Trying to unify the dierent ndings in the eld of bacteria

Get Complete Project Material File(s) Now! »

Brief overview of the mammalian cell cycle

The regulatory network driving cell cycle progression in mammalian cells relies on the same principles than in yeasts although the expansion of the family of cyclins and cyclin-dependent kinases (Cdk) through genetic evolution adds another layer of complexity Malumbres (2014). The precise description of the regulatory mechanisms that set the cell cycle clock in mammalian cells is beyond the scope of this work. Here we only brie y mention the main characteristics of this regulation with an emphasis on the checkpoints that potentially play a role in cell size control. Brie y, the classical view of cell cycle progression involves four dierent couples of cyclin-Cdk that regulate the progression through the cell cycle (gure 4.2). Upon mitogen stimulation (i.e. EGF), cyclin expression starts and marks the beginning of G1. Cyclin D-Cdk4/6 complex induces the expression of cyclin E in mid-G1. cyclin E-cdk2 drives the G1/S transition after which cyclin E is rapidly degraded. Finally, cyclin A-cdk1 and cyclin B-cdk1 complexes drive the transition from G2 to M when they are degraded (reviewed in Malumbres and Barbacid (2009)). However, all these Cdk except Cdk1 are dispensable for cell cycle progression Santamara et al. (2007), thus suggesting that the core mechanism driving cell cycle progression is driven by a unique cyclin-Cdk complex, similarly to yeast.
The cyclic and timely regulated activity of these cyclin-Cdk complexes is at the heart of the successive cell cycle checkpoints that drive cell cycle progression Kastan and Bartek (2004). Although much is known about the molecular networks underlying each of these cell cycle transitions, often, the initial trigger is unknown. Mitotic entry for instance is well known to depend on a bistable switch in Cdk1 activity regulated by Cdc25/Wee1 but the initial trigger to this switch is still mysterious Mchedlishvili et al. (2015). Similarly, G1/S transition occurs in a switch-like manner via a network that involves cyclin E-Cdk2 Barr et al. (2016); Bertoli et al. (2013); Cappell et al. (2016) and several experimental results suggest that this transition depends on cell growth or cell size (discussed in the next section 4.2.1). Yet, the molecular links between these two observations are still mostly unknown. We will rst describe the known links between growth regulatory pathways such as mTOR and cell cycle checkpoints during G0 and G1. In the next section, we will then discuss the results suggesting that progression through G1 is also size-dependent.

Commitment to a new the cell cycle: checkpoint(s) for growth during G0 and G1

After division, cells can either enter a quiescent state (G0) or commit to a new round of division and
progress through G1 phase (gure 4.2). A restriction point (R) occuring in G1 and checking nutrient
availability, similarly to Start checkpoint in S. cerevisiae (section 3.2) was rst proposed in Pardee (1974). The existence of such commitment point was further supported by the description of a point
in early G1 after which cells complete their cell cycle independently of serum removal Zetterberg and Larsson (1985). The molecular basis of the restriction point is thought to be the hyperphosphorylation of Rb that leads to the activation of E2F, a major transcription factor for cell cycle Bertoli et al. (2013) Since these rst in uential studies, much work to identify the mechanism(s) underlying this restriction point has been done. A unied vision of how these mechanisms dene the restriction point is however still missing (reviewed in Fisher (2016)). A rst interesting idea emerging from these works is that signaling during the previous cell cycle plays a role in the decision to commit into a new round of division. Newborn MCF10A cells were shown to be separated into two sub-populations displaying either low or intermediate Cdk2 activity Spencer et al. (2013). The low levels of Cdk2 caused the cells to enter a transient G0-like quiescent state in contrast with the high level Cdk2 cells that immediately entered G1. The proportion of low Cdk2 cells increased as serum was removed during a period of time lasting 8 hours prior to division. Therefore, in addition to the classical view of cyclin D-Cdk4/6 dependent R passage, Cdk2 activity inherited at birth also plays a role in cell-cycle commitment.
A second interesting emergent concept is that decision to commit to the next cell cycle depends on several signaling processes. It has been proposed that two independent checkpoints exist during G1.
The rst depends on growth factors via the Ras/Raf/MAPK/Cyclin D pathway and the second one occurs later, checks nutrient availability via the mTORC1 pathway which ultimately enhances cyclin E-Cdk2 activity Foster et al. (2010). Even more recently, another checkpoint for extracellular lipids was also proposed Patel et al. (2016).
Overall, it seems that several processes occurring in G0 and G1 and that both depends on signaling during the previous cell cycle and cues from the external environment have a role in the cell committing to a new round of division and growth. Better understanding of the links between the growth regulatory pathways such as mTORC1 and the cell cycle network will help rening our understanding of the potentially several restriction points occurring during G1. Furthermore, similarly to S. cerevisiae, G1/S transition is thought to be size-dependent.

