly downstream of the endogenous start codon of the trophoblast giant cell specific Proliferin gene (Plfcre/+) (chapter 3). Faithful cre expression from this locus, as reported by the Rosa26LoxP allele (Plfcre/+;Rosa26LoxP/+), was confirmed by X-gal staining of placental and fetal tissues. This analysis showed expression of cre in TGCs as early as E4.5
(chapter 3), without collateral expression in other trophoblast cell lineages of the placenta or the embryo proper. After crossing Plfcre/+ mice with E2f7f/f, E2f8f/f or E2f7f/f;E2f8f/f mice and confirming the TGC-specific ablation of floxed alleles, as assessed by PCR genotyping of laser capture microdissected samples (Figure 4.6G), we evaluated giant cells for defects in endocycle once again. This analysis showed that all abnormalities observed in E2f7-/-;E2f8-/- trophoblast giant cells, including reduced ploidy, increased cyclin A2 and cyclin B1 expression and ectopic karyokinesis, were recapitulated in conditionally deleted E10.5 placentas (Figure 4.6).
4.2.6 Defects in endocycle in trophoblast giant cells have no physiological consequences in placentation and embryogenesis
96
One of the most interesting questions that we had in our mind, after having observed significant reduction of ploidy in trophoblast giant cells, was the physiological relevance of endocycle in trophoblast giant cells. It is widely believed that polyploidy in trophoblast giant cells is important for the implantation and proper placentation and one would expect embryos with this much significantly defective giant cells to be embryonically lethal. We used Plfcre/+ and
Pl1cre/+ knock-in mice to target cre-mediated recombination in TGCs. Ablation of E2f7 and E2f8 in either or both of two major lineages with cell cycle defects (spongiotrophoblasts and giant cells) resulted in live and phenotypically normal fetuses at every embryonic stage analyzed, including at birth in expected mendalian ratio (Figures 4.7A, lethality data and embryonic tissues for Cyp19-cre and Tpbp-cre generated by Jing Li). Placentas appeared well vascularized without any evidence of architectural disruption (Figure 4.7B). The specific ablation of E2f7 and E2f8 in
STs and TGCs in this context was confirmed by laser capture microdissection (LCM) of the appropriate extra-embryonic cell lineages and PCR genotyping (Figure 4.8).
4.2.7 Endocycle defects in E2f7/E2f8 null giant cells is independent of p57
Recent work from Ullah et. al., utilizing differentiated trophoblast stem cell cultures has led to a widely accepted model of endocycle regulation, where oscillations of p57/Kip2 dictate the cycling of alternating S and G phases (Ullah et al., 2008). We looked at the level of p57 at protein level using immunohistochemistry in E2f7/E2f8 null giant cells and noticed little difference between double mutant and their wild type littermate controls (Figure 4.9). This suggests the defect in endocycle that is observed in E2f7/E2f8 null trophoblast giant cells is driven by very different molecular mechanisms.
97
4.2.8 Deregulation of G2/M targets as the underlying mechanism behind endocycle defects
Given the role of E2F7 and E2F8 in transcriptional repression, we anticipated changes in gene expression as the principal underlying cause for the observed endocycle defects in
E2f7/E2f8 deficient giant cells. Transcriptome sequencing by RNASeq using RNA purified from trophoblast giant cells wild type and double mutant giant cells of embryonic age 10.5 showed that there were significantly more upregulated than downregulated genes in mutant samples, consistent with E2F7/E2F8 functioning as repressors (Figure 4.10). Supervised clustering methods were used to identify differentially expressed genes, deregulated in loss of E2f7and
E2f8 alone and in combination, as we could observe defects in endocycle in all three groups.
We focused our attention on the genes upregulated in all three mutant genetic groups relative to wild type controls. Closer inspection of these genes by Ingenuity Pathway Analysis
(IPA) revealed that most important molecular function affected was cell cycle related with a large proportion of genes with functions in cytokinesis, G2/M progression and various stages of mitosis, such as chromosome condensation, stabilization and segregation (Figure 4.10). By manual annotation, we also noted a bias for gene products with GTPase-activating (GAP) or
GTP exchange factor (GEF) functions, which have the potential to regulate the Rho family of small GTPases that are essential for successful cytokinesis (Normand et al., 2010). As expected, the vast majority of remaining genes involved metabolic pathways and signal transduction. One important point to highlight is the predominant downregulation of DNA damage response genes in knock-out groups. This observation corresponds well with the published role of E2F7 and
E2F8 in DNA damage response and provides a mechanism for the tumor suppressor functions of these proteins.
98
Expression changes of a large subset of these genes were confirmed by NanoString technology (Figure 4.11 and 4.12). For NanoString, we used a number of handpicked genes also based on the observations in liver (discussed in last part of this chapter) and also the known players in the endocycle.
Interestingly, a significant portion of common upregulated genes have E2F binding elements on their promoters, raising the possibility that these may represent direct targets of
E2F7/E2F8. To test this hypothesis, we transfected human embryonic kidney cells (HEK 293) with plasmids expressing flag-tagged E2F7 or E2F8 and performed chromatin immunoprecipitation (ChIP) assays with anti-flag antibodies (done in collaboration with Jing Li).
As shown in figure 4.13, flag-specific antibodies, but not control IgG, co-immunoprecipitated promoter sequences containing E2F binding elements (Ccna2, Chek1, Dtl and E2f1), but not irrelevant sequences lacking E2F binding sites (downstream exonic sequences of E2f1 (E2f1ds) and Tubulin (Tub)). The specificity of these ChIP assays was confirmed by parallel experiments with HEK 293 cells expressing mutant versions of E2F7 and E2F8 that lack DNA-binding capacity.
4.2.9 Similar molecular mechanism underlie regulation of endocycle and acytokinetic mitosis in hepatocytes
The developmentally programmed polyploidization of hepatocytes begins at the time of weaning in rodents, when animals are approximately three weeks of age (Barbason et al., 1974;
Dallman et al., 1974). This process continues throughout normal growth and aging such that up to 90% of hepatocytes in adult mice may be polyploid with ploidy levels of up to 64N reported in literature (Gerhard et al., 1971; Saeter et al., 1988). Interestingly, hepatocytes also undergo a
99 process called acytokinetic mitosis, in which S and M phases are actually completed but cytosolic partioning does not occur (Ullah et al., 2009), resulting in the appearance of bi- or multinucleated cells. The molecular mechanisms leading to the inhibition of cytokinesis are thought to be dependent on modifications of processes involved in actin cytoskeleton assembly and microtubule reorganization (Margall-Ducos et al., 2007), but remain in general much less well understood than mechanisms that control endocycles.
Having noticed defects in endocycle in trophoblast giant cells, we were curious to see whether the role of E2F7 and E2F8 in regulation of mammalian endocycle is in existence in hepatocytes also. To examine their roles during postnatal liver development, another graduate student (Hui-Zi Chen) in our lab crossed E2f7f/f and E2f8f/f mice with the established Albumin- cre mouse model (Alb-cre; Postic and Magnuson, 2000), which expresses cre in hepatocytes around the time of birth (Weisend et al., 2009). Though conditional inactivation of E2f7, E2f8 or both did not significantly alter hepatic mass at any of the ages analyzed (data not shown), evaluation of H&E stained liver sections revealed that the size of hepatocytes and their nuclei in aged Alb-cre;E2f8f/f (Alb-8ko) and Alb-cre;E2f7f/f;E2f8f/f (Alb-78dko) mice were significantly smaller than in control animals (Figure 4.14). These differences were compensated by a corresponding increase in the total number of Alb-8ko and Alb-78dko hepatocytes (data not shown). Moreover, loss of E2f8 or E2f7/E2f8 significantly decreased the proportion of bi- nucleated hepatocytes, suggesting defects in acytokinetic mitosis in these cells. Strikingly, flow cytometry showed that hepatocytes in Alb-8ko and Alb-78dko livers remained diploid over the entire lifetime of the mouse, whereas hepatocytes in Alb-cre;E2f7f/f (Alb-7ko) livers had only a modest reduction in genome ploidy compared to wild type controls. The percentage of Ki67-, cyclin A2, cyclin B1- and P-H3-positive hepatocytes was significantly elevated in 2-month-old
100
Alb-8ko mice and even more so in Alb-78dko mice (data not shown). Surprisingly, Alb-78dko mice containing livers with diploid hepatocytes survived to old age in apparent good health and retained the capacity to regenerate their livers upon chemical (carbon tetrachloride) or physical
(partial hepatectomy) injury (data not shown). From these results, we conclude that E2F8, and to a lesser extent E2F7, participate in promoting hepatocyte endocycles and acytokinetic mitosis.
We then used NanoString technology to compare the changes in gene expression of selected genes with reported G2/M related functions in TGCs isolated from E10.5 wild type,
E2f7-/-, E2f8-/- and E2f7-/-;E2f8-/- placentas. As shown in Figure 4.15, 66 of the 99 genes analyzed had predominantly increased expression in E2f7/E2f8 deficient TGCs, suggesting a common mechanism for how E2F7 and E2F8 may regulate endocycles in TGCs and hepatocytes.
4.3 Discussion
In multi-cellular organisms there exists diversity in the kinetics and composition of cell cycles and presumably, in the mechanisms that regulate them. The endocycle is commonly utilized in flies and plants and hence, more is known about how they are controlled in these organisms than in mammals. Here we exposed a novel role of atypical E2F repressors (E2F7-8) that controls mammalian endocycles. Loss of atypical repressors led to switch from endocycle to mitosis in TGCs and hepatocytes, with significant reductions in ploidy. We, in addition, show that aberrant activation of mitotic program as the underlying cause for these defects. The results of our ChIP assays suggest that E2F7 and E2F8 repress a core group of genes including Ccna2,
Dtl and E2f1 with prominent involvement in the regulation of G2/M related events and provide a molecular basis for the mechanisms underlying endocycle failure in E2f7/E2f8 deficient cells.
101
It has been assumed that polyploidy in mammals, as in flies and plants, is integral to the physiologic functions of placentas and livers. Here we generated the first mouse models with significantly reduced ploidy state to assess this hypothesis. Surprisingly, TGCs and hepatocytes having significantly restricted ploidy levels appeared to impart placentas and livers, respectively, with adequate organ function to carry animals through embryonic and adult development.
Whether altered liver ploidy in mice impacts aging remains to be evaluated. In summary, genetic, systems-based and biochemical approaches revealed a fundamental role for the atypical repressor arms of the E2F network in orchestrating mammalian endocycles in vivo.
4.4 Materials and methods
4.4.1 Mouse strains and genotyping.
All mice were housed in barrier conditions with a 12-hour light/dark cycle and had access to food and water ad libitum. The complete list of genotyping primers are provided in Table 4.1
4.4.2 Ploidy analysis using the Feulgen technique.
For quantification of genome ploidy, placentas and livers were fixed in 10% neutral buffered formalin and subsequently embedded in paraffin. 5μm thick sections were cut, deparaffinized in xylene, treated with 1N hydrochloric acid for eight minutes and stained with
Schiff reagent for one hour. Aqueous light green (1%) was used to counterstain the cytoplasm.
Mounted tissues were imaged using the Axioskop 40 microscope (Zeiss) and nuclear intensity was quantified using the software ImageJ (http://rsbweb.nih.gov/ij/) following standard protocols
(Hardie et al., 2002). TGC ploidy was expressed relative to diploid (2n)-tetraploid (4n)
102 labyrinthine trophoblast cells, and the average percent (± standard deviation) of TGCs in different bins, which represent different ploidy levels, was plotted as histograms. For bi- or multi-nucleated TGCs, the ploidy levels of individual nuclei were summed to show cumulative ploidy of cells. In contrast, hepatocyte ploidy is expressed on a per nucleus basis.
4.4.3 Ploidy analysis using flow cytometry.
Frozen liver tissue (~100mg) was finely minced followed by 5-10 minutes of incubation on ice in a small volume of hypotonic lysis buffer containing 25mM Tris (pH 7.5), 50mM KCl,
2mM MgCl2, 1mM EDTA and fresh 1mM phenylmethylsulfonyl fluoride. Nuclei were then released using a Dounce homogenizer with a tight pestle and washed twice with buffer A, once with sterile PBS, and resuspended in 0.1% PBS-Triton solution containing 50μl of propidium iodide (P.I., 0.5mg/ml, Roche) and 1μl of RNase A (20mg/ml, Invitrogen). A minimum of
30,000 nuclei per liver sample were analyzed for DNA content using LSR II (BD Biosciences), and cell cycle profiles were generated using FlowJo.
4.4.4 RNASeq global gene expression profiling analyses.
TGCs were isolated from hematoxylin and eosin stained 8µm sections of OCT embedded frozen placental tissues using Laser Capture Microdissection (PALM MicroLaser system, Laser
Capture Molecular Core at The Ohio State University Medical Center, https://lcm.osu.edu/) and
RNA was extracted using Arcturus PicoPure RNA Isolation Kit (Applied Biosystems- Cat No.
12204-01) according to manufacturer‟s protocol. 25ng of total RNA was amplified using
NuGEN‟s Ovation RNA-Seq kit according to the manufacturer‟s protocol. RNA/DNA chimeric primers are used in reverse transcription to produce first strand cDNA, and then heteroduplex
103
RNA/DNA double stranded cDNA. This was the template for isothermal linear amplification.
1ug of each amplified cDNA was sonicated with a Covaris S2 to approximately 150bp fragments. Applied Biosystem‟s SOLiD Fragment Library Construction Kit and Fragment
Library Barcoding Kit were used according to the manufacturer‟s instructions to create cDNA libraries as follows. The cDNA fragments were end-repaired; barcoded universal adapters were ligated to them; these were then PCR amplified. Size–selection gels were used after ligation and
PCR to select the appropriate sized products. The barcoded libraries were then pooled for emulsion-PCR and bead enrichment with the EZ Bead system. Sequencing by oligonucleotide ligation was performed on a SOLiD 4 with 50 and 35 base paired-end reads. For each sample, pre-processing of raw signal intensities was performed using the Robust Multi-chip Average
(RMA) method and consisted of background adjustment, quantile normalization and median polish. BRB-Array Tools 3.8.1 was subsequently used to generate class comparisons to determine differentially expressed genes at a significance level of p<0.05 (two-sample T-test with random variance model). For construction of heatmaps, the geometric mean of genes after
RMA normalization from control liver samples was set as reference, with the colors (blue or red) corresponding to ratio between expression of genes in mutant samples and the reference value.
Functional annotation of genes was performed using Ingenuity Pathway Analysis (Ingenuity
Systems, http://www.ingenuity.com).
4.4.5 NanoString gene expression analyses.
The NanoString nCounter system was used to quantitate gene expression in TGCs and validate gene expression changes (as determined through Affymetrix analysis) in the liver. 200ng of RNA was profiled on a custom codeset of the NanoString nCounter system according to
104 manufacturer‟s instructions. The codeset contained 200 genes that were selected in part based on their expression status in the liver, and based on their ability to regulate cellular functions involved in S phase, G2/M transition and M phase of the cell cycle. Additionally, ten housekeeping genes were also included in the codeset. For the analysis of gene expression in
TGCs, we isolated TGCs from 8µm H&E stained sections of OCT (Sakura) embedded frozen placental tissues using laser capture microdissection (PALM MicroLaser system). Total RNA was extracted using Arcturus PicoPure RNA Isolation Kit (12204-01, Applied Biosystems) according to manufacturer‟s protocol. For validation of gene expression changes in the liver, the same TRIzol®-isolated RNA was used for both Affymetrix and NanoString nCounter assays.
The nCounter data is first normalized to the geometric mean of positive control spike counts in each lane of the cartridge, followed by subtracting the minimum negative control spike counts in each lane to remove the background. Finally, expression of each gene in the codeset is normalized with respect to the geometric mean of the house keeping genes. For heatmap construction, the geometric mean of genes from control TGCs was set as reference, with the colors (blue or red) corresponding to fold-change in gene expression between control and mutant
TGCs (p<0.05).
4.4.6 Confocal microscopy 3D reconstruction of TGC nuclei.
Placenta samples were collected from E10.5 control and E2f7/E2f8 deficient embryos, fixed in 4% paraformaldehyde in PBS overnight at 4°C and subsequently embedded in OCT.
70µm-thick frozen sections were cut, washed with 0.2% Triton-X in PBS for 1.5 hours, and stained with 1mM Draq5 (Biostatus) at 4°C overnight. Draq5-stained samples were examined by confocal laser scanning microscopy (Zeiss LSM 510) with the use of a 63X objective and 0.7X scan zoom. Optical sections with between-plane-plane resolution of 0.40μm and axial resolution
105 of 0.42μm were acquired. Four to five non-overlapping regions per placenta were imaged with a field of view of 207μm x 207μm and depth of 40 2μm, resulting in a stack of 100 images for each region. This process was identically carried out for liver samples, except DAPI was used as the nuclear stain. For 3D reconstruction, the image stacks were processed using Insight Toolkit and ITK-SNAP. Each TGC or hepatocyte nucleus was segmented by active contour segmentation using region competition as the stopping criterion. Nuclei of binucleated TGCs were constructed separately and assigned different colors, and only the volumes of individual hepatocyte nuclei were measured.
4.4.7 Chromatin immunoprecipitation (ChIP) Assays.
For ChIP assays, the EZ-CHIPTM assay kit (Millipore) was used as described by the manufacturer. 70% confluent human embryonic kidney (HEK) 293 cells were transiently transfected using the calcium phosphate transfection method with 25μg of plasmids overexpressing wild type flag-E2F7 and flag-E2F8 or mutant flag-E2F7 and flag-E2F8 harboring mutations in their DNA binding domains. After 48 hours, cells were crosslinked and lysed, and the chromatin sonicated to obtain approximately 200-300bp fragments. Sonicated chromatin was incubated with 2μg of anti-flag (M2, Sigma) or normal mouse IgG (Oncogene) antibodies at 4°C with gentle rotation overnight. Antibody-protein-DNA complexes were recovered by addition of
30µl of Salmon Sperm DNA/Protein G agarose slurry (Millipore). Immunoprecipitated DNA fractions were de-crosslinked at 65°C and purified using Qiaquick columns (Qiagen).
Quantitative PCR of de-crosslinked immunopreciptated DNA was performed using SybrGreen
(BioRad) in the iCycler machine with primers specific for E2F binding sites. qPCRs reactions were performed in triplicate with 25μl per reaction with the following conditions: 1 cycle at 95ºC
106 for 5 minutes, followed 20-40 cycles of 95ºC for 15 s, 60-64ºC for 30 s and 72ºC for 30 s. Cycle numbers of genes tested were normalized to the threshold cycle number of 1% total input material.
4.4.8 Immunohistochemistry (IHC) and Immunofluorescence (IF).
5μm paraffin sections of placenta and/or liver tissues were incubated with primary antibodies against BrdU (MO-0744, DAKO), phospho-Histone 3 (Ser10) (06-570, Millipore),
CyclinA2 (sc-596, Santa Cruz), CyclinB1 (sc-752, Santa Cruz), p57Kip2 (sc-8298, Santa Cruz) and placental lactogen 1 (PL-1, a gift from F. Talamantes). TGCs in M phase of cell cycle (in
Figure 2H) were identified by co-immunofluorescence (co-IF) using anti-P-H3 or the lineage- specific anti-PL-1, with DAPI used as counterstain for nuclear DNA. Antigen retrieval was achieved by heat treatment (near boiling) in Target Retrieval Solution (DAKO), followed by blocking and primary antibody incubation at 4°C overnight. For IHC, the DAB (Vector) method was used.
4.4.9 BrdU incorporation assays.
Pregnant females at 10.5 days postcoitum (E10.5) received a single intraperitoneal injection of bromodeoxyuridine (BrdU, Sigma) dissolved in sterile saline at a concentration of
100µg/g body weight thirty minutes prior to sacrifice. Placentas were collected for fixation in
10% neutral buffered formalin, and 5µm paraffin embedded-sections were used for BrdU assays.
