91av

CANCER strikes one in three people in the developed world, and almost all of us are touched by it sooner or later. To find out what causes this devastating disease, scientists are focusing on the genes that choreograph cell behaviour.

Cell division is a crucial part of life from the moment of fertilisation to the day we die. It drives the development of a multicellular organism from a single-celled zygote– a fertilised egg– and in adult life it helps maintain and repair complex tissues and organs. But here’s the rub. Our bodies rely on the cooperation of millions of cells, each of which must perform its own specialised role for the good of the whole. If just one cell starts to pursue its own selfish interests and divides uncontrollably, it could spell doom for the rest of the body. Hence, the body’s need for cell division carries with it the risk of cancer.

Cells that divide willy-nilly produce tumours, which may be benign or malignant. Benign tumours are not cancer. Some may eventually become malignant, but if removed they don’t usually come back or spread to other parts of the body. Malignant tumours are cancer, and can develop in nearly any organ of the body. They grow, then invade and damage nearby tissues, perhaps spreading (metastasising) via the bloodstream to distant sites in the body.

Cancer is caused by accumulated, harmful genetic changes or mutations in the genes that control cell behaviour. To find ways to prevent cancer and treat it, we first need to understand how healthy cells keep a tight control on their growth and division. After several decades of intensive research, scientists have uncovered much of the mechanism that cells use to govern their division. The cell cycle is a complex regulatory loop through which all cells must travel in order to divide (see “Every cycle needs brakes”). It ensures that DNA is faithfully copied, and that the replicated chromosomes make their way into the daughter cells.

The checks and fail-safe mechanisms built into the cell cycle are powerful devices, but there are other controls too. The body reduces the chances of rogue cells appearing by limiting the number of cells capable of dividing. After they have finished cycling, many cells differentiate into specialised cells (such as nerve or muscle) that cannot divide further. Only a select band retain their powers of division, which they use to repair and renew tissues. As a result, they are more likely to become cancerous. For example, many types of leukaemia result from genetic changes in immature blood cells that are still able to divide.

Another check on the development of cancer is cell suicide or apoptosis. If a cell is severely damaged or abnormal, it will self-destruct (see “Secret language of cells”, Inside Science No. 148). This happens in a carefully controlled way, and is a common response to severe chromosome damage. The tightly controlled cell cycle, differentiation and apoptosis all work together to help prevent cells from dividing in a haphazard fashion. But when mutations damage the genes that control these processes, a cell takes the first steps towards cancer: it becomes tumorigenic.

One mutation isn’t enough to cause cancer. The likelihood of cancer increases as a person grows older, and the way in which this risk grows suggests that between four and seven mutations in a single cell are needed. When scientists study DNA from tumour cells, they find many kinds of mutation at multiple sites, ranging from subtle alterations in single bases to obvious changes in the number of chromosomes.

Most cancers are caused by a combination of lifestyle, genetic and environmental factors. Although there’s nothing we can do about the genes we were born with, we can adopt healthier lifestyles. Cigarette smoke contains hundreds of chemicals that damage DNA, so giving up smoking is the best thing a smoker can do to reduce the risk of developing one of several common cancers in later life. Avoiding sunburn helps prevent the often fatal cancer malignant melanoma. And eating lots of fruit and veg, among other benefits, shortens the length of time it takes for food and any DNA-damaging compounds it contains to pass through the gut, so reducing the risk of colon cancer (see “Food, glorious food”, Inside Science No. 104).

Unfortunately, we will probably never be able to banish cancer completely. A low level of spontaneous mutation (for example, caused by mistakes in DNA replication) is unavoidable. Without it, evolution could not happen. Certain features of our environment, such as low levels of ionising radiation or ultraviolet light from the Sun, increase the rate of spontaneous mutation. What’s more, some unlucky people are born with faulty genes that predispose them to cancer. The healthy versions are involved in repairing DNA, others in cell cycle “checkpoints”. In general, the key genes involved in cancer control cell division and fall into two main classes: oncogenes and tumour suppressor genes.