Population-level studies and evidence of size control in early work

Famous work by Zetterberg and Killander using interferometric microscopy on broblasts in the 1960s produced the rst result suggestive of a size-dependent G1 phase. They showed that variability in Figure 4.2: Cell cycle progression in mammalian cells. Top: Early in the cell cycle, the cell can either enter a quiescent state (G0) or commit to a new round of division and enter G1.
This commitment point is thought to be gated by a restriction point (R) that checks that sucient nutrients and growth factors are available. There are however possibly multiple restriction points occuring during G1 (see main text) and G1/S transition is possibly at least partly dependent on cell size. S phase is the phase when replication occurs while G2 is a less well-dened additional growth phase before mitosis. Bottom: Cyclin-Cdk regulate progression through cell cycle phases. CyclinD expression starts at the begining of G1 and assocites with cdk4/6 to induce the expression od cyclin E in mid G1. CyclinE-Cdk2 levels and activity reach a peak at G1/S transition after which cyclin E is rapidly degraded, cyclin A-Cdk1 and cyclin B-Cdk1 drive the transition from G2 to M. size when entering S phase was less important than after birth and that entry into S phase depended more on cell size than on cell age (time spent from birth to S phase) Killander and Zetterberg (1965).
Since then, other works have concluded a same size-dependence progression in G1 phase in dierent cell types (Shields et al. (1978); Gao and Ra (1997), reviewed in Jorgensen and Tyers (2004)). An elegant work from Dolznig et al. (2004) brought strong arguments in favor of a size-sensing mechanism in G1. By inducing avian erythroblasts to proliferate under the control of either normal c-Kit/EpoR physiological cytokines or constitutively active oncogene v-ErB, the authors could tune the rate of proliferation and the nal size of cells. v-ERB cells cycled faster and were larger: they had a faster growth rate and division rate and entered the cell cycle with a volume that was 1.3 fold higher than the control cells. When switched back c-Kit/EpoR, v-ERB cells growth rate was immediately reduced but G1 phase was shorter than control cells so that v-ErB cells would grow less and correct for their larger size. The second cell cycle showed similar G1 phase length in both conditions. This result led the authors Dolznig et al. (2004) and others Ginzberg et al. (2015) to the conclusion that a critical size threshold exists in G1 phase. However, it should be noted from their results that v-ErB cells size takes more than one cell cycle to regress towards the mean size in the c-Kit/EpoR condition, therefore suggesting a mechanism more complex than a simple critical sizer in G1: an adder or a mechanism where a timer limits the reduction of cell cycle length and thus leads to an uncomplete correction could for example be considered.

Results challenging the idea of size control in metazoan cells.

It is important to recall that not all studies agree on the existence of a size-sensing mechanism in metazoan cells, as exposed in the rst part of this introduction. In a study based on Coulter counter
volume measurement performed at the population level, it was proposed that primary Schwann cells grow linearly and therefore do not need to regulate their size Conlon and Ra (2003). In this study, cells were blocked in S phase by aphidicolin and increased linearly their volume over a ve-fold increase during 5 days. Theses conclusion were however challenged by theoretical considerations emphasizing the need of direct and dynamic measurement on single-cells Sveiczer et al. (2004). Moreover, since recent result suggested that a coupling between growth rate and size occurs in G1 Son et al. (2012), it will be interesting to see whether this result came from the fact that cells were blocked later in the cell cycle, in S phase. Later in the same lab, it was shown that growth factors and mitogens independently regulate growth and proliferation and that their amount in the culture media sets cell size, therefore suggesting the absence of size checkpoint in Schwann cells Echave et al. (2007). As the recent progress in the characterization of size control in yeast and bacteria (chapter 3) illustrate, direct measurement of cell growth at the single-cell level is key to make the next steps in the understanding of size control in metazoan cells.

