p53

by Sue Armstrong

  This was clear evidence that ‘mending’ mutant p53 was possible, and Selivanova was very excited. ‘It was fantastic,’ she recalled with a smile. ‘I wanted, of course, to go out and cure tumours – at least in mice!’ But the peptide proved unworkable as a drug: in a living organism these scraps of protein are poor at entering cells and are quickly broken down and recycled. What was needed was a chemical compound, a small molecule that would perform the same tricks.

  Working with Wiman and a new postdoc, Vladimir Bykov, Selivanova screened thousands of compounds from a library of possible candidates provided by the US National Cancer Institute. In 1999, the three discovered a molecule they named PRIMA, an acronym for ‘p53 reactivation and induction of massive apoptosis’ that appealed to the scientists because it also implies something that is first class. Experimenting with the molecule they found, to their gratification, that it is effective with a wide range of p53 mutations, and therefore potentially useful in treating many different tumour types. They published their results soon afterwards, ‘and PRIMA attracted a lot of media attention,’ Wiman, a tall, soft-spoken Swede, told me when I visited him at the Karolinska Institute. ‘I was on TV and in newspapers and journals around the world, because the concept of having a small molecule that will make the cancer cells commit suicide is so appealing.’

  So how does PRIMA-1 work? Wiman and Bykov discovered, to their surprise, that both PRIMA-1 and a very similar compound known as PRIMA-1 MET are converted to another compound that binds tightly to p53 protein and refolds it. ‘This was a very important and exciting finding since it gave us a better understanding of how these compounds can reactivate mutant p53,’ said Wiman.

  In partnership with the Karolinska, he, Bykov and Selivanova set up a small biotech company to develop PRIMA-1 MET for the market. It has been a steep learning curve. ‘As scientists you need to work with company people – a completely different culture,’ commented Wiman. ‘Suddenly there are people in suits, board meetings, talk about money . . . And then you interact with clinicians too. So there are three worlds and you all have to work together all the way through. We had no idea what was involved when we started.’

  The company has taken PRIMA-1 MET through a phase 1 clinical trial, which involved 22 patients with cancer of the prostate and blood being given a short course of the drug by injection. Phase 1 trials are designed to test patients’ tolerance to a potential new drug, and to find out how it disperses in the body and how long it persists. The results, published in 2012, were promising: they showed that PRIMA-1 MET is not toxic and that side effects – including dizziness and fatigue – are mild.

  Phase 2 clinical trials, designed to prove that the drug works in people as it does in the lab and in animal models, are the next step. The Karolinska researchers and their company are hoping to test PRIMA-1 MET in combination with conventional chemotherapy in cancer patients, where the two drugs are expected to act in synergy: while PRIMA-1 MET restores mutant p53 to its normal shape and function, the other drug will cause DNA damage that sends clear signals of stress to trigger apoptosis. But this is where the hurdles en route to the clinic really begin: staging a phase 2 trial for PRIMA-1 MET is likely to cost millions of euros, said Wiman. A tiny biotech company like theirs needs to find a partner with serious money to invest.

  A doctor in Denmark who has seen the effects of PRIMA-1 MET on lung-cancer cells in the lab and in mice is so excited by the drug that he has offered to set up a trial himself. The MD Anderson Center, too, is keen to run a phase 2 trial of PRIMA-1 with cancer patients. But finding the funds for all these activities remains a huge challenge, and so far Big Pharma has shown little interest in small molecules that restore normal shape and function to mutant p53 because it is still not entirely clear how they work.

  SMART THERAPY

  A problem that dogs the field of cancer therapy is the issue of drug resistance. The extreme instability of cancer cells and the terrible speed with which they pick up mutations mean that they are likely to find a way round a targeted drug before too long, no matter how clever the design, as the cells that survive the initial onslaught of treatment give rise to equally hardy clones that grow into resistant tumours. To minimise the prospect of failure, oncologists typically treat their patients with a combination of therapies – either a cocktail of drugs, or a drug together with radiotherapy. With this strategy, cancer cells that are not affected by one drug should be hit by the other.

