TEACHING RESOURCE

C O M P U TAT I O N A L B I O L O G Y nesses of Tyson’s 1991 model (8) are pre-

sented. This discussion emphasizes aspects
Computational Modeling of the Cell Cycle of the biological pathways that have become
apparent subsequent to the model’s publi-
Eric A. Sobie* cation (Slides 8 to 10), and several details
not included in Tyson’s original model (8)
This Teaching Resource provides lecture notes, slides, and a problem set for are mentioned. For instance, autocatalytic
introducing graduate-level students to computational biology through a simple
activation of the dephosphorylation reac-
mathematical model of the cell cycle. The model simulates interactions between
tion is not directly mediated by MPF itself
cyclin B and cyclin-dependent kinase 1, proteins that together form the mitosis-
but instead occurs through the phosphatase
promoting factor (MPF), which initiates the processes leading to mitosis. The
lecture begins with a biological background describing the importance of MPF Cdc25 (Slide 8) (9). In addition, cyclin B
for mitosis, the components of MPF, and the changes in cellular MPF observed degradation does not occur at a constant
during different phases of the cell cycle. The model is compared with newer, rate, as assumed by Tyson, but instead is
more mechanistically detailed models of the same process, which allows for a triggered by MPF (Slide 9) (9). Lastly, the
discussion of the insights that can be gained even from simplified models. The Tyson model did not include a role for the
lecture concludes with a demonstration of how this model can be implemented in protein Wee1 (Slide 10), which is a kinase
the scientific programming language MATLAB and includes a problem set. that phosphorylates Cdk1, which shifts
MPF from the active to inactive form (10).

Downloaded from http://stke.sciencemag.org/ on December 15, 2017

Lecture Notes residues by the phosphatase Cdc25 converts Despite these simplifications, Tyson’s
The signaling pathways controlling the cell pre-MPF into its active form (Slide 5). A essentially phenomenological model gen-
cycle in eukaryotes have been studied inten- rapid increase in active MPF in eukaryotic erated important insights into cell cycle
sively for many decades, and the compo- cells accompanies the transition from G2 regulation (Slides 11 to 13). The model
nents are now known in great detail. Many to M-phase (Slide 4), and this surge in the exhibited distinct behaviors, which could
of the important advances have been made cytosolic concentration of MPF ([MPF]) then be connected to particular biological
using model systems, such as oocytes from initiates many processes that are required states and experimental observations. Ex-
the South African clawed frog Xenopus lae- for orderly cell division. The dephosphoryl­ amples of these model behaviors include the
vis and the budding yeast Saccharomyces ation reaction is autocatalytic so that an ini- spontaneous oscillations in [MPF] that are
cerevisiae. However, the many interlock- tial increase in active MPF increases the rate analogous to the rapid division cycles that
ing loops of activation and inhibition make at which the reaction occurs (Slide 5). MPF occur in the Xenopus oocyte immediately
control of the cell cycle difficult to interpret also triggers the rapid degradation of cyclin after fertilization (Slide 11). The model also
intuitively. As a result of this complexity, B, decreasing cellular [MPF] and allowing revealed two mechanisms for cessation of
computational modeling has been used to the cell cycle to begin again. As a result of oscillations, which represented either meta-
study this system, and several mathemati- these regulatory reactions, the cell cycle is phase arrest or the G0 state of nondividing
cal models have been published (1–8). accompanied by periodic increases and de- somatic cells (Slide 12). Additionally, the
Although all of these models successfully creases in the abundance of cyclin (Slide 4). model showed “excitability” so that under
reproduce certain fundamental features of The core of the lecture is the presenta- some conditions, [MPF] is low, but a small
the cell cycle, they differ greatly from each tion of a mathematical model published by stimulus that initiates an increase in [MPF]
other in terms of the level of detail and the Tyson in 1991 (8) (Slide 6). This relatively can induce a further surge in [MPF] and
processes that are simulated. simple model allows for a discussion of how trigger a cell division (Slide 13). Because
The lecture begins by defining relevant to construct models that consist of systems these model behaviors occurred with differ-
terms and reviewing biological concepts of ordinary differential equations (ODEs), ent combinations of the model’s parameters,
important for understanding the cell cycle. and the results presented in the paper il- they could be connected to different biologi-
Mitosis-promoting factor (MPF; also known lustrate the biological insights that can be cal states and interpreted in this context.
as maturation-promoting factor or M-phase gained through simulations. Tyson’s model To reinforce the issue of simple versus
promoting factor), originally described in simulates the following processes: synthesis more complex models, improvements that
experiments performed on extracts from of cyclin B, binding and unbinding of cyclin have been reported in subsequent publica-
Xenopus oocytes, is made up of a molecule B to Cdk1, phosphorylation and dephos- tions from the same group are discussed
of cyclin B bound to a molecule of Cdk1 phorylation of Cdk1, degradation of cyclin (Slides 14 to 18) (3–5, 11). These examples
(cyclin-dependent kinase 1) (Slide 3). B, and the autocatalytic activation by MPF illustrate the gradual evolution that occurs
When Cdk1 is phosphorylated at Thr14 and of the pre-MPF dephosphorylation reac- with many mathematical representations of
Tyr15, MPF is in an inactive form, known as tion. The result is a system of six nonlinear biological processes.
“pre-MPF.” Dephosphorylation of these ODEs. The lecture discusses a few equa- The lecture concludes with information
tions in detail to illustrate that each ODE that is required for students to implement
describes the difference between the rate of the model in a computing language, such as
Department of Pharmacology and Systems
Therapeutics and Systems Biology Center New
appearance and the rate of disappearance of MATLAB (Slides 19 to 21).
York, Mount Sinai School of Medicine, New a particular chemical species (Slide 7).
York, NY 10029, USA. To illustrate the issue of simplicity ver- Problem Set
*Corresponding author. E-mail, eric.sobie@ sus mechanistic detail that arises repeatedly Introductory Details
mssm.edu in modeling studies, the strengths and weak- The homework assignment includes two
www.SCIENCESIGNALING.org 27 September 2011 Vol 4 Issue 192 tr11 1
TEACHING RESOURCE