After deparaffinization, anti-BrdU (MO-0744, DAKO) antibodies were applied to tissues overnight 4ºC to detect BrdU incorporation. Prior to application of primary antibody overnight, tissues were digested in 2N HCl at 37ºC, neutralized with 10mM sodium borate (pH 8.0), and
107 blocked with 2% BSA. BrdU-stained slides were counterstained with either DAPI (if immunofluorescence) or Mayer‟s hematoxylin (if DAB immunohistochemistry).
4.4.10 IHC quantification and statistical analysis
Images of immunostained placenta and liver sections were captured using either the
Eclipse 50i (Nikon) or Axioskop 40 (Zeiss) microscope. For IHC/IF quantification, cells were counted using Metamorph Imaging 6.1 software photomerged (placenta) or 5-7 random (liver)
40x lens objective fields. The number of binucleated (“karyokinesis positive”) TGCs in control and mutant placenta samples was quantified from confocal images using the LSM image browser
(Zeiss). A minimum of 143 total TGCs were quantified in each genetic group. The number of bi- or multinucleated hepatocytes in control and mutant liver samples was quantified from H&E stained sections with the criteria that nuclei appeared physically in contact and cell membrane could be seen to enclose the nuclei. All results were reported as an average percentage ± SD of positive cells from the total cell population in number (n) of control and mutant samples analyzed. Pairwise comparisons were evaluated by two-tailed Student‟s T-test.
4.4.11 E2F binding site search
Promoter sequences (-3000bp ~ +2000bp) of genes antagonistically regulated by E2F1-3 and
E2F7/E2F8 were obtained from the UCSC genome browser (http://genome.ucsc.edu/), and E2F consensus binding sites were identified using TFSearch
(http://www.cbrc.jp/research/db/TFSEARCH.html).
4.4.12 Statistical Analysis
108
Quantification of IHC/IF was performed on multiple sections per tissue sample.
Percentages of positive cells are reported as average ± standard deviation. Flow cytometry of liver ploidy are representative of at least three independent experiments. One-way ANOVA was used to analyze data. When appropriate, p-values were adjusted by Holm‟s method within each ploidy category.
109
Figure 4.1 Expression of E2f7 and E2f8 in TGCs: (A) NanoString analysis of TGC-specific
E2f7/E2f8 expression in laser capture microdissected wild type TGCs of different embryonic ages. (B) Immunohistochemistry demonstrating E2F7 (top left) and E2F8 (bottom left) expression in wild type E10.5 TGCs but not mutant controls.
110
Figure 4.2 Loss of E2f7 and E2f8 leads to aberrant mitosis in TGCs (A) Representative H&E sections of E10.5 wild type (control) and E2f7-/-;E2f8-/- (dko) placentas. Inset, a dko TGC in metaphase (B) Co-immunofluorescence showing E10.5 dko TGCs in anaphase (left) and metaphase (right). DAPI stained total DNA. (C) Left, quantification of binucleated E10.5 TGCs.
Right, transmission electron micrograph of a dko E10.5 TGC (Right top, arrows indicate two nuclei; Right bottom, enlarged view of boxed area showing separation between nuclear envelopes). (D) Representative confocal images of nuclei in E10.5 control and dko TGCs (top) and 3D reconstruction of a binucleated dko TGC (bottom). Draq5 stained total DNA pseudocolored in green control, E2f7+/+;E2f8+/+; dko, E2f7-/-;E2f8-/- Data in (C) reported as average ± SD. T-test ** p≤0.01.
111
Figure 4.3 Loss of E2f7 and E2f8 causes reregulation of cell cycle genes: Immunostaining and quantification of S (cylin A2, panel A) and M (cylin B1, panel B) phase proteins in E10.5 control and dko TGCs. Quantification data is reported as average ± SD. T-test ** p≤0.01.
112
Figure 4.4 Loss of E2f7 and E2f8 leads to significant reduction of ploidy in TGCs: (A) Feulgen quantification of ploidy in E10.5 TGCs. (B) Representative images of Feulgen stained E10.5 placentas of indicated genotypes.
113
Figure 4.5 Synergistic functions of E2f7 and E2f8 in regulation of endocycle in TGCs: (A-E)
Quantification of cell cycle phase indicators in TGCs of indicated genotypes. T-test *** p<0.01.
(F) Feulgen quantification of ploidy in E10.5 TGCs. One-way ANOVA, * p≤0.05; ** p≤0.01;
***p≤0.001.
114
Figure 4.6 Cell autonomous functions of E2f7 and E2f8 in regulation of TGC endocycle: (A-E)
Quantification of cell cycle phase indicators in TGCs of indicated genotypes. T-test *** p<0.01.
(F) Feulgen quantification of ploidy in E10.5 TGCs. One-way ANOVA, * p≤0.05; ** p≤0.01;
***p≤0.001. (G) PCR genotyping of genomic DNA isolated from laser capture microdissected
E10.5 control and Plfcre/+ (Plfcre/+;E2f7f/f;E2f8f/f ) TGCs.
115
Figure 4.7 Disruption of endocycle in TGCs causes no significant physiological consequences:
(A) Genotypic analysis of embryos derived from intercrosses of Tpbp-cre, Plfcre, Pl1cre and
Cyp19-cre with E2f7loxp/loxp;E2f8loxp/loxp mice. For the Tpbp-cre, Pl1cre & Plfcre experiment, cre refers to presence of both cre alleles in heterozygosity. (B) Representative low and high magnification images of H&E stained E10.5 placental sections (top two rows) and gross appearance of associated fetuses (bottom) with the indicated genotypes. Arrows indicate dilated blood vessels and hemorrhagic areas. Histology scale bars, 100 µm; whole mount scale bars, 1 mm. De., Decidua; La., Labyrinth. Yellow dotted line demarks junctional zone from decidua.
116
Figure 4.8 Confirmation of cell specific deletion of E2f7 and E2f8: Representative PCR genotyping analyses of E2f7 and E2f8. Genomic DNA was isolated from E10.5 whole placentas
(Pla) or laser capture microdissected spongiotrophoblasts (ST) or giant cells (TGC) in Cyp19- cre;E2f7loxp/loxp;E2f8loxp/loxp (Cyp19-cre), Tpbp-cre;E2f7loxp/loxp;E2f8loxp/loxp (Tpbp-cre) and
Plfcre/+;E2f7loxp/loxp;E2f8loxp/loxp (Plfcre/+) respectively, along with cre negative controls (con) and whole fetuses (Fet).
117
Figure 4.9 Endocycle defects in E2f7 and E2f8 null TGCs are p57 independent: Representative images of immunohistocheimical staining for p57 protein levels in E10.5 placentas of indicated genotypes. Tissues with reported high levels of p57 are used as controls.
118
Figure 4.10 Deregulation of key cell cycle genes as molecular basis for endocycle defects in
E2f7/E2f8 null TGCs: (Left) Heatmap generated from RNASeq assay of transcriptome from laser capture microdissected E10.5 TGCs. control, E2f7+/+;E2f8+/+; 78dko, E2f7-/-;E2f8-/-; 8ko, E2f8-/-;
7ko, E2f7-/-. (Right) Gene ontology of genes upregulated and downregulated in all three indicated genetic groups
119
Figure 4.11 Validation of changes in transcriptome in E2f7/E2f8 null TGCs: Heatmap generated from a custom NanoString mRNA codeset in laser capture microdissected E10.5 TGCs showing validation of deregulated genes identified in figure 4.10. control, E2f7+/+;E2f8+/+; 78dko, E2f7-/-
;E2f8-/-; 8ko, E2f8-/-; 7ko, E2f7-/-.
120
Figure 4.12 Derepression of G2/M targets in E2f7/E2f8 null TGCs: Heatmap generated for
G2/M-related genes from custom NanoString mRNA codeset in laser capture microdissected
E10.5 TGCs. control, E2f7+/+;E2f8+/+; 78dko, E2f7-/-;E2f8-/-; 8ko, E2f8-/-; 7ko, E2f7-/-.
121
Figure 4.13 Direct regulation of G2/M targets by E2F7 and E2F8: Chromatin immunoprecipitation (ChIP) assays in HEK 293 cells demonstrating enhanced occupancy of flag-tagged E2F7 and E2F8 on E2F binding sites in the promoter of G2/M (CycA2, Cdt2 and
Chek1) and G1/S (E2f1) genes. ChIP controls E2f1ds (downstream exonic sequence) and Tub
(tubulin) showed pulldown specific to flag-E2F7 and E2F8. F, anti-flag; Ig, immunoglobulin control; w, flag-E2F7 or E2F8 with wild type DNA binding domains (DBD); m, flag-E2F7 or
E2F8 with mutant DBD.
122
Figure 4.14 E2f7 and E2f8 regulate hepatocyte endocycles. (A) Representative H&E sections of aged livers. (B) Flow cytometry of liver nuclei in aged livers. n, number of livers analyzed per genotype. Quantification on the left and representative FACS profiles on right. control,
E2f7f/f;E2f8f/f; Alb-7ko, Alb-cre;E2f7f/f; Alb-8ko, Alb-cre;E2f8f/f; Alb-78dko, Alb- cre;E2f7f/f;E2f8f/f. FACS, fluorescence activated cell sorting. Quantification data in B reported as average ± SD. (This figure was generated independently by Hui-Zi Chen.)
123
Figure 4.15 Similar molecular mechanisms underlie endocycle defects in E2f7/E2f8 null TGCs and hepatocytes: Heatmap generated from custom NanoString mRNA codeset showing expression of 66 genes which were found to be deregulated in E10.5 TGCs and 6 week old hepatocytes. control, E2f7+/+;E2f8+/+; 78dko, E2f7-/-;E2f8-/-; 8ko, E2f8-/-; 7ko, E2f7-/-.
124
Table 4.1 PCR Genotyping Primers for Chapters 4 to 8
Expected genotyping Gene Primer sequences fragment size
Alb-cre 5‟ATGCTTCTGTCCGTTTGCCG Transgene: 260bp Mx-cre 5‟CCTGTTTTGCACGTTCACCG
5‟CACAGCTAAGCCTGGGTAGG Wild type: 186bp Plfcre/+ 5‟TTAATCAGTCTTCTTCATTCCTGA Knockin: 251bp 5‟CGGTTATTCAACTTGCACCA
5‟AGCCACTGGATATGATTCTTGGAC Wild type:188bp E2f1 5‟AGAAGTCACGCTATGAAACCTCAC Knockout: 218bp 5‟AGTGCCAGCGGGGCTGCTAAAG
5‟GCCCCTAACACATGCACCCATTGG Wild type:210bp E2f2ko 5‟CCTGAGCGAGTCGGAGGATGG Knockout: 270bp 5‟ACCAAAGAACGGAGCCGGTTGGCG
5‟TGTGAATAATTTTTGGCATGTTTT Wild type: 152bp E2f3flox 5‟CTTATTCTGAGTGTGGACATACCG Knockout: 340bp E2f3ko 5‟AAGGGAAGGGAAAATTAAATCTGA Conditional: 202bp
5‟TCCCGTCTGCTGATCCGGACG Wild type: 360bp E2f4ko 5‟GCAAGCTATTTATTGTTAGTG Knockout: 250bp 5‟GCCTTCTATCGCCTTCTTGACG
5‟TGCTTTCTTTGTTCTCTTTGTC Wild type: 300 E2f5ko 5‟AGGGCCAAAGGTCTGTCC Knockout: 500 5‟GCCTTCTATCGCCTTCTTGACG
125
5‟AGGCAGCACACTTGACACG Wild type: 300bp E2f7flox 5‟ACTTTTGGGACAGAGGTAGGA Knockout: 400bp E2f7ko 5‟CCAAGATGAAGGCCGAGATGCTAC Conditional: 340bp
5‟TAAAAAGCTTTGCGGTCGTT Wild type: 192bp E2f8flox 5‟AAGCCAACCTCGATGAATTG Knockout: 500bp E2f8ko 5‟CTCGCATCATCGTCTGCTAA Conditional: 230bp
5‟CCTTTGGGCGGATTGTTGTTT Wild type: 680bp CycA1ko 5‟GGTCCTCCTGTACTGCTCAT Knockout: 442bp 5‟CGGTGCGCTTGCTAATC
5‟CGCAGCAGAAGCTCAAGACTCGAC Wild type: 392bp 5‟CACGAAGAATACTTGCTTTATGTC CycA2flox Knockout: 580bp 5‟GTCTTGTGGACCTCCACCAGACCT Conditional: 671bp 5‟CACTCACACACTTAGTGTCTCTGG
5‟GCGAAGAGTTTGTCCTCAACC Wild type: 550bp RosaLoxP 5‟GGAGCGGGAGAAATGGATAT Transgene: 260bp 5‟AAAGTCGCTCTGAGTTGTTAT
Table 4.1 PCR Genotyping Primers for Chapters 4 to 8 contd….
126
Table 4.2 ChIP PCR Primers
Gene ChIP primer sequences Tm
Ccna2 5‟ CTGCTCAGTTTCCTTTGGTTTAC 60ºC
5‟ AAAGACGCCCAGAGATGCAGC
Chek1 5‟ ATCTCCACGTCACCCTTTTG 60ºC
5‟ AGACCCCGAACTCCTTCTTT
Cdt2 5‟ GAAGGGGAAATGACTCTGA 57ºC
5‟ GAAATGCCTCCAAGTTCAGC
E2f1 5‟ CTGCCTGCAAAGTCCCGGCCACTT 64ºC
5‟ AGGAACCGCCGCCGTTGTTCCCGT
E2f1ds 5‟ CGCCCAGACGCCACTTCATC 60ºC
5‟ TTCATTCCCTCACTCATTCAACAA
Tub 5‟ ATGGAGGGATGAATGGTTATGC 59ºC
5‟ CTTTTTGGGTCTGGCTTCTTTCAC
127
Chapter 5: Restoration of G2/M block by ablation of Cyclin A1 and Cyclin A2 significantly
rescues defects in E2F7 and E2F8 null trophoblast giant cells
5.1 Introduction
Analysis of gene expression data from RNASeq studies in TGCs deficient for E2f7 and
E2f8 showed an enrichment of mitosis related genes (Chapter 4). One of the key cell cycle components that were found to be deregulated in these cells is cyclin A2. Studies in Arabidopsis have shown previously that transcriptional repression of all members of the CYCLINA2 family by ILP1 (polyploidy1) forms an important component in endoreplication control in plants, while the loss of CYCA2;1 has been shown to cause an ectopic increase in polyploidy (Yoshizumi et al., 2006). In addition, another A-type cyclin CYCA2;3 when coexpressed with the M phase- associated cyclin dependent kinase CDKB1;1 (equivalent of Cdk1 in mammals) inhibits endocycle initiation and leads to maintenance of mitotic divisions in plant cells (Boudolf et al.,
2009). Though, studies in plants have shown a critical role of cyclin A2 in endocycle control, a similar analysis in mice has not been feasible due to the early embryonic lethality of cyclin A2 null embryos.
The mammalian cyclin A family consists of two isoforms, cyclin A1 and cyclin A2 that are encoded by two distinct genes (Nieduszynski et al., 2002). Published work suggest that expression of murine cyclin A1 is limited to developing testes (Sweeney et al., 1996; Yang et al.,
1997) while cyclin A2 expression is ubiquitous (Pines and Hunter, 1990). Germline deletion of
128 cyclin A2 results in lethality of mouse embryos shortly after implantation by approximately E6.5
(Murphy et al., 1997), highlighting the role of this protein in cell cycle regulation. However, more recently, work from the Sicinski lab has demonstrated that though cyclin A2 function is required in hematopoietic cells and embryonic stem cells, mouse embryonic fibroblasts could survive without Cyclin A2 due to compensation by cyclin E (Kalaszczynska et al., 2009).
Cyclin A2 is induced at the G1/S transition and the activity of cyclin A2-Cdk2 which phosphorlylates Cdh1 is considered essential for cell cycle progression through S phase. One of the consequences this is inhibition of APC/C function during S and G2 phases and gradual accumulation of cyclin B1, which becomes relevant as cyclin A2 degraded rapidly once mitosis begins. On the other hand it is generally believed that cyclin B1 is not critically required for entry into mitosis as cyclin A2 could fulfill this function. Thus cyclin A2 is generally regarded as the essential S phase cyclin, a view that incidentally has now been challenged (Kalaszczynska et al., 2009), and in addition is also required to propel cells into M phase.
Since inactivation of E2f7/E2f8 results in aberrant activation of mitoses without inhibiting DNA replication, we hypothesized that re-establishing a mitotic block would restore polyploidy in E2f7/E2f8 deficient cells. This was further supported by the fact that cyclin A2 expression increased in E2f7/E2f8 deficient cells. Though Cyclin B 1 was also upregulated in double knock out cells, the reasons discussed in the previous paragraph compelled us to test the role of cyclin A instead of cyclin B. In this chapter, we show that quadruple deficiency in cyclin
A1, cyclin A2, E2f7 and E2f8 significantly rescued cell cycle defects and ploidy in TGCs.
Furthermore, we show that ablation of just cyclin A1 and A2 was of little consequence in TGCs
5.2 Results
129
5.2.1 Deregulation of Cyclin A2 in TGCs deficient for E2f7 and E2f8
We had observed the upregulation of cyclin A2 at protein level in trophoblast giant cells upon loss of E2f7 and E2f8. To identify whether this is due to transcriptional deregulation or due to changes in degradation, we decided to check the expression level of cyclin A2 in these cells.
Using NanoString, we observed that double mutant giant cells had significantly higher cyclin A2 mRNA levels compared to wild type litter mates (Figure 5.1). This is quite expected as cyclin A2 is a known E2F target.
5.2.2 Loss of Cyclin A1/A2 has no effect on TGC endocycle
Given the role of cyclin A2 in endocycle regulation in other organisms, we were interested in seeing whether loss of cyclin A will have any consequence in endocycle in trophoblast giant cells. To avoid possible compensation by cyclin A1, we intercrossed cyclin A1-
/-;cyclin A2f/f mice with Plfcre/+ mice to generate offspring with double deficient TGCs (Plf-
A1278dko). To our surprise, we observed that genetic ablation of cyclin A1 and cyclin A2 had no perceivable effect on size or ploidy of trophoblast giant cells (Figures 5.2 and 5.4). This observation was supported by the fact that there were no significant cell cycle defects in cyclin A deficient TGCs (Figure 5.3). It is possible that cyclin E might be compensating for any loss of function induced by ablation of cyclin A2.
5.2.3 Added loss of Cyclin A1/A2 rescues aberrant mitosis in E2f7/E2f8 deficient TGCs
To study consequences of loss of cyclin A1/A2 in E2f7/E2f8 null TGCs, we intercrossed
Ccna1-/-;Ccna2f/f mice38 with Plfcre/+;E2f7f/f;E2f8f/f and generated embryos with quadruply
130 deficient TGCs (Plf-A1278qko). We observed that added loss of cyclin A1 and cyclin A2 rescue in mitotic phenotype seen in E2f7/E2f8 null TGCs, but aberrant S-phase entry persisted in the quadruply deficient TGCs (Figure 5.3). It is interesting to note that the rescue of mitotic phenotype is in the presence of persistence of upregulated cyclin B1 levels. Together these results suggest that cyclin A2 and cyclin A1 are the major driving force underlying the endocycle to mitosis switch seen in loss of E2F7 and E2F8, though the extent to which cyclin B1 is important in this context is not known.
