¿ìè¶ÌÊÓƵ

Few phenomena in biology are quite as dramatic as the cell cycle, the
ordered sequence of events by which cells grow and divide. Under the microscope
the high point of the cycle is undoubtedly division itself, especially
the stately dance of the chromosomes which ensures that each new cell is
born with a full set of genetic instructions. This spectacle – mitosis
– is followed by a much quieter, but no less important, period during which
the new cells grow and copy their DNA, ready for the next division and
another performance of the mitotic dance.

The cycle is controlled by a biochemical machine of exquisite design.
In recent years researchers have taken much of this machinery to pieces,
showing how it works in normal cells and hoping to shed light on its relation
to cancer. They have been particularly interested in the main transitions
of the cycle: the entry into mitosis, the exit from mitosis and the point
at which the cell starts copying its DNA. The most remarkable discoveries
have centred on the control of mitosis.

There are many components in the mitotic machine, but a protein called
cdc2 (cell division cycle) is particularly important. When suitably activated,
cdc2 is a powerful and versatile molecular tool, whose forte is to spur
other molecules into action, an effect which it achieves chemically by ‘persuading’
them to pick up and bind a phosphate group. The addition and removal of
phosphate groups is a standard technique by which cells switch proteins
from dormant to active states and vice versa. The process is common in
many different areas of cell chemistry.

The cdc2 protein is behind many of the spectacular events of cell division.
Although its exact role in the living cell is still being unravelled, the
evidence from the test tube is persuasive. The protein adds phosphate groups
to a network of linked proteins called nuclear lamins, which line the membranous
envelope around the cell’s nucleus. Lamins respond by severing links with
their neighbours, an effect which dissolves the nuclear envelope.

Another of cdc2’s roles is to add phosphate groups to histone H1, one
of a pack of proteins that cling to DNA, sculpting its helices into various
shapes. This chemical transaction appears to be necessary for DNA to collapse
into compact chromosomes ready for the events of mitosis . cdc2 also has
a hand in the construction of the transient structure – a spindle – that
controls the movements of a cell’s chromosomes during mitosis.

Although the details of mitosis vary from organism to organism – in
yeasts, for example, the nuclear membrane does not disintegrate – there
seems to be a broad consensus over the role of cdc2. Five years ago, Melanie
Lee and Paul Nurse demonstrated the point in an extraordinary fashion during
the course of research at the Imperial Cancer Research Fund’s laboratory
in London. Their work used a single-celled organism, the yeast Schizosaccharomyces
pombe – also known as fission yeast to distinguish it from the budding yeast
familiar to bakers and brewers .

Lee and Nurse worked with a mutant strain of yeast in which the gene
responsible for producing cdc2 was defective, causing cell division to be
blocked. Remarkably, they found that they could overcome the block by inserting
a stretch of human DNA into the yeast – a stretch which they subsequently
found to be the human version of the cdc2 gene. The fact that this DNA worked
normally in such an alien environment provided dramatic proof that evolution
has been remarkably conservative in its dealings with the cdc2 gene.

Hundreds of millions of years have elapsed since our ancestors parted
company from those of the yeast – years that saw all the great radiations
of life on earth. And yet our cdc2 protein still bears a striking resemblance
to the yeast’s. When the researchers worked out the structure of human cdc2
and compared it to cdc2 from yeast, they found that they were 63 per cent
similar. This fraternal similarity, in both form and function, opened up
the possibility that organisms of all kinds might control their cell cycles
with the same basic machinery. That idea turned out to be correct. The cdc2
protein is now regarded as a common thread that joins all organisms whose
cells have nuclei – protists, fungi, plants and animals.

Although the cdc2 protein is the centrepiece of the cell cycle, it is
only a part of the overall design. The concentration of this influential
protein does not vary during the cell cycle and yet its effects are unleashed
only at certain times. This implies that it is switched on and off by companion
molecules. Recent research has uncovered a retinue of such molecules, which
hold cdc2 in check, spur it on, or work with it.

One approach focuses on a specific group of proteins whose concentration
varies during the cell cycle. A decade ago Tim Hunt, then at the University
of Cambridge, and his colleagues found the first such protein in developing
embryos of a sea urchin. The protein accumulated throughout the cell cycle,
right up until the moment of division, when it suddenly disappeared. The
same behaviour was repeated in each successive cycle. Hunt and his colleagues
christened the protein cyclin, an apt name for a molecule with such habits.
Tests on other embryos revealed more cyclins with the same basic properties.