Oncogenes were first spotted in viruses that cause cancer in animals. When scientists introduced these genes into cells, the cells started to behave like cancer cells, dividing uncontrollably. But an even bigger surprise was in store. It turned out these viruses had picked up genes involved in cell division from their hosts, and during this process the captured genes had acquired harmful mutations. The healthy counterparts of these viral oncogenes are called proto-oncogenes, and all play a role in the normal functioning of the cell. It is only when they go wrong that they become oncogenes. About a hundred have now been identified. One famous example is the Ras gene (see “Smart proteins”, Inside Science No. 139). Normally, the protein it codes for helps relay signals from the cell membrane to the nucleus. But a mutated form of Ras was discovered in a retrovirus that causes cancer in mice and rats. About a quarter of all cancers contain a mutated version of Ras.

Investigating what proto-oncogenes do in healthy cells helps researchers to get a handle on oncogenes. Proto-oncogenes are categorised according to the activities of their protein products. Many are involved in receiving signals that stimulate cell growth. Others, known as cyclins and Cdk enzymes, regulate the cell cycle. A third group regulates apoptosis.

Normally, the protein encoded by a proto-oncogene is only produced – or is only active – when cells receive the right signals. Mutations that transform proto-oncogenes into oncogenes effectively flip a switch that turns the gene or protein permanently “on”, even if no stimulatory signals are present. This forces the cell to grow even if the second copy of the gene is normal. So biologists call these gene defects dominant, “gain-of-function” mutations.

Permanently switched on

There are various ways in which signalling pathways that stimulate growth can become permanently switched on (see Figure). Some tumorigenic cells synthesise their own growth-promoting signals to keep these pathways firing. In others, mutations in the receptor proteins that cells use to receive signals switch the receptors constantly “on”, fooling the cell into behaving as if signalling molecules were always bound to it.

The proteins that relay signals from the receptors to the nucleus can also be affected. The Ras protein is normally only switched on when receptors on the cell membrane bind to growth factor molecules. But in many cancers, Ras is always switched on, so the cell divides as if it were constantly being stimulated by growth factor.

Finally, things can go wrong at the end of the signalling pathway: the nucleus. Here, proteins called transcription factors normally respond to relayed growth signals by activating growth-promoting genes (see “Control centre”, Inside Science No. 122). In cancer, transcription factors may keep these genes switched on. For example, mutations in genes that make the myc family of transcription factors have been found in many common cancers.

But oncogenes can’t cause cancer on their own. The “brakes” also have to be released by mutations in tumour-suppressor genes. In a healthy person, tumour-suppressor genes halt the same processes that are switched on by proto-oncogenes. Some suppress cell growth and division, while others stimulate apoptosis. Mutated tumour-suppressor genes have been found in many cancer cells, but unlike oncogenes, both copies have to be affected to contribute to the development of cancer. It’s a bit like having one brake left working on your bicycle: one working gene can still put a stop to uncontrolled cell division. So these gene defects are known as recessive, “loss-of-function” mutations.

About 20 tumour-suppressor genes have been identified. The first to be collared was the RB gene. Losing both copies of RB causes retinoblastoma, a cancer of the retina that usually affects young children. Many if not all signals that inhibit cell division are channelled via the retinoblastoma protein (pRB), or two closely related proteins. When pRB is switched on, it inactivates certain transcription factors, stopping the cell cycle.

One important tumour-suppressor gene is BRCA1, involved in some familial cases of breast cancer (see “What causes breast cancer?”). Another, p53, seems to have a finger in every pie (see Figure). Mutations in p53 are found in as many as half of all tumours. In healthy cells, the p53 protein halts the cell cycle at the G2/M checkpoint until DNA damage has been repaired. When p53 is mutated, however, the cycle rolls on without stopping for repairs, increasing the risk that other cancer-causing mutations will occur.

Another way p53 protects against cancer is by forcing a cell to commit suicide if its DNA is damaged beyond repair, or if the cell is behaving abnormally. If mutations knock out p53, the chances of apoptosis are reduced. The protein’s central role in preventing cancer has earned it the nickname Guardian of the Genome. People who have a genetic disease called Li-Fraumeni syndrome have faulty p53, and as a result most of them develop cancer by the age of 30.

It’s clear that mutations in genes involved in regulating cell division play a key role in cancer. But how do they arise in the first place, and what can be done to prevent them?