5.2.4 Polyploidy defects in E2f7/E2f8 deficient TGCs is significantly ameliorated by loss of
Cyclin A1/A2
We further evaluated whether the rescue in cell cycle defects seen in quadruply deleted
TGCs is translated into improvement in ploidy levels. As revealed by Feulgen analysis in figure
5.4, TGCs in E10.5 Plf-A1278qko placentas had larger nuclei with higher ploidy levels than
TGCs in Plf-78dko placentas, with some quadruple knockout cells having reached 1024N. These results support our hypothesis that the major cause of decrease in ploidy is TGCs deficient for
E2F7 and E2F8 is due to aberrant activation of mitotic program.
5.3 Discussion
Fundamental biology of TGCs could possibly explain the reasons why we did not see any consequence of cyclin A1/A2 ablation in TGCs. Germline deletion experiments to analyze the effects of loss of cyclin A family on TGCs is not feasible due to early embryonic lethality.
Plfcre/+ mediated ablation of cyclin A2, starts only after terminal differentiation of TGCs. By this time mitosis to endocycle switch has already set in and (Cross, 2005) and would naturally have
131 required the complete inhibition of cyclin A2 expression, thus rendering its genetic or „artificial‟ inactivation rather obsolete. The additional loss of cyclin A1/A2 from E2f7/E2f8 deficient TGCs restored their ploidy to near wild type levels but without completely correcting all aberrant molecular events, which also included the ectopic accumulation of cyclin B1 proteins. The persistent level of cyclin B1 in Plf-A1278qko TGCs might have been the result of compensatory upregulation due to absence of cyclin A2. Another possibility is that cyclin B1 proteins accumulated over multiple earlier division cycles due to deregulated APC/C function. The observation that many Plf-A1278qko TGCs stained strongly positive for cyclin B1 in the cytoplasm suggests a lack of Cdk1 activation in the nucleus, which would explain how Plf-
A1278qko TGCs did not exhibit decrease in ploidy despite high levels of B1. Interestingly, the
Kaldis laboratory recently showed that hepatocyte specific ablation of Cdk1 function, utilizing also the Alb-cre transgenic system, led to a hyperploidy phenotype in the liver reminiscent of that accomplished by cyclin A2 deletion (G. Leone and P. Kaldis, personal communication). On a broader level, these results suggest that mechanisms regulating endocycle, no matter how divergent they may appear, may ultimately converge on the control of the M phase associated
Cdk1 function.
In summary, interfering with the mitotic machinery (by Ccna1/Ccna2 inactivation) re- established a mitotic block and reinstated higher levels of polyploidy in E2f7/E2f8 deficient cells.
5.4 Materials and methods
5.4.1 Mouse strains and genotyping.
132
All mice were housed in barrier conditions with a 12-hour light/dark cycle and had access to food and water ad libitum. The complete list of genotyping primers are provided in Table 4.1
5.4.2 Ploidy analysis using the Feulgen technique.
For quantification of genome ploidy, placentas and livers were fixed in 10% neutral buffered formalin and subsequently embedded in paraffin. 5μm thick sections were cut, deparaffinized in xylene, treated with 1N hydrochloric acid for eight minutes and stained with
Schiff reagent for one hour. Aqueous light green (1%) was used to counterstain the cytoplasm.
Mounted tissues were imaged using the Axioskop 40 microscope (Zeiss) and nuclear intensity was quantified using the software ImageJ (http://rsbweb.nih.gov/ij/) following standard protocols
(Hardie et al., 2002). TGC ploidy was expressed relative to diploid (2n)-tetraploid (4n) labrynthine trophoblast cells, and the average percent (± standard deviation) of TGCs in different bins, which represent different ploidy levels, was plotted as histograms. For bi- or multi- nucleated TGCs, the ploidy levels of individual nuclei were summed to show cumulative ploidy of cells. In contrast, hepatocyte ploidy is expressed on a per nucleus basis.
5.4.3 Immunohistochemistry (IHC) and Immunofluorescence (IF).
5μm paraffin sections of placenta and/or liver tissues were incubated with primary antibodies against BrdU (MO-0744, DAKO), phospho-Histone 3 (Ser10) (06-570, Millipore),
CyclinA2 (sc-596, Santa Cruz) and CyclinB1 (sc-752, Santa Cruz). Antigen retrieval was achieved by heat treatment (near boiling) in Target Retrieval Solution (DAKO), followed by blocking and primary antibody incubation at 4°C overnight. For IHC, the DAB (Vector) method was used.
133
5.4.4 BrdU incorporation assays.
Pregnant females at 10.5 days postcoitum (E10.5) received a single intraperitoneal injection of bromodeoxyuridine (BrdU, Sigma) dissolved in sterile saline at a concentration of
100µg/g body weight thirty minutes prior to sacrifice. Placentas were collected for fixation in
10% neutral buffered formalin, and 5µm paraffin embedded-sections were used for BrdU assays.
After deparaffinization, anti-BrdU (MO-0744, DAKO) antibodies were applied to tissues overnight 4ºC to detect BrdU incorporation. Prior to application of primary antibody overnight, tissues were digested in 2N HCl at 37ºC, neutralized with 10mM sodium borate (pH 8.0), and blocked with 2% BSA. BrdU-stained slides were counterstained with either DAPI (if immunofluorescence) or Mayer‟s hematoxylin (if DAB immunohistochemistry).
5.4.5 IHC quantification and statistical analysis.
Images of immunostained placenta and liver sections were captured using either the
Eclipse 50i (Nikon) or Axioskop 40 (Zeiss) microscope. For IHC/IF quantification, cells were counted using Metamorph Imaging 6.1 software photomerged (placenta). The number of binucleated (“karyokinesis positive”) TGCs in control and mutant placenta samples was quantified from confocal images using the LSM image browser (Zeiss). All results were reported as an average percentage ± SD of positive cells from the total cell population in number (n) of control and mutant samples analyzed. Pairwise comparisons were evaluated by two-tailed
Student‟s T-test.
5.4.6 Statistical Analysis.
134
Quantification of IHC/IF was performed on multiple sections per tissue sample.
Percentages of positive cells are reported as average ± standard deviation. Flow cytometry of liver ploidy are representative of at least three independent experiments. One-way ANOVA was used to analyze data. When appropriate, p-values were adjusted by Holm‟s method within each ploidy category.
135
Figure 5.1 Transcriptional derepression of cyclin A2 in E2f7 and E2f8 null TGCs: (A)
NanoString expression analysis of cyclin A2 in RNA samples collected by laser capture microdissected E2f7/E2f8 null and wild type TGCs. Data shown as average of two samples each
± standard deviation after normalizing to housekeeping genes. (B) Representative images of immunohistocheimical staining for Cylin A2 protein levels in E10.5 placentas of indicated genotypes. control, E2f7f/f;E2f8f/f; 78dko, Plfcre/+;E2f7f/f;E2f8f/f.
136
Figure 5.2 Added loss of cyclin A1/A2 rescues phenotypic defects in E2f7/E2f8 null TGCs:
Representative hematoxylin and eosin stained E10.5 placental sections from indicated genotypes.
Original magnification 10X control, E2f7f/f;E2f8f/f; Plf-A12, Plfcre/+;cyclin A1-/-;cyclin A2f/f; Plf-
78dko, Plfcre/+;E2f7f/f;E2f8f/f; Plf-A1278qko, Plfcre/+;cyclin A1-/-;cyclin A2f/f;E2f7f/f;E2f8f/f.
137
Figure 5.3 Added loss of cyclin A1/A2 partially rescues aberrant mitosis in E2f7/E2f8 null TGCs:
Quantification of cell cycle phase indicators in E10.5 TGCs of indicated genotypes. control,
E2f7f/f;E2f8f/f; Plf-A12, Plfcre/+;cyclin A1-/-;cyclin A2f/f; Plf-78dko, Plfcre/+;E2f7f/f;E2f8f/f; Plf-
A1278qko, Plfcre/+;cyclin A1-/-;cyclin A2f/f;E2f7f/f;E2f8f/f.
138
Figure 5.4 Added loss of cyclin A1/A2 restores polyploidy in E2f7/E2f8 null TGCs: Feulgen quantification of ploidy in E10.5 TGCs of indicated groups. One-way ANOVA, * p≤0.05; ** p≤0.01; ***p≤0.001. control, E2f7f/f;E2f8f/f; Plf-A12, Plfcre/+;cyclin A1-/-;cyclin A2f/f; Plf-78dko,
Plfcre/+;E2f7f/f;E2f8f/f; Plf-A1278qko, Plfcre/+;cyclin A1-/-;cyclin A2f/f;E2f7f/f;E2f8f/f.
139
Chapter 6: Canonical E2F repressors E2F4 and E2F5 has no significant role in regulation of
endocycle in physiological conditions
6.1 Introduction
The canonical E2F repressors E2F4 and E2F5 share majority of the domains with E2F activators. Unlike atypical E2F repressors, E2F4 and E2F5 bind to DNA as dimers with
DP1/DP2 and are highly responsive to cyclin-cdk pathway. Thought to be major repressors of
E2F-responsive genes, they are expressed throughout cell cycle unlike E2F6, E2F7 and E2F8.
E2F4 and E2F5 are known to play significant role in cell cycle exit and maintenance of quiescence through interaction with C/EBP (Iakova et al., 2003). Though there is no evidence to support their role in regulation of mammalian endocycle, having observed the defects in endocycle caused by loss of E2F7 and E2F5, it was imperative to evaluate tissues from germline deleted mice. Second reason why we wanted to exclude role of canonical E2Fs is the partial phenotype observed in TGCs deficient in E2F7 and E2F8. We wondered whether this could be due to compensation by canonical E2F repressors.
In this chapter, we show by lineage-specific gene ablation strategies that the canonical
E2F repressors do not play critical role in regulation of mammalian endocycle under normal physiological conditions.
6.2 Results
140
6.2.1 Effect of E2f7/E2f8 ablation in TGCs on canonical repressors
To analyze whether loss of E2F7 and E2F8 leads to compensatory upregulation of canonical E2F repressors, we looked for changes in mRNA levels of these factors in transcriptome of TGCs of embryonic age 10.5 using NanoString. The result shown in figure 6.1 suggests that loss of E2f7 and E2f8 leads to significant increase in E2f4 levels, while expression of E2f5 and E2f6 stayed unchanged.
6.2.2 Expression of canonical E2F repressors in TGCs
Having observed the compensatory upregulation of E2f4 in E2f7/E2f8 null TGCs, we decided to look at the expression level of these genes. As shown in Figure 6.2, E2f5 and E2f6 levels were persistently low throughout the different tie points of TGCs. On contrast, there is significant decline in E2f4 levels once endocycle sets in though a second wave was observable late in the life cycle of TGCs. These results suggest that E2F4 is the most abundant among E2F repressors in TGCs and could potentially have independent roles in endocycle regulation.
6.2.3 Loss of E2f4 and E2f5 has no significant effect in TGC biology
We bred mice which are germline deleted for E2f4 and E2f5 singly to generate placentas that are null for E2f4 or E2f5. As shown in figure 6.3, neither loss of E2f4 nor E2f5 in mice appeared to affect polyploidization of TGCs.
6.3 Discussion
141
In this chapter we explored the role of canonical E2F repressors in endocycle control.
Results from our analysis suggest that E2F4 and E2F5 may not have significant role in regulation of endocycle in TGCs. One of the important drawback of this study is that we did not have access to E2f6 germline deleted mice and could not analyze E2f6 null TGCs.
While we cannot rule out a role for these canonical repressors in endocycle control that might otherwise be exposed in different genetic or environmental contexts, it would appear that their involvement is not essential. This is especially significant as we see upregulation of E2f4 levels in TGCs deficient in atypical repressors. The biological significance of this observation is not known, but a favored hypothesis would be that in absence of E2F7 and E2F8, E2F4 could have functional significance. This could explain the incomplete phenotype observed in E2f7/E2f8 null TGCs. Though this is an easy hypothesis to test, we did not venture into triply deleting E2f4,
E2f7 and E2f8. In summary, from these studies it appears that E2f7/E2f8 evolved to specifically regulate variant cell cycles in metazoans in a manner distinct from the function of canonical E2F repressor proteins as well as their plant counterparts.
6.4 Materials and methods
6.4.1 Mouse strains and genotyping.
All mice were housed in barrier conditions with a 12-hour light/dark cycle and had access to food and water ad libitum. Genotyping of mice was done from DNA preparation from tail samples using allele specific primers shown in table 4.1.
6.4.2 NanoString gene expression analyses.
142
The NanoString nCounter system was used to quantitate gene expression in TGCs and validate gene expression changes (as determined through Affymetrix analysis) in the liver. 200ng of RNA was profiled on a custom codeset of the NanoString nCounter system according to manufacturer‟s instructions. The codeset contained 200 genes that were selected in part based on their expression status in the liver, and based on their ability to regulate cellular functions involved in S phase, G2/M transition and M phase of the cell cycle (full list of genes is provided in Supplementary Table 3). Additionally, ten housekeeping genes were also included in the codeset. For the analysis of gene expression in TGCs, we isolated TGCs from 8µm H&E stained sections of OCT (Sakura) embedded frozen placental tissues using laser capture microdissection
(PALM MicroLaser system). Total RNA was extracted using Arcturus PicoPure RNA Isolation
Kit (12204-01, Applied Biosystems) according to manufacturer‟s protocol. For validation of gene expression changes in the liver, the same TRIzol®-isolated RNA was used for both
Affymetrix and NanoString nCounter assays. The nCounter data is first normalized to the geometric mean of positive control spike counts in each lane of the cartridge, followed by subtracting the minimum negative control spike counts in each lane to remove the background.
Finally, expression of each gene in the codeset is normalized with respect to the geometric mean of the house keeping genes. For heatmap construction, the geometric mean of genes from control
TGCs was set as reference, with the colors (blue or red) corresponding to fold-change in gene expression between control and mutant TGCs (p<0.05).
6.4.3 Histology
Standard protocols were used for preparation of 5μm thick paraffin embedded sections of placentas and for hematoxylin and eosin staining.
143
Figure 6.1 Loss of E2f7 and E2f8 leads to upregulation of E2f4 levels in TGCs: NanoString expression analysis of E2f activators in RNA samples collected by laser capture microdissected
E2f7/E2f8 null TGCs. Data shown as average of two samples each ± standard deviation after normalizing to housekeeping genes
144
Figure 6.2 Expression of canonical E2F activators in TGCs: NanoString expression analysis of
E2f activators in RNA samples collected by laser capture microdissected wild type TGCs. Data shown as average of two samples each ± standard deviation after normalizing to housekeeping genes
145
Figure 6.3 Canonical E2F repressors E2F4 and E2F5 have no significant role in regulation of endocycle in TGCs: Representative hematoxylin and eosin stained placental sections. Original magnification 20X
146
Chapter 7: Role of E2F activators in regulation of TGC endocycle
7.1 Introduction
Contrary to classical belief that canonical activator subclass of mammalian E2Fs, consisting of E2F1, E2F2 and E2F3 (E2F1-3) is essential for cell cycle progression, more and more evidence is appearing to suggest that E2F1-3 function is not that critical for mammalian cell proliferation, and moreover the so-called „activators‟ may also have context-dependent repressor function (Chen et al., 2009b; Chong et al., 2009; Wenzel et al., 2011). Nonetheless,
E2F-mediated transcriptional activation has been shown to contribute to endocycle control in flies (Lee et al., 2009), but until now its role in mammalian endocycles has not been explicitly evaluated. In this chapter, we utilized a genetic approach to investigate the role of E2F1-3 in
TGC endocycles.
The classic view of cell cycle regulation, E2F transcriptional activators and repressors are portrayed to antagonistically orchestrate a gene expression program essential for cell cycle progression. Though simple and well supported by more than two decades of experimentation in invertebrate and mammalian cell culture systems, major components of this model remain unsubstantiated by in vivo models. The first unresolved component is that loss of repressors or activators in tissues would lead to increased or decreased expression, respectively, of a common set of transcriptional targets. In the first part of this manuscript, we provide first in vivo evidence for the same. The second component the needs to be resolved is that animals lacking E2f
147 repressors should manifest phenotypes reasonably opposite to those lacking E2f activators. In this chapter we show evidence for opposing phenotypes upon ablation of canonical activators and atypical repressors.
7.2 Results
7.2.1 Effect of loss of atypical repressors on expression of canonical activators
Given the role of E2F7 and E2F8 in downregulating E2f1 expression one would expect a corresponding upregulation of E2f activators in context of loss of E2f7 and E2f8. We hypothesized that such an upregulation could be the driving force behind the cell cycle defects observed in TGCs lacking atypical repressors. We used NanoString to analyze the expression of
E2f activators in RNA samples collected by laser capture microdissected E2f7/E2f8 null TGCs.
As shown in figure 7.1, we observed a significant upregulation of E2f3 messages, while the levels of E2f1 and E2f2 remain relatively unperturbed.
7.2.2 Expression of E2F activators in TGCs
Expression analysis with NanoString technology showed that E2f1 and E2f2 mRNA levels were relatively low in wild type TGCs throughout placental development (from E6.5-
E17.5), whereas E2f3 mRNA levels were high prior to E8.5 and following E13.5. Thus levels of
E2f activators appeared to be at a minimum during the time when TGCs are actively endocycling
(Cross, 2005).
7.2.3. Loss of E2F activators promote TGC endocycle
148
To analyze the role of E2F activators in regulation of endocycle in TGCs, we inbred mice which are germline deleted for E2f1, E2f2 and E2f3. To avoid functional compensation among
E2F activators (Tsai et al., 2008; Danielian et al., 2008), we decided to limit our analysis to
TGCs deficient for the entire E2F activator subclass (E2f1-3). Interestingly, loss of all E2F activators produced a marked increase in the genome ploidy of TGCs at E9.5 as determined by the Feulgen technique. Analysis of E2f1-/-;E2f2-/-;E2f3-/- TGCs at later stages of development was precluded due to embryo lethality at E10.5 (Chong et al., 2009). Thus our analysis of genome ploidy levels in E2f1-3 deficient TGCs demonstrates that mammalian E2F activators have a physiologic role in suppressing endocycles.
7.3 Discussion
E2F mediated transcription contributes to endocycle control in flies and plants, but its role in mammals has never been evaluated. E2F activators are thought to be critical for normal cell cycle progression. But, studies in mice have now demonstrated that mitotic cell cycles in a variety of cell lineages proceed without E2F1-3 (Chen et al., 2009b; Chong et al., 2009; Wenzel et al., 2011). Moreover, the prediction that mice with mutant alleles of E2f4, E2f5 or E2f6 would harbor phenotypes opposite to that observed in mice lacking E2f1-3 has yet to be substantiated.
Here in this chapter we show that E2F activators antagonize the functions of atypical E2F7 and
E2F8 repressors in the regulation of endocycles in vivo. To the best of our knowledge, these findings represent the first demonstration in which distinct arms of the E2F program coordinate gene expression to control fundamental aspects of mammalian variant cell cycles in vivo.
In this chapter, we reveal a novel role of E2f1-3 in suppressing mammalian endocycles.
TGCs became aberrantly polyploid with the ablation of E2f1-3, suggesting that loss of E2F
149 activators either prematurely precipitated the onset of M-E transition or accelerated the kinetics of G/S oscillations (thereby resulting in more rounds of genome reduplication in the same span of time) subsequent to endocycle onset. It has been reported that a number of E2F1-DP target genes promote G2/M transition and M phase progression (Maqbool et al., 2010; Ishida et al.,
2011). Therefore we presumed that a high level of E2F activators would prevent endocycles and conversely, low level of E2F activators would permit endocycles.
Loss of function studies in Drosophila (Duronio et al., 1995; Royzman et al., 1997) and now as we show in mice suggest a critical role for E2F mediated transcriptional activation in regulating endocycles. But surprisingly, the mechanisms involved in flies and mice appear, at least on the surface, to be reversed. In flies, loss of dE2f1 diminishes the capacity of cells to enter
S phase and as a result both the mitotic cycle and endocycles cease, whereas dE2f1 overexpression accelerates endocycles and cells gain further ploidy. Presumably, overproduced dE2F1 protein can still be targeted for degradation late in S phase by the E3 ubiquitin ligase
Cul4cdt2 and thus, total dE2F1 level still oscillates during the cell cycle (Shibutani et al., 2008).