The team’s belief that the cyclins would turn out to play a role in
cell division was amply rewarded. Researchers have since found a vast array
of cyclins of various types in all sorts of organisms, including yeasts,
and a consensus of sorts has emerged about their role: they attach themselves
to the cdc2 protein and help activate it. So the agent that sets mitosis
in motion is not cdc2 pure and simple; it is a complex formed from the temporary
marriage of a cdc2 molecule and a cyclin molecule. In this way the ebb and
flow of the cyclins can influence the progress of the cell cycle. (Other
influences also come into play, of which more later.)

A striking demonstration of the role of cyclins has come from an extraordinary
‘stripped down’ version of the cell, whose behaviour has been explored
by researchers such as Andrew Murray and Marc Kirschner at the University
of California, San Francisco. The technique involves preparing extracts
of the eggs of Xenopus laevis (an African frog much used in laboratory work)
and then adding nuclei from sperm. The result is a mixture in a test tube
that will undergo three or more cell cycles at the behest of a researcher,
while remaining amenable to all manner of chemical treatments. Such mixtures
do many of the things that intact cells do, including the production of
proteins.

SHOOTING THE MESSENGER

Murray and Kirschner were able to show that the extracts must be allowed
to make cyclin if they are to undergo mitosis. Genes make proteins via an
intermediary called messenger RNA, so the first step was to destroy all
the mRNA in the extracts, by means of a selective poison, and see what happened.
The extracts were unable to switch into mitosis. The researchers then added
to these disabled extracts a form of mRNA carrying the instructions for
making a cyclin molecule (from sea urchins, as it happened, but cells are
not fussy about the origin of molecules involved in the cell cycle, as we
have already seen). This treatment, which allowed the extracts to start
making cyclin again, was enough to restart mitosis.

Cyclin clearly plays an important part in controlling the cell’s entry
into mitosis. Does it play a comparable role in the exit from mitosis? Recall
that the exit from mitosis is accompanied by a precipitous decline in the
concentration of cyclin. If this decline is prevented, by giving cells an
artificial cyclin that is unusually resistant to destruction, mitosis cannot
be completed. Destruction of cyclin, it seems, is what turns off cdc2, allowing
cells to finish mitosis and get on with the next phase of the cycle. This
vital demolition job is the work of a protein-destroying enzyme (a protease)
which, some researchers believe, could have the additional task of splitting
the chromosomes in two and permitting them to separate. So the cdc2 protein,
with its cyclin, engineers the onset of mitosis; the protease destroys the
engineer and completes mitosis.

Cyclins are necessary for the smooth operation of the mitotic machinery
but they are, again, only a part of the overall picture. Work on a variety
of organisms, notably fission yeast, has uncovered a network of additional
regulatory molecules, whose function is to adjust the timing of mitosis.

How do these regulators work? Ultimately their role is to modify the
cdc2 protein, according to Nurse and his colleagues at the ICRF’s Cell Cycle
Group at the University of Oxford. The cdc2 protein can be kept dormant
if a phosphate group is installed at a certain site within the molecule,
at an amino acid called tyrosine. Removal of the phosphate group has the
opposite effect: it activates cdc2, allowing mitosis to take place. By balancing
these two effects against one another, the cell can control its entry into
mitosis.

A protein called cdc25 is the molecular ‘accelerator’; it removes phosphate
from cdc2, either directly, or through an intermediary – probably the former,
according to recent evidence. Mutant strains of fission yeast that make
too much cdc25 are too keen to divide; those with a faulty version of the
gene cannot enter mitosis. Human cells also make cdc25 and this substance
can activate cdc2 from starfish eggs – another example of the universal
nature of the cell cycle.

Fission yeast also has two molecular ‘brakes’, wee1 and mik1, whose
roles have been explored by Karen Lundgren and her colleagues at the Howard
Hughes Medical Institute in Cold Spring Harbor, New York state. The wee1
and mik1 proteins work against cdc25, making sure that cdc2 retains its
phosphate and so keeping it in check. There are also molecules that regulate
the regulators.

While researchers have been building up this picture of events around
mitosis, there has been a parallel effort to understand the entry into S
phase, the part of the cycle during which the cell copies its DNA ready
for division. In an extraordinary development, it now seems that cdc2 itself
controls entry into this phase of the cycle. This conclusion has emerged
from a number of studies, including work on yeasts which carry a mutated
form of the cdc2 gene and which consequently cannot duplicate their DNA.
The situation in higher organisms, such as frogs and humans, is a little
more complicated. In such creatures a close relative of the cdc2 protein
operates at this point, not cdc2 itself, but the biochemical properties
of this molecule are similar to those of its more celebrated relative.