  Researchers are also investigating the use of drug combinations in a novel kind of p53-based treatment called cyclotherapy. One of the biggest shortcomings of conventional chemotherapy, which is ‘cytotoxic’ (meaning that it’s a cell poison) and targets the body’s rapidly dividing cells, is that it is indiscriminate. Cancer cells are by definition fast-dividing, but so too are the cells in the hair follicles, lining of the gut and bone marrow, which sustain collateral damage during chemotherapy. But hair loss, nausea, diarrhoea, anaemia and depletion of the immune system are not just distressing side effects for the patient, they are potentially deadly and they limit the dose of cytotoxic drugs the oncologist can administer to attack the cancer.

  The principle behind cyclotherapy is that patients be given one drug to ‘protect’ the healthy cells from the chemotherapy by temporarily stopping them from dividing, while their cancer cells (which continue to divide and therefore remain targets of the chemotherapy) are blasted with a second, cytoxic drug given simultaneously. With healthy cells protected, the theory goes, the oncologist will be able to increase the dose of the cytotoxic drug and thus maximise its potential to wipe out the tumour. But even if it falls short of clearing the cancer completely, cyclotherapy will make chemo a lot less unpleasant for the patient because it will limit the side effects by sparing the cells of the hair, gut, bone marrow, etc. from the full force of treatment.

  In laboratory tests, Nutlin is looking the most promising of a number of similar drugs used to protect the healthy cells. However, cyclotherapy is still a few years from the clinic. Researchers still need to work out which combinations of drugs work best, with what tumour types and in what quantities. The arrest of healthy cells mid-cycle must be reversible: too high a dose of the protective drug, for example, could cause healthy cells to senesce, but too low a dose might not arrest them for long enough to protect them from the cytotoxic drug. And no one is sure yet how well cyclotherapy works in living organisms: as of 2012 there was only one published report of an experiment in mice. However, one of the main constraints on cyclotherapy is the fact that neither Nutlin nor any of the other potential ‘protectors’ has yet been approved for use in the clinic in its own right.

  NEW LIGHT ON OLD TREATMENT

  Despite the frustratingly slow progress of brand new p53-based therapies, scientists’ understanding of p53 is already beginning to have an impact on the treatment of cancer patients: it enables oncologists to make more rational decisions about the use of conventional chemo- and radiotherapy.

  Chemotherapy has a colourful, if unfortunate, history. Its origins go back to World War I, when the Germans used mustard gas in the trenches of Europe to devastating effect. The use of chemical weapons was banned by the Geneva Protocol of 1925, but not the possession of such weapons, and the Americans continued to develop and stockpile them. In December 1943, a US cargo ship, the SS Harvey, secretly carrying mustard-gas bombs to the Mediterranean war front, was sunk in a German raid on the port of Bari, southern Italy, and a cloud of gas drifted over the city. No one knows how many civilians were affected, but more than 600 military personnel were hospitalised and 83 died. During autopsies of the victims, pathologists found evidence that the normally fast-dividing cells of the bone marrow and lymphoid tissues had been suppressed. From this observation came the idea that perhaps such an agent could be used to attack the rapidly dividing cells of cancer.

  Soon scientists were doing experiments with mustard gas in mice. Encouraged by the results, they moved cautiously on to testing the agent in humans. The first h
uman subject was a patient with lymphoma – cancer of the lymphoid tissue – and his doctors observed with delight the dramatic shrinkage of his tumours after administration of the drug. Unfortunately the effect was short-lived, but it galvanised the cancer community: here at last was a new way of treating the disease. For many centuries, surgery had been the only option for getting rid of tumours, and patients’ long-term survival chances were minimal.

  Over the decades since, many different cytotoxic drugs have been developed – all on the same principle, that they are poisonous to cells. But while chemotherapy has been found to work wonderfully well in some tumours, it does not work at all in others. And in some it works for a while and then stops. Why is the response so varied? Until p53 research began offering clues, no one had an answer. Today we know that both chemo- and radiotherapy work not by killing cancer cells directly, in a sledgehammer kind of way as had been assumed, but much more subtly: typically, these therapies work by inducing cancer cells to commit suicide in response to damage of their DNA – the normal response to cell stress, mediated by p53.