There are three main causes of DNA damage. The most well known are environmental agents (including ultraviolet light in sunlight, ionising radiation and “genotoxic” chemicals). The second group are damaging by-products of metabolism (such as the highly reactive “free-radical” molecules generated during respiration). Finally, some of the chemical bonds in DNA break spontaneously. Repairing these sorts of damage is crucial, and cells devote a whole set of specialised proteins to the job. Mutations in genes involved in these repair systems increase the risks of certain cancers dramatically. For example, a mutation in one of the DNA repair genes causes the rare disease xeroderma pigmentosum. People with this disease are a thousand times as likely to get Sun-induced skin cancer, and are also prone to internal tumours.

DNA damage disrupts many cell processes. For example, double-stranded DNA breaks induced by X-rays or chemicals may delete or add extra chromosomes. They can also cause “translocations” – when bits of one chromosome break off and join onto others – which may turn proto-oncogenes into oncogenes. They are a common cause of cancers of bone marrow (leukaemia) and lymph nodes (lymphoma).

Unravelling all the cellular pathways that can be affected by cancer-causing mutations is a huge undertaking. Fortunately, just as nearly all mammalian cells use similar molecular machinery to regulate their growth, division and death, so all cancer cells appear to share a common set of characteristics.

A fine balance between signals that stimulate cell division and those that inhibit it ensures that cells divide just enough to satisfy demands for growth and repairs but no further. Many mutations in proto-oncogenes keep pathways that stimulate growth permanently active. Equally important is tumour cells’ ability to ignore anti-growth signals, which are usually transmitted by tumour-suppressor genes.

Always keep a few spares

Research now shows that many molecules involved in these signalling pathways perform the same tasks. So if one protein stops working, another can compensate. For example, if tumour cells have lost some proteins involved in apoptosis, they will probably possess others that could fulfil the same role. A possible strategy for tackling cancer is to develop drugs that boost the apoptotic pathways that are still working or that act as a bridge between two pathways.

There’s one more mechanism that cancer cells must overcome to carry on dividing selfishly– a “clock” that may limit the number of divisions. In a process called senescence, normal cells stop after they have divided between 60 and 70 times (see “Age-old story”, Inside Science No. 117). This brake seems to be missing in most types of tumour cells that are grown in culture, suggesting that they have somehow become immortal. The structures that limit the number of times a cell can divide lie at the ends of chromosomes. They’re called telomeres and comprise a short, six-base sequence repeated several thousand times (see Figure). Telomeres play a crucial role protecting the ends of chromosomes. The mechanics of DNA copying mean that the enzyme DNA polymerase is physically incapable of copying a DNA strand right to the end. So with each cell division, between 50 and 100 base pairs of DNA disappear. If there were no telomeres, key genes and DNA at the ends of chromosomes would rapidly be lost. Eventually, telomeres become so short they can no longer protect the ends of chromosomes. The unprotected ends join up, resulting in chromosomal chaos and cell death.

Nearly all types of malignant cell can maintain the length of their telomeres. Most do this by boosting the expression of telomerase, an enzyme that rebuilds telomeres. Scientists hoped that a drug to block telomerase would be a magic bullet to combat cancer. But the importance of telomeres in cancer is still disputed. Some researchers believe cell senescence only occurs in cell cultures in labs. Growing cells outside their natural environment in the body may make them more sensitive to senescence. For developing cancer cells, on the other hand, overcoming senescence might not be too difficult.

A more promising approach may be to starve tumours to death. To survive, nearly all cells must be within 100 micrometres of a capillary to absorb enough oxygen and nutrients. To this end, the growth of blood vessels and tissues is closely coordinated during embryonic development and after injury in adulthood. Once a tissue has formed, however, the growth of new blood vessels– known as angiogenesis– is tightly controlled. So cells in a developing tumour must somehow stimulate new blood vessel growth.

Angiogenesis, like cell growth and division, is controlled by a balancing act between promoting and inhibiting signals. More than 50 proteins involved in this process have been found so far. Tumour cells appear to flip an “angiogenic switch” in the early or middle stages of their development by changing the balance of angiogenic inducers and inhibitors they produce.