parts, a programming component (Part 1 min−1, respectively. Keeping k4 constant at Slides. Computational modeling of the
1) and a second component that reveals 180 min−1, run the model for different val- cell cycle.
whether the students have gained biologi- ues of k6. Either decreasing or increasing MATLAB code. Three MATLAB files
cal insight (Part 2). For Part 1, the students k6 will eventually abolish oscillations. How for computational analysis of the cell cycle
are provided with a model that solves Ty- small does k6 have to become before oscil- with ordinary differential equations.
son’s model (8) using Euler’s method, and lations cease? How large does k6 have to be- Problem set answer key file is available
are asked to convert this into a program that come before oscillations cease? upon request
uses MATLAB’s ODE solvers. Because it 2. On a single graph, plot [MPF] versus
is generally quite difficult for students with- time for k4 = 180 min−1 and the following References and Notes
1. B. D. Aguda, A quantitative analysis of the kinet-
out programming experience to produce three values of k6: 0.2 min−1, 1 min−1, and 5 ics of the G(2) DNA damage checkpoint system.
such a program de novo, an incomplete pro- min−1. Only k6 = 1 min−1 should show spon- Proc. Natl. Acad. Sci. U.S.A. 96, 11352–11357
gram is provided as a template. In Part 2, taneous oscillations in [MPF]. Considering (1999).
the students vary parameters to reproduce the biochemical reactions simulated by the 2. K. C. Chen, A. Csikasz-Nagy, B. Gyorffy, J. Val, B.
Novák, J. J. Tyson, Kinetic analysis of a molecular
several of the figures in the original Tyson model, why do very small values of k6 cause model of the budding yeast cell cycle. Mol. Biol.
publication and then interpret the biological oscillations to cease? Why do very large Cell 11, 369–391 (2000).
meaning of how changing parameters alters values of k6 cause oscillations to cease? 3. K. C. Chen, L. Calzone, A. Csikasz-Nagy, F. R.
the model’s response. 3. Although either a large increase or a Cross, B. Novak, J. J. Tyson, Integrative analysis
of cell cycle control in budding yeast. Mol. Biol.
The Supplemtary Materials include three large decrease in k6 will abolish spontane- Cell 15, 3841–3862 (2004).