Table of contents :

Acknowledgments
Abstract
List of abbreviations
1 The question of cell size regulation 
1.1 Dening cell size
1.1.1 Cell mass and cell volume
1.1.2 Cell size in proliferating cells: coordination between growth and cell cycle progression
1.2 Cell size homeostasis in metazoan cells
1.2.1 Cell size vs. Organ size
1.2.2 Intrinsic vs. extrinsic regulation of cell size
1.2.2.1 Flexibility upon environmental changes
1.2.2.2 Coupling of cell growth and cell cycle progression?
1.2.3 Proliferating cells in multicellular organisms
1.2.4 Evolutionary point of view on the requirement of growth regulation in unicellular and multicellular organisms
1.3 Conclusion
2 Size homeostasis study: question, challenges and concepts step by step 
2.1 Sizer, timer & adder
2.2 Coordination of cell cycle progression to growth
2.3 Models of size-sensing mechanism
2.3.1 Geometric measurement of absolute length
2.3.2 Volume measurement through titration-based mechanism
2.3.3 Surface area measurement
2.3.4 Conclusion: the importance of cell volume
2.4 An emergent role for growth regulation in size control
2.4.1 The regulation of growth
2.4.1.1 Growth in volume
2.4.1.2 Growth in mass
2.4.1.3 Growth and metabolism
2.4.1.4 Density throughout the cell cycle
2.4.1.5 Conclusion
2.4.2 Measuring cell growth at the single-cell level
2.4.2.1 New techniques enabling single-cell measurement
2.4.2.2 From population-level studies to single-cell measurement
2.5 Combination of processes?
2.5.1 Independent control in each cell cycle phase
2.5.2 One overarching mechanism
2.5.3 Parallel and concurrent processes
2.6 Robustness and exibility of size control
2.6.1 Apparent exibility of size thresholds
2.6.2 What we can learn from studying cells in dierent growth environments
2.7 Conclusion
3 Current understanding of size homeostasis in yeast and bacteria 
3.1 Cell size homeostasis in bacteria
3.1.1 Brief description of the cell cycle of bacteria
3.1.2 Phenomenological description and recent emergence of the adder model
3.1.3 Beyond the adder observation, current mechanisms debated
3.1.3.1 Control in each sub-periods
3.1.3.2 Molecular players
3.1.4 Trying to unify the dierent ndings in the eld of bacteria
3.1.4.1 How do these controls combine to generate an adder?
3.1.4.2 Lessons from model-free approaches
3.2 Cell size homeostasis in yeast
3.2.1 Size control in S. pombe
3.2.1.1 S. pombe, cell cycle and size-checkpoints
3.2.1.2 S. pombe is the classical example of a sizer in wild-type cells
3.2.1.3 S. pombe. grows in a bilinear fashion
3.2.1.4 In search for the mechanism generating geometric sensing of size
3.2.2 Size control in S. cerevisiae
3.2.2.1 S. cerevisiae, cell cycle and checkpoints
3.2.2.2 Lack of consensus about the growth behaviour of S. cerevisiae
3.2.2.3 Daughter cell of S. cerevisiae behave in an adder-like manner
3.2.2.4 Whi5 and the inhibitor-dilution size sensor model for G1/S transition
3.3 Lessons from yeast and bacteria
4 Cell size homeostasis in metazoan cells 
4.1 Growth and cell cycle pathways in mammalian cells
4.1.1 Pathways regulating the growth rate
4.1.1.1 The mTOR pathway
4.1.1.2 Other pathways
4.1.1.3 Conclusion
4.1.2 Cell cycle regulation
4.1.2.1 Brief overview of the mammalian cell cycle
4.1.2.2 Commitment to a new the cell cycle: checkpoint(s) for growth during G0 and G1
4.2 Population-level studies and evidence of size control in early work
4.2.1 Evidence of size-sensing in G1
4.2.2 Results challenging the idea of size control in metazoan cells
4.3 The challenging problem of characterizing cell growth in metazoan cells
4.4 Current views on the homeostatic process in metazoan cells
4.4.1 Cells grow exponentially or super-linearly
4.4.2 No clear role for time modulation, possible existence of a growth-rate modulation
4.5 Conclusion
5 General conclusion and aims of this study 
