Biology Reference
In-Depth Information
kinase
[87]
. Hence, young yeast cells with damage block at
the early G1 checkpoint rather than the late G1 checkpoint.
DNA damage received later in the cycle causes a block at
the M/A transition in budding yeast cells
[88]
, whereas
most other cell types block at the G2/M transition.
Unicellular eukaryotes also differ from metazoan cells
in the strength, duration and consequences of checkpoints.
The purpose of checkpoints is to block progression through
the cell cycle if problems arise that compromise successful
replication of the cell and its genome. If the problem can be
repaired, the checkpoint can be lifted and the cell can
proceed with the replication
e
division cycle. But what
should the cell do if the problem cannot be repaired? For
a unicellular organism, the best strategy is to bypass the
checkpoint after some time and proceed to cell division.
The worst thing that can happen is that the problem is lethal
and the cell dies. But in many cases the problem is not
lethal, the cell survives and reproduces, and the daughter
cells 'get on with life'. Maybe they carry some new
mutations, maybe they are aneuploid or polyploid, but at
least they are alive. For metazoans, on the contrary, these
damaged cells inhabit a larger organism, and the mutations
they carry may prove advantageous for the cell but fatal for
the whole organism (think of malignant cancer cells).
Hence, mammalian cells (which have been most thor-
oughly studied in this regard) tend to have stronger
checkpoints, and if the damage cannot be repaired then the
surveillance mechanism redirects the cell toward pro-
grammed cell death
[89]
. Better that the damaged cell be
destroyed than that its progeny destroy the whole organism.
are size-regulated at the G1/S transition. In budding yeast, it
should be noted, size control is strong in small daughter cells
but weak in large mother cells
[90]
. It matters little that
mother cells get progressively larger from generation to
generation, because they eventually senesce and die.
Size control is less evident in metazoans than in yeast
[96,97]
. Unlike yeast cells, which maintain a stable size
distribution over many generations of growth and division
[98]
, cell lineages in metazoans are more variable in size,
especially lineages of restricted proliferative potential. We
would expect size control to be more evident in cells with
high proliferative capacity (germline cells and stem cells).
Eggs and early embryos are interesting cases. Eggs grow
very large and arrest in meiosis II with a single, haploid
nucleus. After fertilization, the diploid zygote undergoes
a series of rapid mitotic cycles (without growth) to create
a ball of small, mononucleated cells (the blastula). Although
details of these early embryonic divisions vary from one
organism to another, in general the cell cycle checkpoints are
not operating as usual. Around the mid-blastula stage the
mitotic division cycles change dramatically, acquiring G1
and G2 phases and checkpoint controls, including growth
controls. At this point the embryo starts to grow and develop,
using resources stored in the egg or provided by the placenta.
The fourth feature, that the control system must be
robust in the face of unavoidable molecular noise in the
small environs of a cell, is beyond the scope of this chapter.
Suffice it to say that in recent publications we have studied
realistic stochastic models of bistability and irreversibility
in the cell cycle engine and found that the control system
shows exactly the same sort of robustness
e
variability
exhibited by proliferating yeast cells
[99
e
101]
.
If we have done our job well, then everything we have
said should seem natural and intuitively appealing. If
everything is so obvious, the skeptical reader might ask,
'Why do we need mathematical models and bifurcation
diagrams? It's all right there in the reaction networks!' We
hope most readers will not be so jaded. These ideas were
not so obvious to us until we started thinking about cell
cycle regulation in terms of mathematical models. We are
convinced that a mathematical, systems-level approach to
regulatory networks is absolutely essential to a correct
understanding of physiological control systems like the
eukaryotic cell cycle.
CONCLUSIONS
If our vision of the eukaryotic cell cycle control system is
correct, then it should account naturally for the four char-
acteristic features of mitotic cell division enumerated in the
introduction. The first feature, that cell cycle progression is
unidirectional and irreversible, and the third feature, that
checkpoints guard the major transitions of the cell cycle, are
the fundamental ideas behind our theory. We have shown
how these features are based on the dynamics of the inter-
acting genes and proteins that govern cell cycle progression.
The second feature, balanced growth and division, is
a special case of the checkpoint paradigm. In unicellular
eukaryotes, growth to a minimal size is a requirement for
passing one of the checkpoints of the cell cycle. In budding
yeast the size requirement is enforced at Start
[55,90]
.In
fission yeast, it is enforced at the G2/M transition
[91]
,asis
also true of the acellular slime mold Physarum poly-
cephalum
[92]
. Strict size control is also evident in the
physiology of Stentor
[93]
and Amoeba
[94,95]
,although
the molecular details have never been worked out. By
mutations, the size checkpoint can be moved to a different
transition point; for example, wee1
REFERENCES
[1] Hartwell LH. Nobel Lecture. Yeast and cancer. Biosci Rep
2002;22:373
e
94.
[2] Hunt T. Nobel Lecture. Protein synthesis, proteolysis, and cell
cycle transitions. Biosci Rep 2002;22:465
e
86.
[3] Nurse PM. Nobel Lecture. Cyclin dependent kinases and cell
cycle control. Biosci Rep 2002;22:487
e
99.
[4] Nurse P. Universal control mechanism regulating onset of M-
phase. Nature 1990;344:503
e
8.
D
mutants of fission yeast