  Scott Lowe, whom we met in Chapter 12 creating mouse models and making groundbreaking discoveries about apoptosis and p53, was one of the first to recognise the tumour suppressor’s central role in conventional therapy. To recap, Lowe subjected the highly sensitive thymus glands of his mice to radiation and discovered that the cells with normal, functioning p53 died very quickly by apoptosis, but cells with mutant or no p53 were resistant to radiation and survived. Confirming p53’s role in apoptosis in response to radiation set Lowe wondering more generally: could p53 be responsible for the effect of radiation – and perhaps cytotoxic drugs also – in cancer therapy? As an idea, it was incredibly simple, and so obvious in retrospect, but revolutionary at the time.

  ‘Here was a situation where the hypothesis was that if p53 is mutant, the standard chemotherapy drugs are less likely to work,’ explained Lowe. ‘In the case of leukaemias and lymphomas what we would have predicted holds true. But now, 17 years of subsequent research says of course it’s more complicated than that.’

  In leukaemia and lymphoma cells, p53 is almost always normal and, as one would expect, these cancers are highly sensitive to chemo- and radiotherapy. But in solid tumours (cancers of the organs rather than the blood), the picture is much less predictable – and sometimes it is counter-intuitive. In some types of cancer, cells with mutant p53 are more responsive to cytotoxic drugs than are cells with normal p53. This is the case with glioblastoma, an aggressive tumour of the brain, for example. So what is going on?

  One explanation is that in these cases, the cells with mutant p53 are indeed killed in the sledgehammer way oncologists originally imagined. They sustain severe damage to their DNA that cannot be repaired because p53 is out of action, nor can cell division be arrested. The cells carry on chaotically through the cycle and eventually succumb to what is called ‘mitotic catastrophe’ – wholesale failure of the machinery of replication. This scenario implies that it is essential for oncologists to know which way a tumour type will react to conventional therapy, depending on its p53 status. But things can get even more complicated.

  In some cases, giving chemo- or radiotherapy to patients whose cancers have normal p53 can actually make things worse. Cell death, as we know, is just one of several options chosen by p53 in response to damaged DNA. It can also choose to arrest the cell mid-cycle and send in the repair team before releasing the cell to carry on replication. Or it can condemn the cell to senescence – permanent arrest, which we know from the chapter on ageing can eventually stimulate cancer in neighbouring cells. Thus cancer cells that are not killed by chemo- or radiotherapy can be the seed stock for further tumours – and sometimes these new tumours are especially aggressive simply because the cells are survivors of highly toxic treatment, and bred for resistance.

  This makes sense, but it is only a hypothesis at present – there are no experimental data to prove it definitively. One source of confusion is the fact that, in the vast wealth of research that is carried out on p53, there is so little consistency in the methodology that it is hard to compare results. ‘In experimental systems we have all kinds of effects,’ said Pierre Hainaut. ‘You can always get an experimental system to behave as you would like it to, as an investigator! Now if you go to real life . . .’ Hainaut sat me down in front of the computer in his study at his Lyon home and brought up a paper he was about to submit. It was an analysis of a number of clinical trials involving the use of a common chemotherapy drug, Cisplatin, in lung-cancer cases. Overall, the effect of the drug was small, but he and his colleagues wanted to know whether the p53 status of an individual patient’s tumour influenced the outcome of Cisplatin treatment. For their analysis they had before them the biggest data set of its kind. It contained information about the outcome of treatment, plus the p53 status of the tumours, for 1,200 cancer patients from four trials, conducted in Canada, the US and Europe.

  The researchers found – unsurprisingly – that patients whose tumours had normal p53 did a lot better than those with mutant p53. But what did surprise them was that patients with some specific mutants – but, crucially, not others – got dramatically worse after Cisplatin treatment. Their tumours spread aggressively and many patients died even more quickly than they would likely have done with no treatment at all. Hainaut was not certain, at that point, whether it was the metastases that killed the people – he was awaiting further information from a statistician – but that was his hunch.