As this picture has emerged, researchers have wondered whether cyclins
might also have a role to play in the entry into S phase. Their suspicions
have proved well founded, for several different types of cyclin have now
come to light in addition to those involved in mitosis (mitotic cyclins).
Some of these cyclins accumulate in what biologists call the G1 phase of
the cycle and disappear once the cell has begun to copy its DNA. The behaviour
of these G1 cyclins echoes that of the mitotic cyclins around mitosis, so
they may perform a similar job: joining forces with the cdc2 protein, activating
it and perhaps helping to direct it at an appropriate set of targets. Those
targets are still a matter for debate, but they could include proteins that
react to the addition of phosphate by switching on genes whose products
are essential for the copying of DNA.

Much of the early evidence for G1 cyclins came from budding yeast, but
recent developments have again revealed a common thread that runs right
through the living world. A case in point is the work of Daniel Lew, Vjekoslav
Dulic and Steven Reed at the Scripps Research Institute in La Jolla, California.
The team worked with a strain of budding yeast whose genes for G1 cyclins
had been disabled so as to block cell division. They added pieces of human
DNA to these cells, looking for genes that would correct the deficiency.
They found several. Of these, one in particular, cyclin E, is a promising
candidate for a G1 cyclin in human cells.

If these ideas prove correct the result will be a symmetrical picture,
in which cdc2 (or a close relative, depending on the cell in question) together
with various different cyclins, controls all the main transitions of the
cell cycle. It is an elegant, persuasive idea and one which has wide support.
However, many researchers are cautious about drawing too close a parallel
between the two transitions until more evidence is to hand on the role
of cdc2 and its allies in the S phase of the cycle.

That evidence may come all the sooner thanks to recent work on budding
yeast by Stephen Bell and Bruce Stillman of the Cold Spring Harbor Laboratory
in New York state. Last month the researchers reported in Nature that they
had isolated a complex of proteins that bind to chromosomes at the special
sites where the copying of DNA begins. The implication is that these proteins
may set the copying process in motion and hence play a key part in touching
off the S phase of the cell cycle.

MOLECULAR CHECKPOINTS

The story so far is of a team of molecules whose business is to control
the main transitions of the cell cycle. But how are their activities coordinated?
The answer depends on checkpoints, an idea proposed by Leland Hartwell and
Ted Weinert of the University of Washington. At a checkpoint, the cell
makes sure that one process is complete before authorising the next. Such
mechanisms are necessary because mistakes in timing are extremely costly.
Cells cannot afford to begin mitosis, for example, until they have finished
copying their DNA. Any such negligence would mean that their progeny would
not have complete copies of the genetic library. Similarly, the chromosomes
cannot be permitted to split in two at mitosis until the spindle is assembled
correctly.

Checkpoints are so neatly woven into the fabric of the cell cycle that
their presence only becomes noticeable when the cycle is disrupted. If
researchers add poisons that block the duplication of DNA, for example,
the checkpoint mechanism senses that something has gone wrong and forces
the cell cycle to stop. Work on frogs by Mary Dasso and John Newport of
the University of California at San Diego suggests that this effect is achieved
by keeping cdc2 phosphorylated – and a similar story has emerged from work
on fission yeast.

As well as keeping abreast of internal events, cells must also monitor
the world outside their borders. It would be fatal for a yeast cell, for
example, to try to divide if sufficient nutrients were lacking. So the cell
must assess its chances and suspend the cell cycle if times are hard. In
budding yeast there is a point in G1 called ‘START’, which cells pass only
if they are sufficiently large and well fed. If all is well, they begin
copying their DNA and go on to divide. If conditions are unpromising, or
if they get wind of a prospective sexual partner, they come to a halt at
START.

FROM HARMONY TO MUTINY

Just as yeasts assess their future prospects at START, human cells cultured
in the laboratory make a similar decision at an analogous stage called the
restriction point. However, human cells, like those of other multicellular
creatures, are not their own masters. Because they must live in harmony
with their neighbours within the body, they are subject to many more controls
over their behaviour than a yeast cell. For example, human cells in culture
need tiny amounts of chemicals called growth factors if they are to pass
the restriction point successfully. Without growth factors they simply become
quiescent and stop dividing altogether. If cells in the body refuse to stay
in this state – if they keep dividing when they ought to desist – the stage
is set for the development of a tumour. Many genes that cause tumours when
they misbehave (oncogenes) affect this transition from quiescence to proliferation,
often by giving cells the misleading impression that they should divide.

Other genes that have been implicated in cancer may turn out to have
even closer links with the cell cycle. The oncogene bcl-1, for example,
has come under intense scrutiny because researchers have discovered that
it makes a protein that belongs to the cyclin family. Andrew Arnold and
his colleagues at Harvard Medical School have found that this gene is overly
active in cells taken from certain benign tumours of the parathyroid glands
(organs in the neck whose function is to make a hormone that controls the
body’s use of calcium). The normal function of the gene has yet to be decided,
but it may operate as a regulator of the cell cycle.