5.1 General conclusion
5.2 Aims of this study
6 Methods 
6.1 Long time-lapse acquisition in animal cells to characterize growth: state of the art and challenges
6.2 Fluorescence-exclusion based volume measurement
6.2.1 Volume measurement method: principle and validation
6.2.1.1 Method principle
6.2.1.2 Theoretical validation
6.2.1.3 Experimental validation
6.2.2 Image analysis optimization
6.2.3 Protocol optimization for long time-lapse acquisition
6.2.3.1 Improve nutrient access
6.2.3.2 Standardized cell culture protocol to try reduce variability
6.2.3.3 Analytical validation of the quality of growth in the experiments
6.2.3.4 Improving the uorescent probes
6.3 Cell growth analysis: clonal & single-cell curves, cell cycle transitions
6.3.1 Single-cell curves
6.3.1.1 Careful control of the sources of volume curves uctuations
6.3.1.2 Semi-automated analysis of hundreds of single-growth curves
6.3.2 Clonal growth curves
6.3.3 Cell cycle transitions keypoint analysis
6.4 Developing tools to induce asymmetrical divisions
6.4.1 Inducing asymmetrical divisions using micro-channels
6.4.1.1 Micro-channels induce asymmetrical cell divisions and allow volume measurement
6.4.1.2 Optimization of the micro-channels device to improve nutrient access
6.4.2 Asymmetrical patterns
6.4.3 Drug-induced asymmetrical distribution of organelles
6.5 Choice of cell types
6.5.1 Description of the cell types used in this study
6.5.2 Establishing new stable lineages
6.6 Statistical analysis
7 Results 
7.1 Summary
7.2 Manuscript in preparation for submission
7.3 Concluding remarks
7.3.1 Analysis planed before the submission
7.3.2 Important experiments needed to complete the work
8 Discussion 
8.1 Single cell growth measurement in mammalian cells
8.1.1 Single cell measurement of volume with the FXm
8.1.1.1 FXm allows measurement of both suspended and adherent cells
8.1.1.2 FXm allows measurement of cell volume
8.1.1.3 Results from direct measurement of single cell growth might precise some of the previous ndings
8.1.1.4 FXm measurements provide a robust set of data for the study of size homeostasis
8.1.2 Current limitations of the FXm
8.1.2.1 Limitation to cells which do not internalize the probe
8.1.2.2 FXm currently requires a device conning the cells
8.2 Growth and time modulation
8.2.1 Time modulation
8.2.1.1 G1 duration is correlated with volume in HT29 and HeLa cells and gated by a minimum duration
8.2.1.2 S and G2 phases
8.2.1.3 Conclusion on the role of time adaptation
8.2.2 Growth modulation
8.2.2.1 Experimental evidence
8.2.2.2 Growth regulation and cellular homeostasis
8.2.2.3 Conclusions on the role of growth modulation
8.2.3 Conclusion on the respective contribution of time and growth modulation in size control
8.2.3.1 Classication of the results into three types of combinations of growth and time modulation
8.2.3.2 Identication of three distinct rate-limiting processes for cell cycle and cell growth
8.3 Is there a unique mechanism or several processes resulting in the adder?
8.3.1 Generality of the adder, from bacteria to mammalian cells
8.3.1.1 Generality of the adder in our results
8.3.1.2 Generality of the adder in the litterature
8.3.2 Is there enough evidence to conclude that size homeostasis is an adder in mammalian cells?
8.3.2.1 Statistical resolution is currently lacking to conrm the adder in any organism
8.3.2.2 Testing the adder
8.3.3 Size homeostasis is a exible process
8.3.3.1 Bacteria and S. cerevisiae: not always an adder
8.3.3.2 Why exibility is important when trying to build a model for size homeostasis
8.3.3.3 Conclusion: phenomenological adder and molecular adder
8.3.4 Several scenarios can explain the apparent adder in mammalian cells
8.4 Perspectives: a combination of processes, both single and tissue-level determined?
9 Conclusion 
Bibliography

GET THE COMPLETE PROJECT