Consistent with a requirement for dE2F1 oscillatory expression that in turn drives periodic transcription of its key target cyclin E, salivary gland cells that express a constitutively stable form of dE2F1 protein arrest in S phase and fail to endocycle (Shibutani et al., 2008). In contrast to flies, even the absence of all three E2F activators in the mouse embryo is insufficient to preclude cells from entering S phase or completing mitotic cell division (Chong et al., 2009).
Furthermore, their absence in endocycling TGCs and hepatocytes as we have shown leads to pronounced ploidy gains. This raises an intriguing paradox of why altering the levels of activating E2Fs in flies and mice yield opposite consequences. One possible explanation for this
150 difference may be the number of E2F family members that each species encodes, contributing to the availability of compensatory mechanisms that operate in flies and mice.
In preparation for the M-E transition, it may be that one or multiple upstream signaling pathways would serve to downregulate E2f1-3 expression in order to cease further mitotic cell division. Consistent with this model, we found that levels of E2f1-3 expression were lowest at a time when TGCs are actively endocycling. The result that an increase in TGC-specific E2f3 expression corresponded with a period during which TGCs have stopped endoreplicating may be evidence to suggest that E2F3 has non-endocycle related function in these tissues. Based on these results, we suggest that attenuating E2F1-3 function may be a key event in determining the timing of endocycle onset in mammals. In support of this hypothesis, work from the Calvi laboratory (Maqbool et al., 2010) recently demonstrated that the expressions of dE2f1 and its obligate binding partner dDp were significantly decreased in fly endocycling tissues (larval salivary gland and fat body). As a result, many E2F1-DP target genes were also expressed at significantly lower levels in these tissues than in mitotic cycling tissues (larval brain and imaginal disc). The precise upstream signaling events that determine how and when E2F1-3 activity becomes downregulated are currently unknown, but may involve components of the
Notch pathway.
A second intriguing point that is raised by our study is how DNA replication during endo
S phase might occur in the absence of E2F1-3. It has already been shown that n-Myc was sufficient to stimulate DNA replication in retinal progenitor cells lacking E2f1-3 (Chen et al.,
2009b), eliciting the possibility that n-Myc might function in a similar capacity in E2f1-3 deficient TGCs and hepatocytes. Thus, the previous observation that hepatocyte-specific overexpression of Myc resulted in precocious polyploidy (Conner et al., 2001) might be
151 explained by the potential overlapping actions of Myc and E2F1-3 at the G(1)/S transition.
Finally, the fact that Dp1 ablation in the mouse, which resulted in lethality by E8.5-9.5, suppressed TGC endocycles (Kohn et al., 2003) might seem at first contradictory to our result that loss of E2f1-3 prompted ectopic endocycles in the same cells. However, because trophoblast stem cell differentiation appeared to be severely compromised by the loss of DP1, it was difficult to ascertain how TGC endocycles might have been truly impacted. Furthermore, the DNA binding activity of E2F4-6 was also presumably abrogated by the loss of Dp1, which would have confounded the analysis of the relative contribution of activator versus canonical repressors to endocycle control.
In summary, we showed by utilizing targeted gene ablation strategies that mammalian
E2F activators suppress endocycles in vivo, and presented data to suggest that decreased expression of E2f1-3 may serve as an important developmental trigger to endocycle onset in
TGCs
7.4 Materials and methods
7.4.1 Mouse strains and genotyping.
All mice were housed in barrier conditions with a 12-hour light/dark cycle and had access to food and water ad libitum. Allele specific primers were used for genotyping mice.
7.4.2 Ploidy analysis using the Feulgen technique.
For quantification of genome ploidy, placentas and livers were fixed in 10% neutral buffered formalin and subsequently embedded in paraffin. 5μm thick sections were cut,
152 deparaffinized in xylene, treated with 1N hydrochloric acid for eight minutes and stained with
Schiff reagent for one hour. Aqueous light green (1%) was used to counterstain the cytoplasm.
Mounted tissues were imaged using the Axioskop 40 microscope (Zeiss) and nuclear intensity was quantified using the software ImageJ (http://rsbweb.nih.gov/ij/) following standard protocols
(Hardie et al., 2002). TGC ploidy was expressed relative to diploid (2n)-tetraploid (4n) labyrinthine trophoblast cells, and the average percent (± standard deviation) of TGCs in different bins, which represent different ploidy levels, was plotted as histograms. For bi- or multi-nucleated TGCs, the ploidy levels of individual nuclei were summed to show cumulative ploidy of cells. In contrast, hepatocyte ploidy is expressed on a per nucleus basis.
7.4.3 NanoString gene expression analyses.
The NanoString nCounter system was used to quantitate gene expression in TGCs and validate gene expression changes (as determined through Affymetrix analysis) in the liver. 200ng of RNA was profiled on a custom codeset of the NanoString nCounter system according to manufacturer‟s instructions. The codeset contained 200 genes that were selected in part based on their expression status in the liver, and based on their ability to regulate cellular functions involved in S phase, G2/M transition and M phase of the cell cycle (full list of genes is provided in Supplementary Table 3). Additionally, ten housekeeping genes were also included in the codeset. For the analysis of gene expression in TGCs, we isolated TGCs from 8µm H&E stained sections of OCT (Sakura) embedded frozen placental tissues using laser capture microdissection
(PALM MicroLaser system). Total RNA was extracted using Arcturus PicoPure RNA Isolation
Kit (12204-01, Applied Biosystems) according to manufacturer‟s protocol. For validation of gene expression changes in the liver, the same TRIzol®-isolated RNA was used for both
153
Affymetrix and NanoString nCounter assays. The nCounter data is first normalized to the geometric mean of positive control spike counts in each lane of the cartridge, followed by subtracting the minimum negative control spike counts in each lane to remove the background.
Finally, expression of each gene in the codeset is normalized with respect to the geometric mean of the house keeping genes. For heatmap construction, the geometric mean of genes from control
TGCs was set as reference, with the colors (blue or red) corresponding to fold-change in gene expression between control and mutant TGCs (p<0.05).
7.4.4 Statistical Analysis.
Quantification of IHC/IF was performed on multiple sections per tissue sample.
Percentages of positive cells are reported as average ± standard deviation. Flow cytometry of liver ploidy are representative of at least three independent experiments. One-way ANOVA was used to analyze data. When appropriate, p-values were adjusted by Holm‟s method within each ploidy category.
154
Figure 7.1 Loss of E2f7 and E2f8 leads to upregulation of E2f3 levels in TGCs: NanoString expression analysis of E2f activators in RNA samples collected by laser capture microdissected
E2f7/E2f8 null TGCs. Data shown as average of two samples each ± standard deviation after normalizing to housekeeping genes
155
Figure 7.2 Expression of E2f activators in TGCs: NanoString expression analysis of E2f activators in RNA samples collected by laser capture microdissected wild type TGCs of different embryonic age. Data shown as average of two samples each ± standard deviation after normalizing to housekeeping genes
156
Figure 7.3 Loss of E2f activators promotes TGC endocycles: Top panel shows representative
E9.5 H&E placenta sections showing control and E2f1-/-;E2f2-/-;E2f3-/- (123tko) TGCs. Lower panel shows results of Feulgen quantification of ploidy in E9.5 control and 123tko TGCs. One- way ANOVA, * p≤0.05, ** p≤0.01, *** p≤0.001
157
Chapter 8: E2F1 and E2F3a functionally antagonizes atypical E2F repressors
8.1 Introduction
Progression through the mitotic cell cycle is thought to require coordinated E2F dependent transcriptional activation and repression. Consistent with this paradigm, overexpression of E2f activators (E2f1-3) in a variety of cell lines induces ectopic proliferation
(Johnson et al., 1993; DeGregori et al., 1997) and their combined ablation in mouse embryo fibroblasts (MEFs) elicits a profound cell cycle arrest (Wu et al., 2001). Conversely, overexpression of canonical E2f repressors (E2f4-6) halts cell cycle progression and their combined ablation in MEFs evokes insensitivity to inhibitory mitogenic signals (Gaubatz et al.,
2000). In chapter 2 of this manuscript we have provided in vivo evidence to support this hypothesis. In addition, the observation that E2f1-E2f3 null TGCs have a phenotype exactly opposite to that of E2f7/E2f8 null counterparts, suggest that E2F activators and atypical E2F repressors might be acting antagonistically in regulation of endocycle in TGCs (chapter 7).
Above all, in chapter 7 we have shown evidence to suggest that loss of E2f7 and E2f8 leads to upregulation of E2f3.
Based on these observations, we hypothesized that added loss E2F activators could rescue the endocycle defects observed in E2f7/E2f8 null TGCs To rigorously test this possibility we generated and analyzed mice with TGCs deficient for E2F7, E2F8 and one of the E2F
158 activators, E2F1 (E2f1-/-) and E2F3a (E2f3a-/-). The specific ablation of E2f7f/f and E2f8f/f alleles in TGCs was achieved by Plfcre/+ (Plf-178tko and Plf-3a78tko).
8.2 Results
8.2.1 Added loss of E2f1 and E2f3a rescues defects in nuclear size and nuclear morphology of
E2f7/E2f8 null TGCs
Analysis of placental sections from E10.5 Plf-178tko embryos showed that TGCs with larger nuclei than Plf-78dko and Plf-3a78tko embryos (Figure 8.1). In concordance with rescue of nuclear size, we also observed that most of these cells were mononucleated with very few cells showing karyokinesis. These results suggest that loss of E2F1 and E2F3a could rescue the endocycle to mitosis switch that occurs in E2F7/E2F8 null TGCs.
8.2.2 E2F1 and E2F3a dictate the molecular consequences in E2f7/E2f8 null TGCs
We further went on with analyzing the triply deficient TGCs for the cell cycle defects characteristic of loss of E2f7 and E2f8. We observed partial rescue in markers for aberrant S- phase entry and activation of mitotic machinery, more upon loss of E2f3a compared to E2f1.
Surprisingly, cyclin A2 levels seem unchanged in triple knock out TGCs, suggesting a non- transcriptional component for the observed upregulation E2f7/E2f8 null TGCs.
8.2.3 E2F1 and E2F3a functionally antagonizes E2F7 and E2F8 in regulation of endocycle in
TGCs
159
Having noticed the rescue in nuclear and molecular phenotype, we wondered whether this would translate into improvement in ploidy levels in TGCs. To this end we analyzed Plf-
178tko and Plf-3a78tko TGCs by Feulgen staining. As expected we could see a significant increase in TGCs with higher ploidy levels. These results support findings that we made in the previous chapters and suggest that similar to what was observed in mitotic cell cycles, E2F activators and atypical E2F repressors drive cell cycle programs in an antagonistic fashion in regulation of endocycle in TGCs.
8.3 Discussion
Several lines of evidence (presented in this and previous chapters) support antagonistic roles for canonical E2F activators and atypical repressors in orchestrating mammalian endocycles. The genetic inactivation of activators and atypical repressors in TGCs resulted in opposite cellular phenotypes, with E2f1-3 loss promoting larger nuclei with increased ploidy and
E2f7/E2f8 loss promoting smaller nuclei with reduced ploidy. A recent survey in vitro of E2F1-3 target genes revealed that a subset of these is involved in the positive control of mitosis, including Cyclin A2, Cyclin B1, Cdc25C and Cdc2 (Ishida et al., 2011). Thus, an early decline in
E2F activator function, as „artificially‟ caused by their genetic inactivation, might have been sufficient to induce TGCs and hepatocytes to precociously enter endocycles, an act that would have led to their excess polyploidization.
Here we consider two potential mechanisms for how these two E2F arms might normally coordinate gene expression in vivo. The first involves the recruitment of E2F activators and repressors to distinct E2F binding elements present on target promoters. The occupancy of either activators or repressors on these promoter elements would thus be predicted to dictate
160 transcriptional output. Interestingly, a survey performed within this study of genes regulated by
E2F1-3 and E2F7/E2F8 revealed that approximately 40% of target promoters contained just a single E2F binding element within 5.0kb distance spanning the transcriptional start site. This raises a second possibility where competitive binding to the same element would rely on different DNA binding affinities of activators and repressors, or on the spatial and/or temporal regulation of their expression and activity.
Finally, although we did not closely examine these promoters for the presence of other transcriptional binding elements, it is possible that co-factors might also have been intimately involved in recruiting E2Fs to these target promoters. This latter scenario would not be surprising given that both E2F activators and repressors have been found to be capable of associating with a rather diverse cohort of transcriptional factors in chromatin-bound complexes.
Because we and others (Maqbool et al., 2010) have demonstrated that E2F1-3 expression appears to be significantly diminished in endocycling cells, E2F7/E2F8 mediated repression would thus be expected to dominate and enforce the mitotic block necessary for continuous endocycles after
M-E transition. In line with this interpretation, we propose that the antagonism exhibited by E2F activators and atypical repressors would mainly be responsible for coordinating cellular events that enable the M-E switch.
Finally, in this chapter we show that the inactivation of a single E2F activator (E2F1 or
E2F3a) is sufficient to suppress endocycle defects caused by E2f7/E2f8 deficiency. Interestingly, the level of cyclin A2 protein, which had been elevated in E2f7/E2f8 deleted TGCs and hepatocytes, was not reduced by the ablation of E2f1. This observation would suggest that cyclin
A2, given its critical role at the G2/M transition, either is regulated independently of E2F1/E2F3a
161 at the transcriptional level or perhaps more likely, is predominantly regulated at the post- transcriptional level (e.g. by APC/C).
To summarize, our results in previous chapters suggest that two arms of the E2F program, an activating arm (E2F1-3) that is dependent on dimerization with DP proteins and is regulated by canonical Cdk-Rb signaling and a second atypical repressor arm (E2F7/E2F8) that is both Rb- and DP-independent, converge to antagonistically regulate the endocycle.
8.4 Materials and methods
8.4.1 Mouse Strains and Genotyping
All protocols involving mice were approved by the Institutional Animal Care and Use
Committee at The Ohio State University. Transgenic mice used for this study were maintained in a mixed 129SvEv; C57BL/6; FVB background. Allele-specific (E2f7/8, E2f1, E2f3a and PlfCre) were used for PCR genotyping.
8.4.2 Histology, Immunostaining and Quantification
Standard protocols were used for preparation of 5μm thick paraffin embedded sections of placentas and for hematoxylin and eosin staining. For immunohistochemistry, primary antibodies against P-H3 (06-570, Millipore), BrdU (MO-0744, DAKO), CyclinA2 (sc-596, Santa Cruz) and
CyclinB1 (sc-752, Santa Cruz) were used. Pregnant mice at 10.5 days postcoitum were given intraperitoneal injections of BrdU (100 µg/grams of body weight) 30 min prior to harvesting.
Detection of primary antibodies was done using species specific biotinylated secondary antibodies along with Vectastain Elite ABC reagent (Vector labs) and DAB peroxidase substrate
162 kit (Vector labs). Nuclear counterstaining was done using hematoxylin. Images of immunostained sections were captured using Eclipse 50i (Nikon) and Axioskop 40 (Zeiss) microscopes and positive cells were quantified using Metamorph Imaging 6.1 software. Three sections per sample and at least three different samples for each genotype were analyzed. Data is reported as the average ± SD of percentage of positive cells.
8.4.3 Ploidy analysis using the Feulgen technique.
For quantification of genome ploidy, placentas and livers were fixed in 10% neutral buffered formalin and subsequently embedded in paraffin. 5μm thick sections were cut, deparaffinized in xylene, treated with 1N hydrochloric acid for eight minutes and stained with
Schiff reagent for one hour. Aqueous light green (1%) was used to counterstain the cytoplasm.
Mounted tissues were imaged using the Axioskop 40 microscope (Zeiss) and nuclear intensity was quantified using the software ImageJ (http://rsbweb.nih.gov/ij/) following standard protocols
(Hardie et al., 2002). TGC ploidy was expressed relative to diploid (2n)-tetraploid (4n) labyrinthine trophoblast cells, and the average percent (± standard deviation) of TGCs in different bins, which represent different ploidy levels, was plotted as histograms. For bi- or multi-nucleated TGCs, the ploidy levels of individual nuclei were summed to show cumulative ploidy of cells.
8.4.4 E2F binding site search.
Promoter sequences (-3000bp ~ +2000bp) of genes antagonistically regulated by E2F1-3 and E2F7/E2F8 were obtained from the UCSC genome browser (http://genome.ucsc.edu/), and
163
E2F consensus binding sites were identified using TFSearch
(http://www.cbrc.jp/research/db/TFSEARCH.html).
8.4.5 Statistical Analysis.
Quantification of IHC/IF was performed on multiple sections per tissue sample.
Percentages of positive cells are reported as average ± standard deviation. Flow cytometry of liver ploidy are representative of at least three independent experiments. One-way ANOVA was used to analyze data. When appropriate, p-values were adjusted by Holm‟s method within each ploidy category.
164
Figure 8.1 Added loss of E2f1 or E2f3a rescues nuclear phenotypes in E2f7/E2f8 null TGCs:
Representative H&E sections of E10.5 TGCs. control, E2f7f/f;E2f8f/f; 1ko, E2f1-/-; 3ako, E2f3a-/-;
Plf-178tko, Plfcre/+;E2f1-/-;E2f7f/f;E2f8f/f; Plf-3a78tko, Plfcre/+;E2f3a-/-;E2f7f/f;E2f8f/f
165
Figure 8.2 Ablation of E2f3a and E2f1 partially rescues cell cycle defects in E2f7/E2f8 null
TGCs: Quantification of cell cycle phase indicators in E10.5 TGCs of indicated genotypes. Data shown as average ± Standard Deviation. control, E2f7f/f;E2f8f/f; 1ko, E2f1-/-; 3ako, E2f3a-/-; Plf-
178tko, Plfcre/+;E2f1-/-;E2f7f/f;E2f8f/f; Plf-3a78tko, Plfcre/+;E2f3a-/-;E2f7f/f;E2f8f/f
166
Figure 8.3 Ablation of E2f3a and E2f1 rescues polidy defects in E2f7/E2f8 null TGCs: Feulgen quantification of ploidy in E10.5 TGCs of indicated genotypes. Data shown as average ±
Standard Deviation. control, E2f7f/f;E2f8f/f; 1ko, E2f1-/-; 3ako, E2f3a-/-; Plf-178tko, Plfcre/+;E2f1-/-
;E2f7f/f;E2f8f/f; Plf-3a78tko, Plfcre/+;E2f3a-/-;E2f7f/f;E2f8f/f
167
Chapter 9: Atypical E2Fs play significant role in regulation of endomitosis in megakaryocytes
9.1. Introduction
Megakaryocytes are platelet precursor cells which are characterized by expression of
CD41, CD61, CD42 (glycoprotein Ib) and glycoprotein V (Hodohara et al., 2000; Roth et al.,
1996). Transcription factors involved in commitment of megakaryocytic progenitors to commit to megakaryocyte lineage is reasonably well studied. For example, megakaryocyte specific elimination of GATA1 leads to severe thrombocytopenia due to dysmegakaryopoiesis
(Shivdasani et al., 1997). Other studies have shown that ets transcription factor Fli-1 is essential for megakaryopoeisis and mutations in this transcription factor have been known to be associated with congenital thrombocytopenia in humans (Doubeikovski et al., 1997; Hart et al., 2000).
Identification and cloning of c-Mpl and Tpo (ligand for c-Mpl) has advanced the field of megakaryocyte biology significantly by exposing the biological pathways involved in differentiation of these cells (de Sauvage et al., 1994; Kuter et al., 1994; Lok et al., 1994; Methia et al., 1993; Sohma et al., 1994; Solar et al., 1998; Wendling et al., 1994).
