Dr Graham Currie, director of research at the institute, says: "Our work is part of an international collaborative effort. The sorts of discoveries that our scientists are making will have a major impact on the detection and diagnosis of cancer. This will aid us in identifying new targets for drug development - which we hope to develop with the help of the pharmaceuticals industry."
Scientists already know a great deal about what goes wrong when cancer develops. Normal growth of cells is controlled by two sets of genes. One set - the "on switch" - produces proteins that tell a cell to grow and divide. The second - the "off switch" set - makes proteins that tell the cell to stop dividing.
In many malignant tumours there are faults or mutations in one or both of these sets. Some of these faults result in the switch that controls cell division becoming stuck in the "on" position. Genes with this type of fault are called oncogenes.
In most cancers, cells in the tumour have also lost their copy of the gene that tells the cell to stop dividing - the switch that turns off cell growth no longer works. These genes, whose absence results in cancer, are called tumour suppressor genes.
In cells that have lost a tumour suppressor gene, or which have gained an oncogene, cell division can run riot. But in most cancers, a whole series of changes and mutations are needed before a cell takes on the characteristics of a cancer cell.
As people age, the genetic material of their cells accumulates random mutations, perhaps in response to environmental factors such as smoking, diet or radiation. This is why cancer tends to affect mainly older people - because it takes time to accumulate enough errors, and in a certain order, for cancer to develop.
At the Marie Curie Research Institute, seven teams of scientists are adding to this bank of knowledge. One is studying the ways in which viruses infect cells - a process that involves hijacking the cell's normal pattern of growth. Another team has been examining the mechanism by which normal cells divide. And a third group has been combing through samples of tumour cells to try to identify hitherto unknown oncogenes and tumour suppressor genes.
For example, Dr Peter O'Hare and his colleagues have been studying how the herpes virus switches genes on and off in the cells that it infects. His team has found that, after the virus has invaded the cell, a protein from the virus binds to one of the cell's proteins - one which is normally involved in controlling the genes that must be activated for cell division to take place.
The researchers have discovered that this binding is only possible because the cellular protein assumes a different shape in cells that are infected by the virus. This discovery means that it may one day be possible to design drugs that would selectively attack cells that are dividing in an abnormal way.
Dr O'Hare says: "By exploring how this viral protein interacts with the cell's proteins, we are uncovering the basic mechanisms about how proteins control genes.
"This is helpful to our knowledge about cancer, since many proteins involved in triggering cancer exert their effect by controlling genes. The more we find out about these aspects, the closer we are to finding out what goes wrong in cancer."
A second team, led by Dr Rob Cross, is scrutinising the process of normal cell division. When a cell divides, it duplicates its chromosomes, which contain the genetic material. Half of the chromosomes then travel to one end of the cell and half to the other. But how do the chromosomes move across the cell? That is one of the questions that Dr Cross and his team are trying to answer.
During cell division, the cell builds a specialised network or "scaffolding" made up of tiny tubes called microtubules. This surrounds the nucleus. It is now known that individual molecules, powered by the cell's chemical energy, pick up and transport the chromosomes along this network of microtubules. Many different versions of these "molecular motors", as they have been called, have been discovered, which move at different rates and in both directions along the microtubules.
Dr Cross and his colleagues have shown that some molecular motors have two sites which can bind to the microtubules - but that each molecule can only use one of its two binding sites at any one time. It is, he explains, rather as though the molecules have two feet that allow them to "walk" along the microtubules, but can only put one of the feet down at any given moment.
Dr Cross says: "If we are to tackle cancer, we first need to understand normal processes of cell division at the molecular level." One day, he adds, it may be possible to design new drugs that will interfere specifically with these stages of cell division in the cancer cell.
A third team at the institute is searching for new examples of genetic mutations that cause bladder cancers. Dr Maggie Knowles, who is leading this work, explains that all tumours arise because normal cells have accumulated several mutations; in the case of bladder cancer, six to eight mutations are thought to be needed.
"Some of the mutations present in bladder cancer are known - our objective is to track down the ones that are not yet known," she says.
To do so, she and her team have been examining the genetic material present in hundreds of samples of bladder tumours that have been removed from patients. They are also about to embark on a study to find out whether particular mutations are associated with tumours that resist treatment and spread rapidly.
About 20 per cent of people diagnosed as having bladder cancer have tumours which have already grown into the muscle of the bladder. Many of them respond poorly to treatment. The remaining 80 per cent have superficial tumours which have not spread to the muscle of the bladder and can be completely removed. They are likely to respond well to treatment and live for many more years. But, in about one in five of them, the tumour will recur and penetrate the bladder wall. These patients are also difficult to treat.
Dr Knowles and her colleagues are hoping to be able to find out whether tumours that respond poorly to treatment, tumours that spread, and tumours that recur, are associated with particular rogue genes. If they are, knowing which type of tumour patients have when they are first diagnosed could be helpful: doctors might be able to treat certain patients with more powerful drugs, or follow them up more often.
Dr Currie adds: "With modern gene technology, it should be possible to identify these tumours at an early stage by screening the urine for cancer cells."