Downloaded from http://stke.sciencemag.org/ on December 15, 2017

MATLAB programs: “tyson.m”, which ous oscillations, your plot should show that 4. A. Csikász-Nagy, D. Battogtokh, K. C. Chen, B.
uses Euler’s method to solve the system of the steady-state concentrations of MPF are Novák, J. J. Tyson, Analysis of a generic model
of eukaryotic cell-cycle regulation. Biophys. J. 90,
ODEs; “tyson_ode.m”, which is an incom- different in the two cases. To what different
4361–4379 (2006).
plete program that sets up a solution using biological states do these different steady 5. B. Novák, J. J. Tyson, Numerical analysis of a
MATLAB’s solvers; “dydt_tyson.m”, which levels correspond? comprehensive model of M-phase control in Xen-
is an incomplete function that is called by opus oocyte extracts and intact embryos. J. Cell
Sci. 106, 1153–1168 (1993).
“tyson_ode.m”. Educational Details 6. J. R. Pomerening, E. D. Sontag, J. E. Ferrell Jr.,
If “tyson” is typed at the MATLAB Learning Resource Type: Lecture, as- Building a cell cycle oscillator: Hysteresis and bi-
prompt, the program will integrate the equa- signment, digital presentation stability in the activation of Cdc2. Nat. Cell Biol. 5,
tions and plot output similar to that shown Context: Graduate 346–351 (2003).
7. Z. Qu, W. R. MacLellan, J. N. Weiss, Dynamics
in figure 3A of Tyson’s paper (8). Intended Users: Teacher, learner of the cell cycle: Checkpoints, sizers, and timers.
Intended educational use: Learn, plan, Biophys. J. 85, 3600–3611 (2003).
Student Assignment: Part 1— teach 8. J. J. Tyson, Modeling the cell division cycle:
Programming Discipline: Biochemistry, bioengineer- cdc2 and cyclin interactions. Proc. Natl. Acad. Sci.
U.S.A. 88, 7328–7332 (1991).
Using the working program “tyson.m” as a ing, biophysics, cell biology, develop- 9. B. Alberts, A. Johnson, J. Lewis, M. Raff, K.
guide, add lines of MATLAB code to “ty- mental biology, education, physiology, Roberts, P. Walter, Molecular Biology of the Cell
son_ode.m” and “dydt_tyson.m” so that the reproductive biology, theoretical biology (Garland Science, New York, ed. 5, 2007).
10. R. Boutros, V. Lobjois, B. Ducommun, CDC25
equations will be integrated when “tyson_ Keywords: Dynamical systems, math-
phosphatases in cancer cells: Key players? Good
ode” is typed at the MATLAB prompt. The ematical modeling, mitosis, numerical targets? Nat. Rev. Cancer 7, 495–507 (2007).
locations where lines of MATLAB code methods, Xenopus oocytes, simulation, 11. J. C. Sible, J. J. Tyson, Mathematical modeling as
need to be added are indicated in the two in- computer programming, MATLAB a tool for investigating cell cycle control networks.
Methods 41, 238–247 (2007).
complete files. When this is working, typing 12. Acknowledgments: The author thanks Y.-S.
“tyson_ode” should generate a plot similar Technical Details Lee for testing the problem set. Funding: The
to figure 3A in Tyson’s paper (8). Format: PowerPoint (.ppt) development of this course was supported by a
Size: 1.58 MB Systems Biology Center grant (P50 GM071558)
and a training grant in Pharmacological Sciences
Student Assignment: Part 2—Biological Requirements: Microsoft PowerPoint (T32GM062754).
Interpretation Format: MATLAB (.m)
1. Two especially important parameters in Size: 1 KB, 3 KB, 3 KB
10.1126/scisignal.2001985
this model are k4 and k6. In the Tyson pa- Requirements: Mathworks MATLAB
per (8), figure 2 shows that changes in these
parameters lead to qualitatively different Supplementary Materials
model behavior. The control values of these http://stke.sciencemag.org/cgi/content/full/ Citation: E. A. Sobie, Computational modeling of
parameters in the model are 180 min−1 and sigtrans;4/192/tr11/DC1 the cell cycle. Sci. Signal. 4, tr11 (2011).

www.SCIENCESIGNALING.org 27 September 2011 Vol 4 Issue 192 tr11 2

Computational Modeling of the Cell Cycle
Eric A. Sobie

Sci. Signal. 4 (192), tr11.

DOI: 10.1126/scisignal.2001985originally published online September 20, 2011

Downloaded from http://stke.sciencemag.org/ on December 15, 2017

ARTICLE TOOLS http://stke.sciencemag.org/content/4/192/tr11

SUPPLEMENTARY http://stke.sciencemag.org/content/suppl/2011/09/23/4.192.tr11.DC1
MATERIALS

RELATED http://stke.sciencemag.org/content/sigtrans/4/190/tr2.full
CONTENT
http://stke.sciencemag.org/content/sigtrans/6/271/tr2.full
http://science.sciencemag.org/content/sci/353/6306/aaf1420.full

REFERENCES This article cites 10 articles, 5 of which you can access for free
http://stke.sciencemag.org/content/4/192/tr11#BIBL

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Use of this article is subject to the Terms of Service

Science Signaling (ISSN 1937-9145) is published by the American Association for the Advancement of Science, 1200 New
York Avenue NW, Washington, DC 20005. 2017 © The Authors, some rights reserved; exclusive licensee American
Association for the Advancement of Science. No claim to original U.S. Government Works. The title Science Signaling is a
registered trademark of AAAS.