Platelets are formed by fragmentation of pseudopodial projections (proplatelets) from mature megakaryocyte cell membranes (Patel et al., 2005). This process involves massive reorganization of megakaryocyte cytoskeleton leading to generation of a new compartment complementing cytoplasmic components including organelles, granules and soluble macromolecules (Italiano et al., 1999; Kaushansky, 2008). Generally it is believed that each
168 mature megakaryocte will generate close to 3000 platelets and once the life span is over, the residual nuclear material undergoes macrophage mediated phagocytosis (Radley and Haller,
1983; Stenberg and Levin, 1989).
Indirect evidence suggest that size of megakaryocytes could have an impact on the amount of platelets that it could generate, as seen in patients with GATA1, Gylcoprotein Ib/IX complex and Wiskott Aldrich syndrome protein mutations (Geddis and Kaushansky, 2004).
Current consensus is that the number of platelets in circulation is dependent on total number of megakaryocytes and size of individual megakaryocytes (Hoffman, 1989). High ploidy in megakaryocytes is associated with an increase in cell size, mRNA and protein content (Hancock et al., 1993). The process of polyploidy starts with terminal differentiation of megakaryocytes and is associated with a process called endomitosis where nuclear DNA content and cytoplasmic volume increases in the absence of karyokinesis and cytokinesis (Odell and Jackson, 1968). The process of endomitosis is characterized by multiple rounds of DNA synthesis with intervening
Gap phase, synthesis of Cyclin D3 isoform, decrease in CDK1 kinase activity, low levels of
Cyclin B1, centrosomal and centriolar amplification, DNA condensation and breakdown and reorganization of nuclear membrane (Datta et al., 1996; Garcia and Cales, 1996; Gu et al., 1993;
Handeli and Weintraub, 1992; Odell et al., 1968; Vitrat et al., 1998; Vitrat et al., 1996; Zhang et al., 1996). Together these observations suggest that endomitosis as a variant or abortive cell cycle where the cells go through successive rounds of G1, S, G2 and part of M (till anaphase A) leading to cells with ploidy content reported as high as 128N (Nagata et al., 1997; Ravid et al.,
2002; Roy et al., 2001). Several studies have shown that continually proliferating megakaryocytes (induced by overexpression of T antigen c-myc or E2F1), even when expressing
169 differentiation markers, can have low ploidy presumably by failure to switch from mitosis to endomitosis (Guy et al., 1996; Kubota et al., 1996; Ravid et al., 1993).
At what point does the megakaryocytes get committed to endomitosis is still not clear, though current evidence from analysis of ploidy of these cells in different species suggests that majority of the are in 16N, while a small proportion goes further to achieve ploidy levels as high as 128N (Jackson et al., 1997). A switch from cyclin D1 to cyclin D3 happens, but is of unknown significance as transgenic models overexpressing either of it leads to increase in ploidy (Sun et al., 2001; Zimmet et al., 1997). In addition, augmented levels of cyclin A and cyclin E is also noted in these cells (Datta et al., 1998; Garcia et al., 2000). The levels of cyclin B1 is low as compared to cyclin A and has been demonstrated to be due to increased destruction by Anaphase
Promoting Complex resulting in high Cyclin B1/DNA ratio in megakaryocytes of high ploidy
(Vitrat et al., 1998; Zhang et al., 1998). One of the key observations as to how these cells go through abortive mitosis is through the observation that AIM-1 (one of the Aurora kinases which is believed to be essential for progression through mitosis) is transcriptional downregulated in both human and murine megakaryocytes and also in response to TPO, though the cause and effect of this is still not proven (Kawasaki et al., 2001; Zhang et al., 2001).
9.2. Results
9.2.1. Loss of E2f7/E2f8 disrupts endomitosis in megakaryocytes
Since loss of E2f7 and E2f8 led to disruption of polyploidy in two major cell types, hepatocytes and TGCs, we interested in knowing whether these functions of E2F7 and E2F8 are conserved in megakaryocytes also. Germline deletion of E2f7 and E2f8 leads to embryonic
170 lethality. So we depended on megakaryocyte specific targeted deletion strategy using well characterized Pf4-Cre mouse model (ref). This transgenic mouse line is known to express cre recombinase in megakaryocyte specific manner upon terminal differentiation. We bred Pf4-Cre transgenic mice with E2f7f/f;E2f8f/f mice to generate Pf4-Cre; E2f7f/f;E2f8f/f mice. Cre negative littermates were used as controls for the experiments.
We first looked at the morphology of megakaryocytes in cre positive mice and observed that E2f7/E2f8 null megakaryocytes were significantly smaller than those in cre negative littermate controls, suggesting possibility that ploidy of these megakaryocytes could be lower
(Figure 9.1). We then looked at the ploidy in these cells by flow cytometry, using CD41 as a marker for these cells. To our surprise, we found that cre positive megakaryocytes had significant reduction in their ploidy compared to cre negative littermate controls (Figure 9.2). In addition, results from these experiments also showed that this phenotype is dependent on both
E2F7 and E2F8 and that there is a dosage effect as seen in TGCs and hepatocytes.
9.2.2 Disruption of endomitosis by loss of E2f7/E2f8 is of little physiological consequence
Nuclear size and ploidy of megakaryocytes are believed to be two most important determinants for their ability to generate platelets. But in our experience, TGCs and hepatocytes with significantly low ploidy could function well in normal physiological context. Moreover,
Pf4-Cre; E2f7f/f;E2f8f/f mice did not shown any evidence of bleeding disorders in our hands. Still, we went ahead and checked platelet count in these mice. Platelet count in Pf4-Cre; E2f7f/f;E2f8f/f mice were lower than cre negative cases, but were always within normal range for the age of animals. Also, the difference in platelet count did not achieve statistical significance. This data suggests that though loss of E2f7 and E2f8 though leads to decrease in ploidy in megakaryocytes,
171 these cells still could generate sufficient number of platelets to maintain the requirement of the animal.
9.2.3 Molecular mechanisms underlying the phenotype
We did not delve much into identifying molecular mechanisms behind this phenotype, but preliminary results from the changes in expression of genes which were in our custom
NanoString panel suggest that G2-M targets that were deregulated in TGCs and hepatocytes were deregulated in the double mutant megakaryocytes also, suggesting a shared transcriptional derepressional mechanism as underlying cause of this phenotype.
9.2.4 Loss of canonical activators also leads to similar phenotype
Having observed disruption of endomitosis in E2f7/E2f8 megakaryocytes, we hypothesized that if the underlying mechanisms behind the regulation of polyploidy is shared between megakaryocytes, TGCs and hepatocytes, there should be an opposing phenotype in absence of canonical E2F activators. We used Mx1-Cre system to delete E2f1-E2f3 in bone marrow precursors and analyzed megakaryocytes in these mice. To our surprise, we observed a phenotype similar to that was seen in E2f7/E2f8 null megakaryocytes with significant reduction in ploidy. We entertained the possibility that there could be increased apoptosis in triple knock out cells, but sub-G0 population and the percentage of CD41 positive cells did not seem different between the generic groups analyzed.
9.3 Discussion
172
Megakaryocytes, the platelet precursors, are the third subset of cells which achieve endopolyploidy in mammals. Unlike TGCs and hepatocytes, these cells use endomitosis to achieve high levels of ploidy. Previous studies have shown that ability of megakaryocytes to generate platelets is dependent on their ploidy and size. Though some amount of data suggest role of E2F activators in regulation of ploidy in these cells, little is known about any role of atypical repressors. Using targeted gene knockout approach, we show here that atypical E2F repressors E2F7 and E2F8 play key role in regulation of endomitosis in megakaryocytes. Our preliminary data suggests that the underlying mechanism by which E2F7 and E2F8 regulate this process is similar to that seen in TGCs and hepatocytes.
On the other hand, quite unexpectedly, we observed a similar result in E2f1-E2f3 null megakaryocytes also. Multiple factors should be considered before drawing conclusions from this result. First, the cre system that was used in this experiment was different and non-specific compared to the one used in analysis of E2f7 and E2f8. Mx1-Cre is expressed in bone marrow precursors and might have an impact on differentiation of different lineages. Further ongoing experiments using Pf4-Cre will address this question directly. Secondly, loss of E2F activators are known to induce apoptosis in different cell lineages (Chong 2009). Vulnerably of megakaryocytes for apoptotic stimuli is not well studied yet. Our results suggest that this may not be the case, but further studies using apoptosis specific markers like annexin is needed to clarify this question.
In summary, using mouse models we have demonstrated a unique role of E2F activators and atypical E2F repressors in regulation of endomitosis in megakaryocytes, which appears to be similar to that in TGCs and hepatocytes.
173
9.4 Materials and methods
9.4.1 Mouse Strains and Genotyping
All protocols involving mice were approved by the Institutional Animal Care and Use
Committee at The Ohio State University. Transgenic mice used for this study were maintained in a mixed 129SvEv; C57BL/6; FVB background. Allele-specific (E2f7/8) and transgene specific primers (Pf4-Cre and Mx1-Cre) were used for PCR genotyping.
9.4.2 Ploidy analysis using flow Cytometry in megakaryocytes.
Bone marrow from 2 month old mice were collected by flushing femur and tibia.
Samples were washed in processing media (RPMI with 10% FCS) and treated with RBC lysis buffer for 1 minute. After RBC lysis, samples are washed twice in processing media and 2 million cells were aliquoted into new tubes. Cells were spun down and washed with PBS and resuspended in 1 ml of staining media (Sterile PBS with 1% FBS) with 2ug of FITC-conjugated
CD41 antibody (ebiosciences) for 30 minutes in ice. FITC-conjugated isogenic IgG was used as control along with „no antibody‟ controls. After 30 minutes samples were spun down and washed with PBS and treated with propedium iodide solution (50ug PI, 100ug RNase A, per ml of 0.1% sodium citrate solution with 0.1% Triton-X) for 20 minutes at room temperature. The samples were then subjected to flow cytometry at OSUCCC flow cytometry core facility using Aria (BD
Biosciences) and cell cycle profiles were generated using FlowJo.
9.4.3 Platelet analysis
174
0.5 cc of blood was drawn per animal directly from the heart of mice at the time of harvest and collected in EDTA precoated tubes and platelet count was done by automated hemocytometer at the OSU Veterinary College Mouse Phenotyping core facility.
9.4.4 pIpC injections for induction of Mx1-Cre
Six-week-old mice received five intraperitoneal injections of 250μg of polyinosine- polycytidine (pIpC, Sigma) dissolved in sterile PBS every alternate day. Mice were sacrificed 24 hours after the last injection, and their bone marrow harvested for FACS analyses.
175
Figure 9.1 E2f7/E2f8 deficient megakaryocytes have smaller nuclei: Representative micrograms showing megakaryocytes stained with nuclear stain Geimsa. Original magnification 40X
176
Figure 9.2 Loss of E2f7 and E2f8 leads to significant decrease in ploidy in megakaryocytes: Top panel shows quantifications of cells in different phases of cell cycle based on their DNA content as analyzed by flow cytometry. Lower panels shows representative flow cytometry images of PI stained CD41 positive cells
177
Figure 9.3 E2f7/E2f8 null megakaryocytes generate sufficient platelets to maintain platelet count within normal limits
178
Figure 9.4 Loss of E2f1, E2f2 and E2f3 leads to significant decrease in ploidy in megakaryocytes: Top panel shows quantifications of cells in different phases of cell cycle based on their DNA content as analyzed by flow cytometry. Lower panels shows representative flow cytometry images of PI stained CD41 positive cells
179
Table 9.1PCR genotyping primers
Expected genotyping Gene Primer sequences fragment size
Mx-cre 5‟ATGCTTCTGTCCGTTTGCCG Transgene: 260bp Pf4-Cre 5‟CCTGTTTTGCACGTTCACCG
5‟AGCCACTGGATATGATTCTTGGAC Wild type:188bp E2f1ko 5‟AGAAGTCACGCTATGAAACCTCAC Knockout: 218bp 5‟AGTGCCAGCGGGGCTGCTAAAG
5‟GCCCCTAACACATGCACCCATTGG Wild type:210bp E2f2ko 5‟CCTGAGCGAGTCGGAGGATGG Knockout: 270bp 5‟ACCAAAGAACGGAGCCGGTTGGCG
5‟TGTGAATAATTTTTGGCATGTTTT Wild type: 152bp E2f3flox 5‟CTTATTCTGAGTGTGGACATACCG Knockout: 340bp E2f3ko 5‟AAGGGAAGGGAAAATTAAATCTGA Conditional: 202bp
5‟AGGCAGCACACTTGACACG Wild type: 300bp E2f7flox 5‟ACTTTTGGGACAGAGGTAGGA Knockout: 400bp E2f7ko 5‟CCAAGATGAAGGCCGAGATGCTAC Conditional: 340bp
5‟TAAAAAGCTTTGCGGTCGTT Wild type: 192bp E2f8flox 5‟AAGCCAACCTCGATGAATTG Knockout: 500bp E2f8ko 5‟CTCGCATCATCGTCTGCTAA Conditional: 230bp
180
Chapter 10:.Summary and perspectives
In the current study, using targeted gene ablation, we have attempted to address biological functions of newly discovered members of E2F family of transcription factors, E2F7 and E2F8. We explored the cell cycle regulatory functions of E2F7 and E2F8 in mitosis and variant cell cycles and demonstrate they form critical repressor arm which balances cumulative
E2F activity in cell cycle phase dependent manner in mitotic and variant cell cycles.
10.1 Balancing E2F activity in mitosis and variant cell cycles
Previously published work in Drosophila have provided interesting insights into regulation of E2F-responsive genes by demonstrating antagonistic roles for dE2f1 and dE2f2 in the control of cell proliferation and larval development (Ambrus et al., 2007; Cayirlioglu et al.,
2003; Frolov et al., 2001; Frolov et al., 2003; Weng et al., 2003). But surprisingly, further analysis failed to show that activators and repressors regulate same subset of genes (Cayirlioglu et al., 2003; Dimova et al., 2003; Stevaux et al., 2005). It is generally viewed that mammalian activators (E2F1, E2F2 and E2F3) and classic repressors (E2F4, E2F5 and E2F6) also work together to orchestrate the expression of E2F target genes during the cell cycle. However, on a similar note, surprisingly little evidence in vivo exists to support the notion that these two subsets of mammalian E2Fs regulate the same set of target genes or even function in the same tissues.
181
Mouse placenta is one physiological context that has provided insightful clues into the critical functions of the Cdk-Rb-E2F pathway in mammals (Geng et al., 2003; Kohn et al., 2003;
Kohn et al., 2004; Kozar et al., 2004; Malumbres et al., 2004; Parisi et al., 2003; Wenzel et al.,
2007) (Geng et al., 2003; Kohn et al., 2004; Wenzel et al., 2007). Previous work from our lab demonstrated that the targeted disruption of E2f7 and E2f8 in mice leads to lethality by embryonic age 11.5, and consistent with their high expression in placental tissues, we now show that double mutant embryos exhibit severe placental abnormalities attributable to cell cycle defects. Using placenta as model system, we show here that a single activator (E2F3a) and the two atypical repressors (E2F7/E2F8) are utilized in vivo to co-regulate cell proliferation by antagonistically regulating same subset of genes during development. This observation is evolutionarily remarkable given the divergent nature of their DNA binding domains, with E2F3a utilizing inter-molecular interactions with its partner protein DP (Zheng et al., 1999) and
E2F7/E2F8 utilizing both inter- and intra-molecular interactions between its two DNA binding domains (Di Stefano et al., 2003; Logan et al., 2005; Maiti et al., 2005a).
Cell cycle studies in Drosophila have also yielded significant knowledge into the role of
E2Fs in endocycle onset and regulation. Numerous target genes activated by dE2F1 are required for G/S transition and DNA synthesis. Among these targets is cyclin E and indeed, genetic or functional ablation of dE2F1 results in drastic reduction in DNA synthesis and endocycle progression in the Drosophila embryo (Duronio and O‟Farrell, 1995; Duronio et al., 1995, 1998;
Royzman et al., 1997). Conversely, expression of a mutant form of dE2F1 unable to undergo ubiquitin-mediated proteolysis results in constitutive cyclin E expression and consequently, a block in endoreplication in larval salivary glands (Shibutani et al., 2008). Loss of dE2F2, a component of the multi-protein Myb-Muv or dREAM repressor complex, also led to continuous
182 cyclin E expression and severe developmental defects in tissues that endocycle (Weng et al.,
2003). Finally, an important point to note in these loss and gain of function studies is that DNA replication defects were not strictly limited to endocycling tissues in dE2F mutant flies. Thus, these studies also lend support to the notion that mitotic cell cycles and endocycles may share a subset of regulatory factors to ensure faithful genome duplication.
Though it is widely believed that E2Fs might play a similar role in regulation of mammalian variant cell cycles, any attempt to study this is compounded by the innate redundancy among different E2F family members. In this study, we systematically analyzed effects of ablation of different groups of E2F family members and demonstrate that canonical
E2F activators and atypical E2F repressors play antagonistic roles in regulation of endocycle in
TGCs. To our surprise we found that loss of E2F1-3 and E2F7-8 resulted in exactly opposite effects in ploidy of TGCs. In parallel studies, we show that these functions are conserved in different mammalian polyploidy cells like hepatocytes and megakaryocytes. This set of data provides first in vivo evidence to show phenotypic antagonism between E2F activators and repressors, and complement the results that we observed in mitotic cell cycle.
Results from our experiments suggest that the basis for how E2F1-3 and E2F7/E2F8 proteins coordinate gene expression is dependent on their temporal occupancy on target promoters during the cell cycle. Studies performed in multiple cell lines have shown that E2F activator protein levels peak at G1/S, followed by a precipitous drop in early-mid S phase (Ishida et al., 2001; Leone et al., 1998). In contrast, levels of E2F7 and E2F8 begin to increase in mid- late S phase, peak in G2 and decline as cells approach M and the next G1 phase (de Bruin et al.,
2003; Maiti et al., 2005a). The sequential accumulation of E2F activators followed by
E2F7/E2F8 proteins support a mechanism where the loading of E2F3a on target promoters at
183
G1/S leads to their activation, including that of E2f7 and E2f8. As atypical E2F protein levels increase in mid-to-late S phase and E2F3a protein is targeted for degradation (Leone et al.,
1998), G1/S targets become efficiently repressed by E2F7 and E2F8, leading to their decline by
G2/M. Thus, the E2F3a-dependent activation of E2f7 and E2f8 that we have demonstrated in our current work may be viewed as a mechanism to ensure that rhythmic waves of E2F-dependent activation and repression drive cell cycle dependent gene expression.
10.2 Biological functions of polyploidy
The fact that physiological polyploidy is observed in organisms ranging from the unicellular protists to eukaryotes as complex as metazoans suggests an evolutionary significance to this process. Many biological advantages of being polyploidy over diploid have been proposed, including an augmented response to strenuous cellular metabolic demands, buffering against environmental, metabolic-derived or replication-associated DNA damage, dampened sensitivity to apoptotic stimuli such as telomeric attrition and increased p53-p21 checkpoint function and evolutionary functions (Lee et al., 2009; Otto and Whitton, 2000; Lilly and
Spradling, 1996; Lazzerini Denchi et al., 2006; Mehrotra et al., 2008). On the other hand, it is observed that the higher the organism is in the evolutionary hierarchy, the lesser is the incidence of polyploidy. This clearly brings up an alternative hypothesis as to whether this physiological process is just an evolutionary vestige for higher organisms. It is believed that polyploidy in mammals, as in plants and flies, is integral to the biological functions of endocycling cells.
Surprisingly, we show here that TGCs, hepatocytes and megakaryocytes with severe reductions in ploidy can nonetheless bestow placentas, livers and platelets, respectively, with apparently normal physiological function. Though it could just be that the parameters that we have chosen
184 in this study were insufficient to identify significance of polyploidy, our studies clearly question the physiological role of polyploidy in mammals. Further studies are required to identify the role of polyploidy in mammals.