As with so many other anti-cancer mechanisms, the p53 protein seems to be important, as it stimulates the production of a protein that inhibits angiogenesis, thrombospondin-1. When the p53 gene is mutated, levels of thrombospondin-1 fall, and blood vessel cells are allowed to proliferate. Commandeering a blood supply in this way allows the tumour to grow rapidly from a small group of cells to a large mass.

Angiogenesis is an important focus of research because most if not all tumours need it to thrive. Crucially, angiogenesis is almost unique to tumour cells, so targeting it with drugs is less likely to harm normal cells. A number have been developed and are now in clinical trials.

Even when well supplied with blood vessels, a tumour eventually runs out of nutrients and space. During the development of most cancers, “pioneer” cells move out of a tumour, invade neighbouring tissue and travel to distant sites to found new colonies. These metastases are the cause of 90 per cent of cancer deaths. The body’s mechanisms for preventing tissue invasion and metastasis are not yet well understood. But somehow, cells loosen the tethers that anchor them to their environment, and secrete enzymes that allow them to invade tissue. Many proteins that are normally involved in anchoring cells or wound healing are altered in tumour cells.

As we have seen, a vast number of genetic changes contribute to the development of cancer. Studying these mutations, and working out how normal cells control their growth and division, should lead to advances in treatment and prevention. One day it may even become possible to obtain a readout of all the mutations present in a particular tumour, then treat the patient with a personalised suite of drugs designed to compensate for those defects. Conversely, a drug that mimics the effects of p53 could treat a wide range of tumours. Drugs that target other processes common to all cancer cells, such as angiogenesis, could also save many lives.

Billions of dollars are spent every year trying to discover why cells turn traitor and become cancerous. As well as providing extraordinary insights into the workings of cells, these efforts are beginning to pay off with a steady improvement in the success rates of treatment. There may be a long way to go, but we’re starting to get the measure of this merciless disease.

The cell cycle, through which all actively dividing cells must pass, has four distinct phases (see Figure). Immediately after it has undergone cell division or mitosis, a cell pauses in G1 phase (the G stands for gap). This allows it to grow and await signals from its environment giving it the green light to enter S phase. During S phase (S for synthesis), nuclear DNA is copied. Then the cell pauses in G2 phase, checking that all DNA synthesis is complete and accurate. Only then will it take the plunge into M phase (mitosis) and divide.FIG-mg23516702.jpg

Two main families of proteins control the cell cycle, cyclins and cyclin-dependent protein kinases (otherwise known as Cdks). Cdks are enzymes that activate the cell cycle by switching on key proteins. They are themselves controlled by cyclins, which stick to them and switch them on or off.

The cycle contains brakes, or checkpoints, that monitor the cells and their environment for any problems before allowing division to proceed. There are two major checkpoints, one before the transition to S phase, the other before the transition to M phase. If any problems are detected– such as damaged DNA– division is stopped in its tracks.

Breast cancer is very common in the West, and is now the most common cancer in Britain, even though it almost always affects only women. It has been the focus of much research over the past two decades, culminating in the identification of two genes, BRCA1 and BRCA2, involved in hereditary breast cancer.

Only about 5 per cent of breast cancer cases run in families. But inheritance of faulty BRCA1 and BRCA2 accounts for about 45 per cent and 25 per cent of these families, respectively. Women from affected families can be screened to see if they carry mutant versions of one of these genes. A woman with such a mutation in her genome is very likely to develop the disease and may be advised to take tamoxifen, a drug often used to treat breast cancer but which may also help to prevent it.

The hormone oestrogen appears to increase a woman’s risk of breast cancer by promoting cell growth and proliferation, and tamoxifen stops it binding to its receptors. However, taking the drug carries other risks: it increases a woman’s chances of developing uterine cancer and blood clots, so it is only used in high-risk cases.

Even though the inheritance pattern of BRCA1 and BRCA2 could be clearly traced through families known to have an increased risk of breast cancer, it took a huge effort to identify these genes. But 95 per cent of breast cancer sufferers do not have a family history of the disease. Recently, though, researchers discovered six more genes that might predispose women to breast cancer. These genes are linked to another disease called Fanconi anaemia, which makes people more susceptible to cancer.

No one knows how important these genes are in causing non-hereditary breast cancer. It is likely that individual genes make only a small contribution, and that cancer is caused by environmental factors and the additive effects of a number of genes that are harmless by themselves.