Chen2004 - An integrated yeast cell cycle model

December 2018, model of the month by Ashley Xavier
Original model: BIOMD0000000056

Introduction

The replication of an eukaryotic cell is a tightly controlled process involving multiple checkpoints, steady states and phases. Being a complex process that integrates various signals; mathematical models have been used to describe the cell cycle process from the early 1990s. The advantage of this approach is that it allows us to study the emergent behaviours and phenotypes from accurately describing the relatively simpler interactions between the different components. The models are made to reproduce known experimental evidences, which are then extended to make other verifiable predictions. This creates an iterative process where cell cycle models are constantly improved upon with the emergence of new findings.
The Chen 2004 model [1]describes the working of the major phases and checkpoints involved in a cell cycle. This model was able to accurately predict the phenotypes of over 100 mutant deletion strains known at the time.

Model

The model was built using both known and predicted regulatory relationships [2, 3] between the major players in the yeast cell cycle. A key feature of the eukaryotic cell cycle is the control of entry into and exit from the cell cycle phases. Once a wild-type yeast cell has made a commitment to mitotic cell division, it goes through all the phases in sequence. The different phases of the cell cycle are the first Growth phase(G1), DNA Synthesis phase(S), second Growth phase(G2) and the Mitotic phase(M) which need to occur in the right order.

Figure 2

Figure 1. Network diagram of the model with the major interactions visualized. Figure is taken from [1].

G1 phase

The division of a cell divides the mass and components between the daughter cells. A newly formed yeast daughter cell is thus usually small and replication at this stage would be highly detrimental to the survival of the would-be daughter cells. The purpose of the first growth phase is so that the cell can grow to be sufficiently large before it commits to replicating. The cell size is modelled by the buildup of Cln3 and Bck2 in the cell as it grows which leads to the synthesis of Cln2 and Clb5. Cln2 promotes budding, the formation of the initial outgrowth on the cell which eventually divides into a new daughter cell.

Figure

Figure 2. The dynamics of species important to the first growth phase(highlighted). Simulated concentrations in this figure are normalized. The data was generated in COPASI 4.24 and plotted using R 3.5.1.

S and G2 phases

Clb5 promotes the synthesis of new DNA to be divided between the two daughter cells, but is initially unable to execute its function because of Cyclin-dependent Kinase Inhibitors(CKI) forming a complex with it to deactivate it. Cln2 phosphorylates CKI leading to its degradation. Once CKI has been completely degraded the Clb5 can start the synthesis on new DNA.
CKI was also inhibiting another component Clb2. Clb2 promotes its own synthesis by promoting the activity of its own transcription factor. Once CKI is degraded even a very small amount of Clb2 leads to it being synthesized in large amounts due to this feedforward effect. Clb2 also downregulates the factors required for Cln3 and Bck2 to promote the expression of Cln2 and Clb5 preventing reinitiation of budding or DNA synthesis. Clb2 also prepares the Anaphase promoting Complex(APC) required for mitotic exit.

Figure

Figure 3. The dynamics of species important to the DNA synthesis phase(highlighted in red) and the second growth phase(highlighted in green). The data was generated in COPASI 4.24 and plotted using R 3.5.1.

M phase

APC should not initiate mitotic exit till all the DNA/chromosomes are duplicated and ready. This is ensured through the inhibition of APC by Mad2 which is turned off when all chromosomes are attached. Sister chromatid separation is initiated by the APC mediated degradation of Pds1 which in turn releases Esp1. APC also degrades Clb5 and partially degrades Clb2.

Figure

Figure 4. The dynamics of species important to the Mitotic stage(highlighted). The data was generated in COPASI 4.24 and plotted using R 3.5.1.

The species described above were handpicked to describe the overall trend of the model and is not a comprehensive description.
The parameters and values were fit to describe the wild-type trends and predict the phenotypes of 131 mutant strains. The concentrations used in the model are not the absolute concentrations since the values not known at the time but can be considered as effective values.

Results

The model was able to predict the phenotypes such as viability for wild-type, and 120 out of a total of 131 mutant strains. The 11 mutant strains whose phenotypes did not correspond to the model helps identify aspects of the model which can be further improved.
An Intermediate Enzyme with certain interactions had previously been predicted to explain experimental observations and this model suggests that APC is most likely that enzyme, has the predicted interactions and produces the same phenotypes.
The aspect of mitotic exit which later came to be known as FEAR(CDC Fourteen Early Anaphase Release) was elucidated when this model was being created and wasn’t incorporated into it. Interestingly the same behaviour of the FEAR network can be seen in the hypothesized species PPX.
A common feature of these networks is the negative feedback loops, where the events triggered by an entity generally leads to its own degradation. When this feature is disrupted the cell is highly prone to becoming inviable as can be seen in the mutant deletion strains.

Conclusion

Models of cell cycle regulation have been of tremendous importance partly due to the similarity and conservation of the core mechanisms throughout the eukaryotes, and partly due to the prevalent deregulation of its various components in cancer. These models are a representation of our understanding of the underlying mechanisms and help test our theories against experimental evidence. If a new experimental evidence does not match the prediction of the model, this could suggest flaws in our understanding, which can then be used to create better models. This iterative process allows us to accumulate and integrate understanding about these cellular processes in a form that can be widely used and tested.

 

References

  1. Katherine C. Chen, Laurence Calzone, Attila Csikasz-Nagy, Frederick R. Cross, Bela Novak, and John J. Tyson (2004). Integrative analysis of cell cycle control in budding yeast. Molecular Biology of the Cell Vol. 15, No. 8 . DOI:https://doi.org/10.1091/mbc.e03-11-0794
  2. Novak B, Pataki Z, Ciliberto A, Tyson JJ. (2001) Mathematical model of the cell division cycle of fission yeast. Chaos 2001 March 11(1):277-286 DOI:https://doi.org/10.1063/1.1345725
  3. Katherine C. Chen, Attila Csikasz-Nagy, Bela Gyorffy, John Val, Bela Novak, and John J. Tyson(2000) Kinetic analysis of a molecular model of the budding yeast cell cycle. Molecular Biology of the Cell Vol. 11, No. 1. DOI:https://doi.org/10.1091/mbc.11.1.369