10.3 E2f7/E2f8: Tumor suppressor or oncogene
The deregulated expression of E2F7 and E2F8 has been observed in several human cancers and with known role as a transcriptional repressors of genes related to cell cycle progression, they are believed to contribute to the uncontrolled proliferation potential of malignant cells (Reimer et al., 2006, 2007; Endo-Munoz et al., 2009; Deng et al., 2010; Hazar-
Rethinam et al., 2011; de Bruin et al., 2003; Maiti et al., 2005). Atypical E2F repressors also have known functions in checkpoint regulation and DNA damage response (Zalmas et al., 2008).
Loss of function of E2F7 and E2F8 could thus directly lead to carcinogenesis by impairing ability of cells to repair DNA damage, especially in tissues like liver with high exposure to DNA damaging agents. In addition, their ability to suppress G2-M phase of cell cycle and cause polyploidization in developmentally programmed tissues add to their ability to act as tumor suppressors. For example, in mouse models a direct negative correlation between degree of polyploidization and proliferative capacity of hepatocytes was observed in the past (Saeter et al.,
1988; Schwarze et al., 1991 and Gerlyng et al., 1992). Moreover, diploid cells are considered less protected against recessive mutations than polyploid cells (Schwarze et al., 1984) owing to the lower buffering ability, with polyploidy providing additional copies of functional wild type alleles (Otto and Whitton, 2000). Our unpublished observations support this notion and show that double mutant livers with diploid hepatocytes develop both spontaneous and chemical-
185 induced HCC with much more severity than wild type livers composed of hepatocytes with polyploid genomes.
Though it is not difficult to imagine how E2F7 and E2F8 through the above mentioned functions might act as potent tumor suppressors, in reality E2F repressors as a whole have not been frequently found mutated, deleted or their expression silenced in human cancers (Chen et al., 2009a). Instead, quite frequently, increased copy numbers through amplification (E2F5 in breast cancer) as well as overexpression (E2F8 in primary liver and ovarian cancers) have been reported in the past (Polanowska et al., 2000; Umemura et al., 2009; Deng et al., 2010). Several in vitro studies have shown relation of polyploidy to genomic instability with evidence that newly formed polyploid cells display a higher rate of chromosome loss (Storchova and Pellman,
2004). On the same note, based on the observation of polyploidy (genome multiplication beyond four copies) in precancerous lesions of uterine cervix, vocal cord and breast, as well as in tumors like renal oncocytoma, pheochromocytoma, parathyroid and adrenal adenomas and thyroid
Hurthle cell oncocytoma, it has been proposed that depolyploidization (genome reduction cell division) of polyploid intermediates could lead to aneuploidy and further progression of cancer
(Biesterfeld et al., 1994; Storchova and Pellman, 2004). Multiple experimental models in mouse and Drosophila have provided evidence the high error rates seen during „ploidy reversal‟ divisions (Duncan et al., 2010; Duncan et al., 2009; Fox et al., 2010). Thus one could argue that the switch to endocycles during development is not irreversible, but when happens has deleterious consequence and genomic instability. Therefore, though polyploidy in cells protects against consequences of DNA damage, it also leads to a potential scenario of generation of genetically unstable „intermediates‟ that are predisposed to further progression towards malignancy. Thus atypical repressors due to their varied functions could potentially function as
186 either an oncogene or a tumor suppressor based on the context. The explorative screens in the past have not always looked for changes in E2F7 or E2F8 as these E2F family members are relatively newly identified. One would expect more information to come into literature about incidence of misregulation of these E2F factors in context of tumorigenesis more frequently in future.
In nutshell, by using targeted gene ablation strategies in mice we have shown that E2F7 and E2F8 play physiologically relevant role as late S-G2 repressors of E2F-target genes and that these functions are important in promoting TGC, megakaryocyte and hepatocyte polyploidy.
Though we found that reduction in ploidy did not have significant consequences on normal physiological functions of these cells, our preliminary results suggest that these defects could have significant impact on vulnerability of hepatocytes for tumorigenesis opening up a new arena for future studies.
187
Bibliography
Ambrus, A.M., Nicolay, B.N., Rasheva, V.I., Suckling, R.J., and Frolov, M.V. (2007). dE2F2-independent rescue of proliferation in cells lacking an activator dE2F1. Molecular and Cellular Biology 27, 8561-8570.
Andreassen, P., Martineau, S., and Margolis, R. (1996). Chemical induction of mitotic checkpoint override in mammalian cells results in aneuploidy following a transient tetraploid state. Mutat Res 372, 181-194.
Arias, E.E., and Walter, J.C. (2006). PCNA functions as a molecular platform to trigger Cdt1 destruction and prevent re-replication. Nat Cell Biol 8, 84-90.
Arias, E.E., and Walter, J.C. (2007). Strength in numbers: preventing rereplication via multiple mechanisms in eukaryotic cells. Genes Dev 21, 497-518.
Asano, M. (2009). Endoreplication: the advantage to initiating DNA replication without the ORC? Fly (Austin) 3, 173-175.
Attwooll, C., Lazzerini Denchi, E., and Helin, K. (2004). The E2F family: specific functions and overlapping interests. EMBO J 23, 4709-4716.
Awad, M.M., Sanders, J.A., and Gruppuso, P.A. (2000). A potential role for p15(Ink4b) and p57(Kip2) in liver development. FEBS Lett 483, 160-164.
Bermejo, R., Vilaboa, N., and Cales, C. (2002). Regulation of CDC6, geminin, and CDT1 in human cells that undergo polyploidization. Mol Biol Cell 13, 3989-4000.
Biesterfeld, S., Gerres, K., Fischer-Wein, G., and Böcking, A. (1994). Polyploidy in non-neoplastic tissues. J Clin Pathol 47, 38-42.
Blais, A., and Dynlacht, B.D. (2007). E2F-associated chromatin modifiers and cell cycle control. Curr Opin Cell Biol 19, 658-662.
Blow, J.J., and Hodgson, B. (2002). Replication licensing--defining the proliferative state? Trends Cell Biol 12, 72-78.
Boudolf, V., Lammens, T., Boruc, J., Van Leene, J., Van Den Daele, H., Maes, S., Van Isterdael, G., Russinova, E., Kondorosi, E., Witters, E., et al. (2009). CDKB1;1 forms a functional complex with CYCA2;3 to suppress endocycle onset. Plant Physiol 150, 1482-1493.
188
Boudolf, V., Vlieghe, K., Beemster, G.T., Magyar, Z., Torres Acosta, J.A., Maes, S., Van Der Schueren, E., Inze, D., and De Veylder, L. (2004). The plant-specific cyclin-dependent kinase CDKB1;1 and transcription factor E2Fa-DPa control the balance of mitotically dividing and endoreduplicating cells in Arabidopsis. Plant Cell 16, 2683-2692.
Calvi, B.R., Lilly, M.A., and Spradling, A.C. (1998). Cell cycle control of chorion gene amplification. Genes Dev 12, 734-744.
Calvi, B.R., and Spradling, A.C. (1999). Chorion gene amplification in Drosophila: A model for metazoan origins of DNA replication and S-phase control. Methods 18, 407-417.
Campanero, M.R., and Flemington, E.K. (1997). Regulation of E2F through ubiquitin-proteasome- dependent degradation: stabilization by the pRB tumor suppressor protein. Proc Natl Acad Sci U S A 94, 2221-2226.
Cao, A.R., Rabinovich, R., Xu, M., Xu, X., Jin, V.X., and Farnham, P.J. (2011). Genome-wide analysis of transcription factor E2F1 mutant proteins reveals that N- and C-terminal protein interaction domains do not participate in targeting E2F1 to the human genome. J Biol Chem 286, 11985-11996.
Castellano, M.M., del Pozo, J.C., Ramirez-Parra, E., Brown, S., and Gutierrez, C. (2001). Expression and stability of Arabidopsis CDC6 are associated with endoreplication. Plant Cell 13, 2671-2686.
Cayirlioglu, P., Ward, W.O., Key, S.C.S., and Duronio, R.J. (2003). Transcriptional. repressor functions of Drosophila E2F1 and E2F2 cooperate to inhibit genomic DNA synthesis in ovarian follicle cells. Molecular and Cellular Biology 23, 2123-2134.
Cebolla, A., Vinardell, J.M., Kiss, E., Olah, B., Roudier, F., Kondorosi, A., and Kondorosi, E. (1999). The mitotic inhibitor ccs52 is required for endoreduplication and ploidy-dependent cell enlargement in plants. EMBO J 18, 4476-4484.
Ceol, C.J., and Horvitz, H.R. (2001). dpl-1 DP and efl-1 E2F act with lin-35 Rb to antagonize Ras signaling in C. elegans vulval development. Mol Cell 7, 461-473.
Chen, HZ., Tsai, SY. and Leone, G. (2009). Emerging roles of E2Fs in cancer: an exit from cell cycle control. Nat Rev Cancer 9(11), 785-797.
Chen, D., Pacal, M., Wenzel, P., Knoepfler, P.S., Leone, G., and Bremner, R. (2009). Division and apoptosis of E2f-deficient retinal progenitors. Nature 462, 925-929.
Chong, J.L., Tsai, S.Y., Sharma, N., Opavsky, R., Price, R., Wu, L., Fernandez, S.A., and Leone, G. (2009a). E2f3a and E2f3b contribute to the control of cell proliferation and mouse development. Mol Cell Biol 29, 414-424.
Chong, J.L., Wenzel, P.L., Saenz-Robles, M.T., Nair, V., Ferrey, A., Hagan, J.P., Gomez, Y.M., Sharma, N., Chen, H.Z., Ouseph, M., et al. (2009b). E2f1-3 switch from activators in progenitor cells to repressors in differentiating cells. Nature 462, 930-934.
189
Christensen, J., Cloos, P., Toftegaard, U., Klinkenberg, D., Bracken, A.P., Trinh, E., Heeran, M., Di Stefano, L., and Helin, K. (2005). Characterization of E2F8, a novel E2F-like cell-cycle regulated repressor of E2F- activated transcription. Nucleic Acids Res 33, 5458-5470.
Christensen, R.N., Weinstein, M., and Tassava, R.A. (2002). Expression of fibroblast growth factors 4, 8, and 10 in limbs, flanks, and blastemas of Ambystoma. Dev Dyn 223, 193-203.
Claycomb, J.M., Benasutti, M., Bosco, G., Fenger, D.D., and Orr-Weaver, T.L. (2004). Gene amplification as a developmental strategy: isolation of two developmental amplicons in Drosophila. Dev Cell 6, 145- 155.
Claycomb, J.M., and Orr-Weaver, T.L. (2005). Developmental gene amplification: insights into DNA replication and gene expression. Trends Genet 21, 149-162.
Cloud, J.E., Rogers, C., Reza, T.L., Ziebold, U., Stone, J.R., Picard, M.H., Caron, A.M., Bronson, R.T., and Lees, J.A. (2002). Mutant mouse models reveal the relative roles of E2F1 and E2F3 in vivo. Molecular and Cellular Biology 22, 2663-2672.
Coan, P.M., Conroy, N., Burton, G.J., and Ferguson-Smith, A.C. (2006). Origin and characteristics of glycogen cells in the developing murine placenta. Dev Dyn 235, 3280-3294.
Cookson, S.J., Radziejwoski, A., and Granier, C. (2006). Cell and leaf size plasticity in Arabidopsis: what is the role of endoreduplication? Plant Cell Environ 29, 1273-1283.
Coverley, D., Laman, H., and Laskey, R.A. (2002). Distinct roles for cyclins E and A during DNA replication complex assembly and activation. Nat Cell Biol 4, 523-528.
Cross, J. (2005). How to make a placenta: mechanisms of trophoblast cell differentiation in mice--a review. Placenta 26 Suppl A, S3-9.
Cross, J.C. (2000). Genetic insights into trophoblast differentiation and placental morphogenesis. Semin Cell Dev Biol 11, 105-113.
Cross, J.C., Baczyk, D., Dobric, N., Hemberger, M., Hughes, M., Simmons, D.G., Yamamoto, H., and Kingdom, J.C. (2003). Genes, development and evolution of the placenta. Placenta 24, 123-130.
Croxton, R., Ma, Y., and Cress, W.D. (2002). Differences in DNA binding properties between E2F1 and E2F4 specify repression of the Mcl-1 promoter. Oncogene 21, 1563-1570.
Danielian, P.S., Friesenhahn, L.B., Faust, A.M., West, J.C., Caron, A.M., Bronson, R.T., and Lees, J.A. (2008). E2f3a and E2f3b make overlapping but different contributions to total E2f3 activity. Oncogene 27, 6561-6570.
Datta, N.S., Williams, J.L., Caldwell, J., Curry, A.M., Ashcraft, E.K., and Long, M.W. (1996). Novel alterations in CDK1/cyclin B1 kinase complex formation occur during the acquisition of a polyploid DNA content. Molecular Biology of the Cell 7, 209-223.
190
Datta, N.S., Williams, J.L., and Long, M.W. (1998). Differential modulation of G1-S-phase cyclin- dependent kinase 2/cyclin complexes occurs during the acquisition of a polyploid DNA content. Cell growth & differentiation : the molecular biology journal of the American Association for Cancer Research 9, 639-650. de Bruin, A., Maiti, B., Jakoi, L., Timmers, C., Buerki, R., and Leone, G. (2003). Identification and characterization of E2F7, a novel mammalian E2F family member capable of blocking cellular proliferation. J Biol Chem 278, 42041-42049. de Sauvage, F.J., Hass, P.E., Spencer, S.D., Malloy, B.E., Gurney, A.L., Spencer, S.A., Darbonne, W.C., Henzel, W.J., Wong, S.C., Kuang, W.J., et al. (1994). Stimulation of megakaryocytopoiesis and thrombopoiesis by the c-Mpl ligand. Nature 369, 533-538.
DeGregori, J. (2002). The genetics of the E2F family of transcription factors: shared functions and unique roles. Biochim Biophys Acta 1602, 131-150.
DeGregori, J. (2006). Surprising dependency for retinoblastoma protein in ras-mediated tumorigenesis. Mol Cell Biol 26, 1165-1169.
DeGregori, J., and Johnson, D. (2006a). Distinct and Overlapping Roles for E2F Family Members in Transcription, Proliferation and Apoptosis. Curr Mol Med 6, 739-748.
DeGregori, J., and Johnson, D.G. (2006b). Distinct and Overlapping Roles for E2F Family Members in Transcription, Proliferation and Apoptosis. Curr Mol Med 6, 739-748. del Pozo, J.C., Diaz-Trivino, S., Cisneros, N., and Gutierrez, C. (2006). The balance between cell division and endoreplication depends on E2FC-DPB, transcription factors regulated by the ubiquitin-SCFSKP2A pathway in Arabidopsis. Plant Cell 18, 2224-2235.
Deng, W.M., Althauser, C., and Ruohola-Baker, H. (2001). Notch-Delta signaling induces a transition from mitotic cell cycle to endocycle in Drosophila follicle cells. Development 128, 4737-4746.
DePamphilis, M.L., Blow, J.J., Ghosh, S., Saha, T., Noguchi, K., and Vassilev, A. (2006). Regulating the licensing of DNA replication origins in metazoa. Curr Opin Cell Biol 18, 231-239.
Di Stefano, L., Jensen, M.R., and Helin, K. (2003). E2F7, a novel E2F featuring DP-independent repression of a subset of E2F-regulated genes. EMBO J 22, 6289-6298.
Diaz-Trivino, S., del Mar Castellano, M., de la Paz Sanchez, M., Ramirez-Parra, E., Desvoyes, B., and Gutierrez, C. (2005). The genes encoding Arabidopsis ORC subunits are E2F targets and the two ORC1 genes are differently expressed in proliferating and endoreplicating cells. Nucleic Acids Res 33, 5404- 5414.
Dimova, D.K., and Dyson, N.J. (2005). The E2F transcriptional network: old acquaintances with new faces. Oncogene 24, 2810-2826.
Dimova, D.K., Stevaux, O., Frolov, M.V., and Dyson, N.J. (2003). Cell cycle-dependent and cell cycle- independent control of transcription by the Drosophila E2F/RB pathway. Genes Dev 17, 2308-2320.
191
Doubeikovski, A., Uzan, G., Doubeikovski, Z., Prandini, M.H., Porteu, F., Gisselbrecht, S., and Dusanter- Fourt, I. (1997). Thrombopoietin-induced expression of the glycoprotein IIb gene involves the transcription factor PU.1/Spi-1 in UT7-Mpl cells. J Biol Chem 272, 24300-24307.
Du, W. (2000). Suppression of the rbf null mutants by a de2f1 allele that lacks transactivation domain. Development 127, 367-379.
Du, W., and Dyson, N. (1999). The role of RBF in the introduction of G1 regulation during Drosophila embryogenesis. EMBO J 18, 916-925.
Du, W., Vidal, M., Xie, J.E., and Dyson, N. (1996). RBF, a novel RB-related gene that regulates E2F activity and interacts with cyclin E in Drosophila. Genes Dev 10, 1206-1218.
Duronio, R.J., Bonnette, P.C., and O'Farrell, P.H. (1998). Mutations of the Drosophila dDP, dE2F, and cyclin E genes reveal distinct roles for the E2F-DP transcription factor and cyclin E during the G1-S transition. Mol Cell Biol 18, 141-151.
Duronio, R.J., Brook, A., Dyson, N., and O'Farrell, P.H. (1996). E2F-induced S phase requires cyclin E. Genes Dev 10, 2505-2513.
Duronio, R.J., and O'Farrell, P.H. (1995). Developmental control of the G1 to S transition in Drosophila: cyclin Eis a limiting downstream target of E2F. Genes Dev 9, 1456-1468.
Duronio, R.J., O'Farrell, P.H., Xie, J.E., Brook, A., and Dyson, N. (1995). The transcription factor E2F is required for S phase during Drosophila embryogenesis. Genes Dev 9, 1445-1455.
Edgar, B.A., and Orr-Weaver, T.L. (2001). Endoreplication cell cycles: more for less. Cell 105, 297-306. Erenpreisa, J., Kalejs, M., and Cragg, M.S. (2005a). Mitotic catastrophe and endomitosis in tumour cells: an evolutionary key to a molecular solution. Cell Biol Int 29, 1012-1018.
Erenpreisa, J., Kalejs, M., Ianzini, F., Kosmacek, E.A., Mackey, M.A., Emzinsh, D., Cragg, M.S., Ivanov, A., and Illidge, T.M. (2005b). Segregation of genomes in polyploid tumour cells following mitotic catastrophe. Cell Biol Int 29, 1005-1011.
Field, S.J., Tsai, F.Y., Kuo, F., Zubiaga, A.M., Kaelin, W.G., Jr., Livingston, D.M., Orkin, S.H., and Greenberg, M.E. (1996). E2F-1 functions in mice to promote apoptosis and suppress proliferation. Cell 85, 549-561.
Follette, P.J., Duronio, R.J., and O'Farrell, P.H. (1998). Fluctuations in cyclin E levels are required for multiple rounds of endocycle S phase in Drosophila. Curr Biol 8, 235-238.
Frolov, M.V., and Dyson, N.J. (2004). Molecular mechanisms of E2F-dependent activation and pRB- mediated repression. J Cell Sci 117, 2173-2181.
Frolov, M.V., Huen, D.S., Stevaux, O., Dimova, D., Balczarek-Strang, K., Elsdon, M., and Dyson, N.J. (2001). Functional antagonism between E2F family members. Genes Dev 15, 2146-2160.
192
Frolov, M.V., Stevaux, O., Moon, N.S., Dimova, D., Kwon, E.J., Morris, E.J., and Dyson, N.J. (2003). G1 cyclin-dependent kinases are insufficient to reverse dE2F2-mediated repression. Gene Dev 17, 723-728.