Another gene with close links to the cell cycle is the retinoblastoma
gene, which gets its name from the rare tumours of the eye that occur in
some children who inherit a faulty version of the gene. The normal retinoblastoma
gene has earned a reputation as a tumour suppressor – a gene whose role
is to keep cells in check – and there has been a surge of interest in how
it works over recent years. In the normal cell the protein made by the gene
resides in the nucleus, where it binds avidly to many other molecules and
so inactivates them – it has been likened to a molecular sponge that soaks
up its victims. Among its targets are molecules that turn on genes required
during cell division. In the test tube, cdc2 and its allies can phosphorylate
the retinoblastoma protein – an event which makes it release the molecules
to which it has been clinging.

So the retinoblastoma protein could play an important part in the cell
cycle by gripping and then letting go of other molecular activists. It
is an appealing story and one that would begin to explain the links between
the retinoblastoma gene and cancer; if the gene failed to work properly,
then cells might be freed of one important control mechanism and so more
likely to divide. Another link involves certain viruses that cause tumours.
These make proteins that avidly bind to the retinoblastoma protein – forcing
it to let go of its proper quarry and allowing cells to start dividing.

In a sense all work on the cell cycle is groundwork for the study of
cancer. Yeasts, sea urchins and frogs all provide information relevant to
humans because their cell cycles are so similar to ours. That accord is
the legacy of an unknown creature which lived untold millions of years ago,
but which had already worked out an ingenious and robust solution to the
challenge of cell division.

Stephen Young is a freelance science writer based in Wales.

* * *

Imagine the fate of a new cell, one just born by fission and destined
to go on dividing if circumstances permit. A tranquil period of growth
follows the exertions of division, during which the cell must decide whether
or not to divide again. If it proceeds, then it must first duplicate the
genetic library of DNA molecules in its nucleus so that each of its progeny
can be allocated a complete copy. When the copying is complete, another
quiet period ensues before the cell finally switches on its programme for
fission.

There then follows a dramatic series of events. The DNA molecules in
the cell’s nucleus condense into distinct chromosomes, the membrane around
the nucleus dissolves, an array of filaments – the spindle – appears and
the chromosomes line up in the cell’s centre, attached to the spindle. As
if at some prearranged signal, each chromosome breaks into two identical
pieces, which make off in opposite directions. The cell narrows at its
waist and pinches off two daughter cells, while nuclear membranes reform
around the chromosomes. The cycle is now complete.

These events make up the process of mitosis. A distinct, but similar
two-stage process, meiosis, generates sex cells such as sperms and eggs,
which have half the genetic content of their progenitors.

The phases of the cycle each have an abbreviated name. G1 is the period
immediately after cell division (G stands for gap); S phase is the next
stage, during which the cell synthesises a copy of its DNA. The gap between
the end of S phase and the onset of mitosis is known as G2. Cells may also
leave the cycle altogether and spend long periods without dividing. Mitosis
is itself divided into four phases: prophase sees the appearance of the
condensed chromosomes, which then line up in the middle of the cell at metaphase;
at anaphase the chromosomes divide and at telophase they form new nuclei.
The entire cycle, from one division to the next, takes anything from two
hours in yeasts to about 20 hours in vertebrates. Temporally, mitosis takes
up about 40 per cent of the cycle in the cells of early embryos, but less
than 10 per cent in most other cells.

* * *

2: The menagerie that made it all make sense

Research on the cell cycle owes much to a small inner circle of organisms,
notably yeasts and an African frog Xenopus laevis. Two yeasts have found
favour with geneticists: budding yeast (the one used in brewing and baking)
Saccharomyces cerevisiae and fission yeast Schizosaccharomyces pombe, which,
as the name implies, reproduces by a more equitable process of bisection.
The two yeasts share a similar level of organisation and both can turn sugar
into alcohol, but the two are not close relatives in evolutionary terms.

Thanks to work on yeasts, geneticists have identified a huge number
of genes involved in the cell cycle. The standard method is to search for
mutant strains whose behaviour varies with temperature. Such strains thrive
at moderate temperatures, but divide abnormally, or stop dividing altogether,
when the temperature rises. By noting the point in the cell cycle at which
the mutated gene takes effect, researchers can obtain clues about the normal
role of that gene in cellular affairs.

The African frog has made its contribution in the biochemistry laboratory.
For some years researchers had known that the eggs of this creature, when
mature, contained some chemical that could induce other eggs to ripen –
a process in which they enter M phase (meiosis in this case rather than
mitosis) and then wait to be fertilised. The chemical was called maturation
promoting factor, or just MPF. MPF turned out to have all the right properties
for a controller of cell division. After much painstaking work, it transpired
that MPF was made up of two components – a version of cdc2 and a cyclin.
Genetics and biochemistry came together with the same elegant story.