  Whatever the final cause of death turns out to be, knowing the p53 status of lung tumours will be useful in deciding who should receive Cisplatin therapy and who should not. ‘We are not doing well with lung cancer,’ Hainaut reflected as he scrolled through his paper on the computer screen. ‘There are probably 1.5 million people in the world with this type of cancer. Maybe 500,000 receive this treatment every year – and they receive it “blind”, because p53 is not being tested by mutation in these patients up front. Such a test would clearly improve the outcome. It would be really worthwhile . . . That’s the lesson of our analysis.’

  The situation Hainaut was describing was specific to lung cancer, with certain p53 mutations, treated with Cisplatin. But the lesson holds true more generally. What scientists have discovered about p53 and its role in conventional therapy offers cancer specialists a tool for making more rational decisions about how best to treat their patients. This is especially true when p53 status is part of a wider analysis of the genetic make-up of a tumour, because so many things besides this tumour suppressor have an impact on treatment. At present such tests are rarely offered in cancer clinics, but things are changing fast. As full genome sequencing becomes ever easier, quicker and cheaper to perform – and as the new gene therapies that target the defects specific to an individual patient’s tumour begin to reach the clinic – genetic analysis will become a routine part of diagnosis and treatment. Genetic analysis is an essential part, too, of the latest strategies for cancer prevention.

  THE BEST CHANCE OF SURVIVAL

  Compared to new treatment ideas, prevention studies have a tough time attracting cancer-research funds. The science of prevention is not as sexy; it doesn’t offer the same rewards to Big Pharma; and besides, it’s easier to get excited about tumours that are cured than about tumours that just don’t happen.

  Nevertheless, Bert Vogelstein is not deterred. ‘We believe the major impact on cancer over the next half-century will come not from treating advanced cancers, but from preventing cancer – in particular from detecting tumours at a very early stage,’ he said. ‘Virtually all cancers are treatable by surgery, without the need for any chemotherapy or radiation, if they’re caught early enough. That’s definitely true for colon, but it’s also true for many other tumours. It’s an underlying principle.’

  For a number of years now, Vogelstein’s lab at Johns Hopkins has been busily engaged in developing tools to look for evidence of early cancers. They are focusing their efforts on de
tecting biomarkers – bits of mutant DNA sloughed off by cancer cells that might be swilling around in a sea of normal DNA molecules in the blood, urine, stools or sputum, bearing witness to the presence of furtive disease. ‘The best marker, the best gene, is obviously p53, because it’s mutant in more tumours than any other gene – that’s the bedrock of this test,’ explained Vogelstein.

  The body fluid in which a biomarker is found is often a good indicator of where the tumour is developing: urine suggests bladder cancer, for example, stool suggests colon cancer and sputum suggests lung. By late 2012, Vogelstein’s team had investigated more than 700 cancers, starting with advanced tumours, to see if they could find free-floating biomarkers. ‘In advanced cancers of most tumour types – that is breast, colon, pancreas, lung – you can detect well over 90 per cent of them in the blood,’ he commented. For advanced colon cancer the researchers’ detection rate in stool samples is close to 100 per cent, and even in relatively early, pre-metastatic cancers it is 85–90 per cent. ‘This test is starting to rival colonoscopy in sensitivity,’ said Vogelstein. He reckons that even in blood samples, his team has more than a 50/50 chance of detecting colon cancer before it has spread. ‘And if you can detect even 50 per cent of cancers at a stage when they’re curable that would be massive.’

  Researchers working on the problem of liver cancer in West Africa have found biomarkers in the blood that can be used to screen for the disease before symptoms arise. In this region, you will recall, liver cancer is often associated with aflatoxin contamination of food crops, and DNA molecules released into the blood from a diseased liver show the characteristic fingerprint mutations in the p53 gene. Elsewhere, too, scientists are exploring the possibilities of using the presence of mutant p53 in body fluids to screen for early cancers.