Garcia-Higuera, I., Manchado, E., Dubus, P., Canamero, M., Mendez, J., Moreno, S., and Malumbres, M. (2008). Genomic stability and tumour suppression by the APC/C cofactor Cdh1. Nat Cell Biol 10, 802-811.
Garcia, P., and Cales, C. (1996). Endoreplication in megakaryoblastic cell lines is accompanied by sustained expression of G1/S cyclins and downregulation of cdc25C. Oncogene 13, 695-703.
Garcia, P., Frampton, J., Ballester, A., and Cales, C. (2000). Ectopic expression of cyclin E allows non- endomitotic megakaryoblastic K562 cells to establish re-replication cycles. Oncogene 19, 1820-1833.
Gaubatz, S., Lindeman, G.J., Ishida, S., Jakoi, L., Nevins, J.R., Livingston, D.M., and Rempel, R.E. (2000). E2F4 and E2F5 play an essential role in pocket protein-mediated G1 control. Mol Cell 6, 729-735.
Geddis, A.E., and Kaushansky, K. (2004). Inherited thrombocytopenias: toward a molecular understanding of disorders of platelet production. Current opinion in pediatrics 16, 15-22.
Geng, Y., Yu, Q., Sicinska, E., Das, M., Schneider, J.E., Bhattacharya, S., Rideout, W.M., Bronson, R.T., Gardner, H., and Sicinski, P. (2003). Cyclin E ablation in the mouse. Cell 114, 431-443.
Giangrande, P.H., Zhang, J., Tanner, A., Eckhart, A.D., Rempel, R.E., Andrechek, E.R., Layzer, J.M., Keys, J.R., Hagen, P.O., Nevins, J.R., et al. (2007). Distinct roles of E2F proteins in vascular smooth muscle cell proliferation and intimal hyperplasia. Proc Natl Acad Sci U S A 104, 12988-12993.
Giangrande, P.H., Zhu, W., Schlisio, S., Sun, X., Mori, S., Gaubatz, S., and Nevins, J.R. (2004). A role for E2F6 in distinguishing G1/S- and G2/M-specific transcription. Genes Dev 18, 2941-2951.
Grafi, G., and Larkins, B.A. (1995). Endoreduplication in Maize Endosperm: Involvement of M Phase-- Promoting Factor Inhibition and Induction of S Phase--Related Kinases. Science 269, 1262-1264.
Gu, X.F., Allain, A., Li, L., Cramer, E.M., Tenza, D., Caen, J.P., and Han, Z.C. (1993). Expression of Cyclin-B in Megakaryocytes and Cells of Other Hematopoietic Lineages. Cr Acad Sci Iii-Vie 316, 1438-1445.
Guy, C.T., Zhou, W., Kaufman, S., and Robinson, M.O. (1996). E2F-1 blocks terminal differentiation and causes proliferation in transgenic megakaryocytes. Molecular and cellular biology 16, 685-693.
Hancock, V., Martin, J.F., and Lelchuk, R. (1993). The relationship between human megakaryocyte nuclear DNA content and gene expression. British journal of haematology 85, 692-697.
Handeli, S., and Weintraub, H. (1992). The Ts41 Mutation in Chinese-Hamster Cells Leads to Successive S Phases in the Absence of Intervening G2, M, and G1. Cell 71, 599-611.
Hardie, D.C., Gregory, T.R., and Hebert, P.D. (2002). From pixels to picograms: a beginners' guide to genome quantification by Feulgen image analysis densitometry. J Histochem Cytochem 50, 735-749.
193
Hart, A., Melet, F., Grossfeld, P., Chien, K., Jones, C., Tunnacliffe, A., Favier, R., and Bernstein, A. (2000). Fli-1 is required for murine vascular and megakaryocytic development and is hemizygously deleted in patients with thrombocytopenia. Immunity 13, 167-177.
Hauck, L., Kaba, R.G., Lipp, M., Dietz, R., and von Harsdorf, R. (2002). Regulation of E2F1-dependent gene transcription and apoptosis by the ETS-related transcription factor GABPgamma1. Mol Cell Biol 22, 2147-2158.
Havens, C.G., and Walter, J.C. (2009). Docking of a specialized PIP Box onto chromatin-bound PCNA creates a degron for the ubiquitin ligase CRL4Cdt2. Mol Cell 35, 93-104.
Hayashi, S., Lewis, P., Pevny, L., and McMahon, A.P. (2002). Efficient gene modulation in mouse epiblast using a Sox2Cre transgenic mouse strain. Mech Dev 119 Suppl 1, S97-S101.
Heinemeyer, T., Wingender, E., Reuter, I., Hermjakob, H., Kel, A.E., Kel, O.V., Ignatieva, E.V., Ananko, E.A., Podkolodnaya, O.A., Kolpakov, F.A., et al. (1998). Databases on transcriptional regulation: TRANSFAC, TRRD and COMPEL. Nucleic Acids Res 26, 362-367.
Helin, K., Lees, J.A., Vidal, M., Dyson, N., Harlow, E., and Fattaey, A. (1992). A cDNA encoding a pRB- binding protein with properties of the transcription factor E2F. Cell 70, 337-350.
Higa, L.A., Banks, D., Wu, M., Kobayashi, R., Sun, H., and Zhang, H. (2006). L2DTL/CDT2 interacts with the CUL4/DDB1 complex and PCNA and regulates CDT1 proteolysis in response to DNA damage. Cell Cycle 5, 1675-1680.
Hodohara, K., Fujii, N., Yamamoto, N., and Kaushansky, K. (2000). Stromal cell-derived factor-1 (SDF-1) acts together with thrombopoietin to enhance the development of megakaryocytic progenitor cells (CFU-MK). Blood 95, 769-775.
Hoffman, R. (1989). Regulation of megakaryocytopoiesis. Blood 74, 1196-1212.
Hu, J., and Xiong, Y. (2006). An evolutionarily conserved function of proliferating cell nuclear antigen for Cdt1 degradation by the Cul4-Ddb1 ubiquitin ligase in response to DNA damage. J Biol Chem 281, 3753- 3756.
Humbert, P.O., Verona, R., Trimarchi, J.M., Rogers, C., Dandapani, S., and Lees, J.A. (2000). E2f3 is critical for normal cellular proliferation. Genes Dev 14, 690-703.
Ishida, S., Huang, E., Zuzan, H., Spang, R., Leone, G., West, M., and Nevins, J.R. (2001). Role for E2F in control of both DNA replication and mitotic functions as revealed from DNA microarray analysis. Mol Cell Biol 21, 4684-4699.
Italiano, J.E., Jr., Lecine, P., Shivdasani, R.A., and Hartwig, J.H. (1999). Blood platelets are assembled principally at the ends of proplatelet processes produced by differentiated megakaryocytes. J Cell Biol 147, 1299-1312.
194
Ivanov, A., Cragg, M.S., Erenpreisa, J., Emzinsh, D., Lukman, H., and Illidge, T.M. (2003). Endopolyploid cells produced after severe genotoxic damage have the potential to repair DNA double strand breaks. J Cell Sci 116, 4095-4106.
Jackson, C., Arnold, J., Pestina, T., and Stenberg, P. (1997). Megakaryocyte Biology. In Thrombopoiesis and thrombopoietins: molecular, cellular, preclinical, and clinical biology, D. Kuter, P. Hunt, W. Sheridan, and D. Zucker-Franklin, eds. (Totowa, NJ: Humana Press, Inc.), pp. 3-30.
Jackson, P.K., Chevalier, S., Philippe, M., and Kirschner, M.W. (1995). Early events in DNA replication require cyclin E and are blocked by p21CIP1. J Cell Biol 130, 755-769.
Jakoby, M., and Schnittger, A. (2004). Cell cycle and differentiation. Curr Opin Plant Biol 7, 661-669. Johnson, D.G., and Degregori, J. (2006). Putting the Oncogenic and Tumor Suppressive Activities of E2F into Context. Curr Mol Med 6, 731-738.
Johnson, D.G., Schwarz, J.K., Cress, W.D., and Nevins, J.R. (1993). Expression of transcription factor E2F1 induces quiescent cells to enter S phase. Nature 365, 349-352.
Jordan, K.C., Schaeffer, V., Fischer, K.A., Gray, E.E., and Ruohola-Baker, H. (2006). Notch signaling through tramtrack bypasses the mitosis promoting activity of the JNK pathway in the mitotic-to- endocycle transition of Drosophila follicle cells. BMC Dev Biol 6, 16.
Kaelin, W.G., Jr., Krek, W., Sellers, W.R., DeCaprio, J.A., Ajchenbaum, F., Fuchs, C.S., Chittenden, T., Li, Y., Farnham, P.J., Blanar, M.A., et al. (1992). Expression cloning of a cDNA encoding a retinoblastoma- binding protein with E2F-like properties. Cell 70, 351-364.
Kalejs, M., Ivanov, A., Plakhins, G., Cragg, M.S., Emzinsh, D., Illidge, T.M., and Erenpreisa, J. (2006). Upregulation of meiosis-specific genes in lymphoma cell lines following genotoxic insult and induction of mitotic catastrophe. BMC Cancer 6, 6.
Kaushansky, K. (2008). Historical review: megakaryopoiesis and thrombopoiesis. Blood 111, 981-986.
Kawasaki, A., Matsumura, I., Miyagawa, J., Ezoe, S., Tanaka, H., Terada, Y., Tatsuka, M., Machii, T., Miyazaki, H., Furukawa, Y., et al. (2001). Downregulation of an AIM-1 kinase couples with megakaryocytic polyploidization of human hematopoietic cells. The Journal of cell biology 152, 275-287.
Kinross, K.M., Clark, A.J., Iazzolino, R.M., and Humbert, P.O. (2006). E2f4 regulates fetal erythropoiesis through the promotion of cellular proliferation. Blood 108, 886-895.
Klapholz, B., Dietrich, B.H., Schaffner, C., Heredia, F., Quivy, J.P., Almouzni, G., and Dostatni, N. (2009). CAF-1 is required for efficient replication of euchromatic DNA in Drosophila larval endocycling cells. Chromosoma 118, 235-248.
Kohn, M.J., Bronson, R.T., Harlow, E., Dyson, N.J., and Yamasaki, L. (2003). Dp1 is required for extra- embryonic development. Development 130, 1295-1305.
Kohn, M.J., Leung, S.W., Criniti, V., Agromayor, M., and Yamasaki, L. (2004). Dp1 is largely dispensable for embryonic development. Mol Cell Biol 24, 7197-7205.
195
Kondorosi, E., and Kondorosi, A. (2004). Endoreduplication and activation of the anaphase-promoting complex during symbiotic cell development. FEBS Lett 567, 152-157.
Kosugi, S., and Ohashi, Y. (2002). E2Ls, E2F-like repressors of Arabidopsis that bind to E2F sites in a monomeric form. J Biol Chem 277, 16553-16558.
Kovesdi, I., Reichel, R., and Nevins, J.R. (1986). Identification of a cellular transcription factor involved in E1A trans-activation. Cell 45, 219-228.
Kozar, K., Ciemerych, M.A., Rebel, V.I., Shigematsu, H., Zagozdzon, A., Sicinska, E., Geng, Y., Yu, Q., Bhattacharya, S., Bronson, R.T., et al. (2004). Mouse development and cell proliferation in the absence of D-cyclins. Cell 118, 477-491.
Krek, W., Xu, G., and Livingston, D.M. (1995). Cyclin A-kinase regulation of E2F-1 DNA binding function underlies suppression of an S phase checkpoint. Cell 83, 1149-1158.
Kubota, K., Tamauchi, H., and Nakazato, K. (1996). Deregulated c-myc expression is insufficient for emergence of the tumorigenic phenotype in malignancy-suppressed intratypic T-cell lymphoma hybrids: an in vitro model for multistep lymphomagenesis. Cancer letters 101, 199-203.
Kuter, D.J., Beeler, D.L., and Rosenberg, R.D. (1994). The purification of megapoietin: a physiological regulator of megakaryocyte growth and platelet production. Proc Natl Acad Sci U S A 91, 11104-11108.
Lammens, T., Boudolf, V., Kheibarshekan, L., Zalmas, L., Gaamouche, T., Maes, S., Vanstraelen, M., Kondorosi, E., La Thangue, N., Govaerts, W., et al. (2008a). Atypical E2F activity restrains APC/CCCS52A2 function obligatory for endocycle onset. Proc Natl Acad Sci U S A 105, 14721-14726.
Lammens, T., Boudolf, V., Kheibarshekan, L., Zalmas, L.P., Gaamouche, T., Maes, S., Vanstraelen, M., Kondorosi, E., La Thangue, N.B., Govaerts, W., et al. (2008b). Atypical E2F activity restrains APC/CCCS52A2 function obligatory for endocycle onset. Proc Natl Acad Sci U S A 105, 14721-14726.
Lammens, T., Li, J., Leone, G., and De Veylder, L. (2009). Atypical E2Fs: new players in the E2F transcription factor family. Trends Cell Biol 19, 111-118.
Lazzerini Denchi, E., Celli, G., and de Lange, T. (2006). Hepatocytes with extensive telomere deprotection and fusion remain viable and regenerate liver mass through endoreduplication. Genes Dev 20, 2648- 2653.
Lee, H.O., Davidson, J.M., and Duronio, R.J. (2009). Endoreplication: polyploidy with purpose. Genes Dev 23, 2461-2477.
Leiva-Neto, J.T., Grafi, G., Sabelli, P.A., Dante, R.A., Woo, Y.M., Maddock, S., Gordon-Kamm, W.J., and Larkins, B.A. (2004). A dominant negative mutant of cyclin-dependent kinase A reduces endoreduplication but not cell size or gene expression in maize endosperm. Plant Cell 16, 1854-1869.
Lengronne, A., and Schwob, E. (2002). The yeast CDK inhibitor Sic1 prevents genomic instability by promoting replication origin licensing in late G(1). Mol Cell 9, 1067-1078.
196
Leone, G., DeGregori, J., Yan, Z., Jakoi, L., Ishida, S., Williams, R.S., and Nevins, J.R. (1998). E2F3 activity is regulated during the cell cycle and is required for the induction of S phase. Genes Dev 12, 2120-2130.
Li, F.X., Zhu, J.W., Tessem, J.S., Beilke, J., Varella-Garcia, M., Jensen, J., Hogan, C.J., and DeGregori, J. (2003). The development of diabetes in E2f1/E2f2 mutant mice reveals important roles for bone marrow-derived cells in preventing islet cell loss. Proc Natl Acad Sci U S A 100, 12935-12940.
Li, J., Ran, C., Li, E., Gordon, F., Comstock, G., Siddiqui, H., Cleghorn, W., Chen, H.Z., Kornacker, K., Liu, C.G., et al. (2008a). Synergistic function of E2F7 and E2F8 is essential for cell survival and embryonic development. Developmental cell 14, 62-75.
Li, J., Ran, C., Li, E., Gordon, F., Comstock, G., Siddiqui, H., Cleghorn, W., Chen, H.Z., Kornacker, K., Liu, C.G., et al. (2008b). Synergistic function of E2F7 and E2F8 is essential for cell survival and embryonic development. Developmental Cell 14, 62-75.
Lilly, M.A., and Spradling, A.C. (1996). The Drosophila endocycle is controlled by Cyclin E and lacks a checkpoint ensuring S-phase completion. Genes Dev 10, 2514-2526.
Lindeman, G.J., Dagnino, L., Gaubatz, S., Xu, Y.H., Bronson, R.T., Warren, H.B., and Livingston, D.M. (1998). A specific, nonproliferative role for E2F-5 in choroid plexus function revealed by gene targeting. Gene Dev 12, 1092-1098.
Liu, P., Jenkins, N., and Copeland, N. (2003). A highly efficient recombineering-based method for generating conditional knockout mutations. Genome Res 13, 476-484.
Logan, N., Graham, A., Zhao, X.J., Fisher, R., Maiti, B., Leone, G., and La Thangue, N.B. (2005). E2F-8: an E2F family member with a similar organization of DNA-binding domains to E2F-7. Oncogene 24, 5000- 5004.
Lok, S., Kaushansky, K., Holly, R.D., Kuijper, J.L., Lofton-Day, C.E., Oort, P.J., Grant, F.J., Heipel, M.D., Burkhead, S.K., Kramer, J.M., et al. (1994). Cloning and expression of murine thrombopoietin cDNA and stimulation of platelet production in vivo. Nature 369, 565-568.
Lopez-Schier, H., and St Johnston, D. (2001). Delta signaling from the germ line controls the proliferation and differentiation of the somatic follicle cells during Drosophila oogenesis. Genes Dev 15, 1393-1405.
Lozano, E., Saez, A.G., Flemming, A.J., Cunha, A., and Leroi, A.M. (2006). Regulation of growth by ploidy in Caenorhabditis elegans. Curr Biol 16, 493-498.
Maines, J.Z., Stevens, L.M., Tong, X., and Stein, D. (2004). Drosophila dMyc is required for ovary cell growth and endoreplication. Development 131, 775-786.
Maiti, B., Li, J., de Bruin, A., Gordon, F., Timmers, C., Opavsky, R., Patil, K., Tuttle, J., Cleghorn, W., and Leone, G. (2005a). Cloning and characterization of mouse E2F8, a novel mammalian E2F family member capable of blocking cellular proliferation. The Journal of biological chemistry 280, 18211-18220.
197
Maiti, B., Li, J., de Bruin, A., Gordon, F., Timmers, C., Opavsky, R., Patil, K., Tuttle, J., Cleghorn, W., and Leone, G. (2005b). Cloning and characterization of mouse E2F8, a novel mammalian E2F family member capable of blocking cellular proliferation. J Biol Chem 280, 18211-18220.
Malumbres, M., Sotillo, R., Santamaria, D., Galan, J., Cerezo, A., Ortega, S., Dubus, P., and Barbacid, M. (2004). Mammalian cells cycle without the D-type cyclin-dependent kinases Cdk4 and Cdk6. Cell 118, 493-504.
Margottin-Goguet, F., Hsu, J.Y., Loktev, A., Hsieh, H.M., Reimann, J.D., and Jackson, P.K. (2003). Prophase destruction of Emi1 by the SCF(betaTrCP/Slimb) ubiquitin ligase activates the anaphase promoting complex to allow progression beyond prometaphase. Dev Cell 4, 813-826.
Mariconti, L., Pellegrini, B., Cantoni, R., Stevens, R., Bergounioux, C., Cella, R., and Albani, D. (2002). The E2F family of transcription factors from Arabidopsis thaliana. Novel and conserved components of the retinoblastoma/E2F pathway in plants. J Biol Chem 277, 9911-9919.
Mayhew, C., Bosco, E., Fox, S., Okaya, T., Tarapore, P., Schwemberger, S., Babcock, G., Lentsch, A., Fukasawa, K., and Knudsen, E. (2005). Liver-specific pRB loss results in ectopic cell cycle entry and aberrant ploidy. Cancer Res 65, 4568-4577.
Meckert, P.C., Rivello, H.G., Vigliano, C., Gonzalez, P., Favaloro, R., and Laguens, R. (2005). Endomitosis and polyploidization of myocardial cells in the periphery of human acute myocardial infarction. Cardiovasc Res 67, 116-123.
Mehrotra, S., Maqbool, S.B., Kolpakas, A., Murnen, K., and Calvi, B.R. (2008). Endocycling cells do not apoptose in response to DNA rereplication genotoxic stress. Genes Dev 22, 3158-3171.
Methia, N., Louache, F., Vainchenker, W., and Wendling, F. (1993). Oligodeoxynucleotides antisense to the proto-oncogene c-mpl specifically inhibit in vitro megakaryocytopoiesis. Blood 82, 1395-1401.
Murga, M., Fernandez-Capetillo, O., Field, S.J., Moreno, B., Borlado, L.R., Fujiwara, Y., Balomenos, D., Vicario, A., Carrera, A.C., Orkin, S.H., et al. (2001). Mutation of E2F2 in mice causes enhanced T lymphocyte proliferation, leading to the development of autoimmunity. Immunity 15, 959-970.
Myers, T.R., and Greenwald, I. (2005). lin-35 Rb acts in the major hypodermis to oppose ras-mediated vulval induction in C. elegans. Dev Cell 8, 117-123.
Nagata, Y., Muro, Y., and Todokoro, K. (1997). Thrombopoietin-induced polyploidization of bone marrow megakaryocytes is due to a unique regulatory mechanism in late mitosis. The Journal of cell biology 139, 449-457.
Nakayama, H., Liu, Y., Stifani, S., and Cross, J.C. (1997). Developmental restriction of Mash-2 expression in trophoblast correlates with potential activation of the notch-2 pathway. Dev Genet 21, 21-30.
Narbonne-Reveau, K., Senger, S., Pal, M., Herr, A., Richardson, H.E., Asano, M., Deak, P., and Lilly, M.A. (2008). APC/CFzr/Cdh1 promotes cell cycle progression during the Drosophila endocycle. Development 135, 1451-1461.
198
Odell, T.T., Jackson, C.W., and Reiter, R.S. (1968). Generation Cycle of Rat Megakaryocytes. Experimental Cell Research 53, 321-&.
Odell, T.T., Jr., and Jackson, C.W. (1968). Polyploidy and maturation of rat megakaryocytes. Blood 32, 102-110.
Parisi, T., Beck, A.R., Rougier, N., McNeil, T., Lucian, L., Werb, Z., and Amati, B. (2003). Cyclins E1 and E2 are required for endoreplication in placental trophoblast giant cells. EMBO J 22, 4794-4803.
Park, J.A., Ahn, J.W., Kim, Y.K., Kim, S.J., Kim, J.K., Kim, W.T., and Pai, H.S. (2005). Retinoblastoma protein regulates cell proliferation, differentiation, and endoreduplication in plants. Plant J 42, 153-163.
Park, S.Y., and Asano, M. (2008). The origin recognition complex is dispensable for endoreplication in Drosophila. Proc Natl Acad Sci U S A 105, 12343-12348.
Patel, S.R., Hartwig, J.H., and Italiano, J.E., Jr. (2005). The biogenesis of platelets from megakaryocyte proplatelets. J Clin Invest 115, 3348-3354.
Pesin, J.A., and Orr-Weaver, T.L. (2008). Regulation of APC/C activators in mitosis and meiosis. Annu Rev Cell Dev Biol 24, 475-499.
Pohlers, M., Truss, M., Frede, U., Scholz, A., Strehle, M., Kuban, R.J., Hoffmann, B., Morkel, M., Birchmeier, C., and Hagemeier, C. (2005). A role for E2F6 in the restriction of male-germ-cell-specific gene expression. Current Biology 15, 1051-1057.
Polager, S., and Ginsberg, D. (2008). E2F - at the crossroads of life and death. Trends Cell Biol 18, 528- 535.
Prieur-Carrillo, G., Chu, K., Lindqvist, J., and Dewey, W.C. (2003). Computerized video time-lapse (CVTL) analysis of the fate of giant cells produced by X-irradiating EJ30 human bladder carcinoma cells. Radiat Res 159, 705-712.
Puig, P.E., Guilly, M.N., Bouchot, A., Droin, N., Cathelin, D., Bouyer, F., Favier, L., Ghiringhelli, F., Kroemer, G., Solary, E., et al. (2008). Tumor cells can escape DNA-damaging cisplatin through DNA endoreduplication and reversible polyploidy. Cell Biol Int 32, 1031-1043.
Rabinovich, A., Jin, V.X., Rabinovich, R., Xu, X., and Farnham, P.J. (2008). E2F in vivo binding specificity: comparison of consensus versus nonconsensus binding sites. Genome Res 18, 1763-1777.
Radley, J.M., and Haller, C.J. (1983). Fate of senescent megakaryocytes in the bone marrow. Br J Haematol 53, 277-287.
Ravid, K., Li, Y.C., Rayburn, H.B., and Rosenberg, R.D. (1993). Targeted expression of a conditional oncogene in hematopoietic cells of transgenic mice. The Journal of cell biology 123, 1545-1553.
Ravid, K., Lu, J., Zimmet, J.M., and Jones, M.R. (2002). Roads to polyploidy: the megakaryocyte example. J Cell Physiol 190, 7-20.
199
Rawn, S.M., and Cross, J.C. (2008). The evolution, regulation, and function of placenta-specific genes. Annu Rev Cell Dev Biol 24, 159-181.
Regenstreif, L.J., and Rossant, J. (1989). Expression of the c-fms proto-oncogene and of the cytokine, CSF-1, during mouse embryogenesis. Dev Biol 133, 284-294.
Rempel, R.E., Saenz-Robles, M.T., Storms, R., Morham, S., Ishida, S., Engel, A., Jakoi, L., Melhem, M.F., Pipas, J.M., Smith, C., et al. (2000). Loss of E2F4 activity leads to abnormal development of multiple cellular lineages. Molecular Cell 6, 293-306.
Rinkenberger, J.L., Cross, J.C., and Werb, Z. (1997). Molecular genetics of implantation in the mouse. Dev Genet 21, 6-20.
Rossant, J., and Cross, J.C. (2001). Placental development: lessons from mouse mutants. Nat Rev Genet 2, 538-548.
Roth, G.J., Yagi, M., and Bastian, L.S. (1996). The platelet glycoprotein Ib-V-IX system: regulation of gene expression. Stem Cells 14 Suppl 1, 188-193.
Roy, L., Coullin, P., Vitrat, N., Hellio, R., Debili, N., Weinstein, J., Bernheim, A., and Vainchenker, W. (2001). Asymmetrical segregation of chromosomes with a normal metaphase/anaphase checkpoint in polyploid megakaryocytes. Blood 97, 2238-2247.
Royzman, I., Whittaker, A.J., and Orr-Weaver, T.L. (1997). Mutations in Drosophila DP and E2F distinguish G1-S progression from an associated transcriptional program. Genes Dev 11, 1999-2011.
Saavedra, H.I., Wu, L., de Bruin, A., Timmers, C., Rosol, T.J., Weinstein, M., Robinson, M.L., and Leone, G. (2002). Specificity of E2F1, E2F2, and E2F3 in mediating phenotypes induced by loss of Rb. Cell Growth Differ 13, 215-225.
Sauer, K., Knoblich, J.A., Richardson, H., and Lehner, C.F. (1995). Distinct modes of cyclin E/cdc2c kinase regulation and S-phase control in mitotic and endoreduplication cycles of Drosophila embryogenesis. Genes Dev 9, 1327-1339.
Schaeffer, V., Althauser, C., Shcherbata, H.R., Deng, W.M., and Ruohola-Baker, H. (2004). Notch- dependent Fizzy-related/Hec1/Cdh1 expression is required for the mitotic-to-endocycle transition in Drosophila follicle cells. Curr Biol 14, 630-636.
Schlisio, S., Halperin, T., Vidal, M., and Nevins, J.R. (2002). Interaction of YY1 with E2Fs, mediated by RYBP, provides a mechanism for specificity of E2F function. EMBO J 21, 5775-5786.
Senga, T., Sivaprasad, U., Zhu, W., Park, J.H., Arias, E.E., Walter, J.C., and Dutta, A. (2006). PCNA is a cofactor for Cdt1 degradation by CUL4/DDB1-mediated N-terminal ubiquitination. J Biol Chem 281, 6246-6252.
Sferruzzi-Perri, A.N., Macpherson, A.M., Roberts, C.T., and Robertson, S.A. (2009). Csf2 null mutation alters placental gene expression and trophoblast glycogen cell and giant cell abundance in mice. Biol Reprod 81, 207-221.
200
Shcherbata, H.R., Althauser, C., Findley, S.D., and Ruohola-Baker, H. (2004). The mitotic-to-endocycle switch in Drosophila follicle cells is executed by Notch-dependent regulation of G1/S, G2/M and M/G1 cell-cycle transitions. Development 131, 3169-3181.
Shibutani, S.T., de la Cruz, A.F., Tran, V., Turbyfill, W.J., 3rd, Reis, T., Edgar, B.A., and Duronio, R.J. (2008). Intrinsic negative cell cycle regulation provided by PIP box- and Cul4Cdt2-mediated destruction of E2f1 during S phase. Dev Cell 15, 890-900.
Shivdasani, R.A., Fujiwara, Y., McDevitt, M.A., and Orkin, S.H. (1997). A lineage-selective knockout establishes the critical role of transcription factor GATA-1 in megakaryocyte growth and platelet development. EMBO J 16, 3965-3973.
Simmons, D., Fortier, A., and Cross, J. (2007). Diverse subtypes and developmental origins of trophoblast giant cells in the mouse placenta. Dev Biol 304, 567-578.
Simmons, D.G., Rawn, S., Davies, A., Hughes, M., and Cross, J.C. (2008). Spatial and temporal expression of the 23 murine Prolactin/Placental Lactogen-related genes is not associated with their position in the locus. BMC Genomics 9, 352.
Sohma, Y., Akahori, H., Seki, N., Hori, T., Ogami, K., Kato, T., Shimada, Y., Kawamura, K., and Miyazaki, H. (1994). Molecular cloning and chromosomal localization of the human thrombopoietin gene. FEBS Lett 353, 57-61.
Solar, G.P., Kerr, W.G., Zeigler, F.C., Hess, D., Donahue, C., de Sauvage, F.J., and Eaton, D.L. (1998). Role of c-mpl in early hematopoiesis. Blood 92, 4-10.
Soriano, P. (1999). Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet 21, 70- 71.
Sozzani, R., Maggio, C., Varotto, S., Canova, S., Bergounioux, C., Albani, D., and Cella, R. (2006). Interplay between Arabidopsis activating factors E2Fb and E2Fa in cell cycle progression and development. Plant Physiol 140, 1355-1366.
Stenberg, P.E., and Levin, J. (1989). Mechanisms of platelet production. Blood cells 15, 23-47. Stevaux, O., Dimova, D.K., Ji, J.Y., Moon, N.S., Frolov, M.V., and Dyson, N.J. (2005). Retinoblastoma family 2 is required in vivo for the tissue-specific repression of dE2F2 target genes. Cell Cycle 4, 1272- 1280.
Storchova, Z., and Pellman, D. (2004). From polyploidy to aneuploidy, genome instability and cancer. Nat Rev Mol Cell Biol 5, 45-54.
Sun, J., Smith, L., Armento, A., and Deng, W.M. (2008). Regulation of the endocycle/gene amplification switch by Notch and ecdysone signaling. J Cell Biol 182, 885-896.
Sun, S., Zimmet, J.M., Toselli, P., Thompson, A., Jackson, C.W., and Ravid, K. (2001). Overexpression of cyclin D1 moderately increases ploidy in megakaryocytes. Haematologica 86, 17-23.
201
Susaki, E., Nakayama, K., Yamasaki, L., and Nakayama, K.I. (2009). Common and specific roles of the related CDK inhibitors p27 and p57 revealed by a knock-in mouse model. Proc Natl Acad Sci U S A 106, 5192-5197.
Sylvestre, Y., De Guire, V., Querido, E., Mukhopadhyay, U.K., Bourdeau, V., Major, F., Ferbeyre, G., and Chartrand, P. (2007). An E2F/miR-20a autoregulatory feedback loop. J Biol Chem 282, 2135-2143.
Takahashi, K., Kobayashi, T., and Kanayama, N. (2000). p57(Kip2) regulates the proper development of labyrinthine and spongiotrophoblasts. Mol Hum Reprod 6, 1019-1025.
Tanaka, S., and Diffley, J.F. (2002). Deregulated G1-cyclin expression induces genomic instability by preventing efficient pre-RC formation. Genes Dev 16, 2639-2649.
Tetzlaff, M.T., Yu, W., Li, M., Zhang, P., Finegold, M., Mahon, K., Harper, J.W., Schwartz, R.J., and Elledge, S.J. (2004). Defective cardiovascular development and elevated cyclin E and Notch proteins in mice lacking the Fbw7 F-box protein. Proc Natl Acad Sci U S A 101, 3338-3345.
Trikha, P., Sharma, N., Opavsky, R., Reyes, A., Pena, C., Ostrowski, M.C., Roussel, M.F., and Leone, G. (2011). E2f1-3 are critical for myeloid development. J Biol Chem 286, 4783-4795.
Trimarchi, J.M., and Lees, J.A. (2002). Sibling rivalry in the E2F family. Nat Rev Mol Cell Biol 3, 11-20. Tsai, K.Y., Hu, Y., Macleod, K.F., Crowley, D., Yamasaki, L., and Jacks, T. (1998). Mutation of E2f-1 suppresses apoptosis and inappropriate S phase entry and extends survival of Rb-deficient mouse embryos. Mol Cell 2, 293-304.
Tsai, S.Y., Opavsky, R., Sharma, N., Wu, L., Naidu, S., Nolan, E., Feria-Arias, E., Timmers, C., Opavska, J., de Bruin, A., et al. (2008). Mouse development with a single E2F activator. Nature 454, 1137-1141.
Ullah, Z., Kohn, M.J., Yagi, R., Vassilev, L.T., and DePamphilis, M.L. (2008). Differentiation of trophoblast stem cells into giant cells is triggered by p57/Kip2 inhibition of CDK1 activity. Genes Dev 22, 3024-3036.
Ullah, Z., Lee, C.Y., and Depamphilis, M.L. (2009a). Cip/Kip cyclin-dependent protein kinase inhibitors and the road to polyploidy. Cell Div 4, 10.
Ullah, Z., Lee, C.Y., Lilly, M.A., and DePamphilis, M.L. (2009b). Developmentally programmed endoreduplication in animals. Cell Cycle 8, 1501-1509. van den Heuvel, S., and Dyson, N.J. (2008). Conserved functions of the pRB and E2F families. Nat Rev Mol Cell Biol 9, 713-724.
Varmuza, S., Prideaux, V., Kothary, R., and Rossant, J. (1988). Polytene chromosomes in mouse trophoblast giant cells. Development 102, 127-134.
Vitrat, N., Cohen-Solal, K., Pique, C., Le Couedic, J.P., Norol, F., Larsen, A.K., Katz, A., Vainchenker, W., and Debili, N. (1998). Endomitosis of human megakaryocytes are due to abortive mitosis. Blood 91, 3711-3723.
202
Vitrat, N., LeCouedic, J.P., Pique, C., Debili, N., and Vainchenker, W. (1996). Megakaryocyte endomitosis are really abortive mitosis. Blood 88, 1135-1135.
Wang, Z.M., Yang, H., and Livingston, D.M. (1998). Endogenous E2F-1 promotes timely G0 exit of resting mouse embryo fibroblasts. Proc Natl Acad Sci U S A 95, 15583-15586.
Weiss, A., Herzig, A., Jacobs, H., and Lehner, C.F. (1998). Continuous Cyclin E expression inhibits progression through endoreduplication cycles in Drosophila. Curr Biol 8, 239-242.
Welch, C., Chen, Y., and Stallings, R.L. (2007). MicroRNA-34a functions as a potential tumor suppressor by inducing apoptosis in neuroblastoma cells. Oncogene 26, 5017-5022.
Wendling, F., Maraskovsky, E., Debili, N., Florindo, C., Teepe, M., Titeux, M., Methia, N., Breton-Gorius, J., Cosman, D., and Vainchenker, W. (1994). cMpl ligand is a humoral regulator of megakaryocytopoiesis. Nature 369, 571-574.
Weng, L., Zhu, C., Xu, J., and Du, W. (2003). Critical role of active repression by E2F and Rb proteins in endoreplication during Drosophila development. EMBO J 22, 3865-3875.
Wenzel, P., Wu, L., de Bruin, A., Chong, J., Chen, W., Dureska, G., Sites, E., Pan, T., Sharma, A., Huang, K., et al. (2007). Rb is critical in a mammalian tissue stem cell population. Genes Dev 21, 85-97.
Wenzel, P.L., Chong, J.L., Saenz-Robles, M.T., Ferrey, A., Hagan, J.P., Gomez, Y.M., Rajmohan, R., Sharma, N., Chen, H.Z., Pipas, J.M., et al. (2011). Cell proliferation in the absence of E2F1-3. Dev Biol 351, 35-45.
Wenzel, P.L., and Leone, G. (2007). Expression of Cre recombinase in early diploid trophoblast cells of the mouse placenta. Genesis 45, 129-134.
Wiemers, D.O., Shao, L.J., Ain, R., Dai, G., and Soares, M.J. (2003). The mouse prolactin gene family locus. Endocrinology 144, 313-325.
Winn, J., Carter, M., Avery, L., and Cameron, S. (2011). Hox and a newly identified E2F co-repress cell death in Caenorhabditis elegans. Genetics 188, 897-905.
Yamaguchi, M., Ogren, L., Endo, H., Soares, M.J., and Talamantes, F. (1994). Co-localization of placental lactogen-I, placental lactogen-II, and proliferin in the mouse placenta at midpregnancy. Biology of reproduction 51, 1188-1192.
Yamasaki, L., Jacks, T., Bronson, R., Goillot, E., Harlow, E., and Dyson, N.J. (1996). Tumor induction and tissue atrophy in mice lacking E2F-1. Cell 85, 537-548.
Yeung, E.C., and Meinke, D.W. (1993). Embryogenesis in Angiosperms: Development of the Suspensor. Plant Cell 5, 1371-1381.
Zhang, Y., and Kalderon, D. (2000). Regulation of cell proliferation and patterning in Drosophila oogenesis by Hedgehog signaling. Development 127, 2165-2176.
203
Zhang, Y., Sun, S., Chen, W.C., Kaluzhny, Y., Chinnappan, D., Yu, G., and Ravid, K. (2001). Repression of AIM-1 kinase mRNA as part of a program of genes regulated by Mpl ligand. Biochemical and biophysical research communications 282, 844-849.
Zhang, Y., Wang, Z., Liu, D.X., Pagano, M., and Ravid, K. (1998). Ubiquitin-dependent degradation of cyclin B is accelerated in polyploid megakaryocytes. The Journal of biological chemistry 273, 1387-1392.
Zhang, Y., Wang, Z.G., and Ravid, K. (1996). The cell cycle in polyploid megakaryocytes is associated with reduced activity of cyclin B1-dependent Cdc2 kinase. Journal of Biological Chemistry 271, 4266-4272.
Zheng, N., Fraenkel, E., Pabo, C.O., and Pavletich, N.P. (1999). Structural basis of DNA recognition by the heterodimeric cell cycle transcription factor E2F-DP. Gene Dev 13, 666-674.
Zhong, W., Feng, H., Santiago, F.E., and Kipreos, E.T. (2003). CUL-4 ubiquitin ligase maintains genome stability by restraining DNA-replication licensing. Nature 423, 885-889.
Zhu, W., and Depamphilis, M.L. (2009). Selective killing of cancer cells by suppression of geminin activity. Cancer Res 69, 4870-4877.
Ziebold, U., Reza, T., Caron, A., and Lees, J.A. (2001). E2F3 contributes both to the inappropriate proliferation and to the apoptosis arising in Rb mutant embryos. Genes Dev 15, 386-391.
Zielke, N., Querings, S., Rottig, C., Lehner, C., and Sprenger, F. (2008). The anaphase-promoting complex/cyclosome (APC/C) is required for rereplication control in endoreplication cycles. Genes Dev 22, 1690-1703.
Zimmet, J.M., Ladd, D., Jackson, C.W., Stenberg, P.E., and Ravid, K. (1997). A role for cyclin D3 in the endomitotic cell cycle. Mol Cell Biol 17, 7248-7259.
204