FOR years, they鈥檝e been depicted as the good guys. Stem cells help renew many parts of our bodies and, we鈥檙e told, with a little coaxing might provide cures for anything from Alzheimer鈥檚 to diabetes. But stem cells can also turn bad. Very bad.
Rogue stem cells are now known to be involved in at least some cancers, including breast cancer, and they are resilient, relentless killers. They seem immune to most treatments, lurking in the background as the other cancer cells around them die. Then they churn out new cancer cells that can make old tumours regrow or seed new ones 鈥� perhaps the reason why so many cancers come back after seeming to have disappeared.
鈥淩ogue stem cells are known to be resilient, relentless killers鈥�
Advertisement
The discovery of cancer stem cells is bringing about a fundamental shift in our understanding of cancer, a shift that includes radically rethinking how we tackle the disease. Most cancer treatments are not designed to kill cancer stem cells, which may be why so many fail so miserably. So some researchers are trying to develop an arsenal of weapons designed specifically to eliminate cancer stem cells, in the hope of ridding patients of their cancer once and for all.
The field is buzzing with excitement, but it鈥檚 not going to be easy. Cancer stem cells are masters of disguise. They share many similarities with normal stem cells, a feature that makes them hard to spot 鈥� and even harder to kill. A lot of painstaking work is needed to establish just what role stem cells play in different cancers, let alone to develop therapies. Nevertheless, a handful of treatments are already nearing trials.
The idea that stem cells are involved in cancer goes back more than half a century. The trouble has been proving it. True stem cells can divide asymmetrically, producing one daughter cell that remains a stem cell and a second, called a progenitor, that is slightly more specialised. The progenitor divides more rapidly and ultimately gives rise to many highly specialised cell types that cannot divide any more. In this way, a relatively small number of blood stem cells, for example, can generate the billions of new blood cells needed by the body every day. With cancer, similarly, the idea was that a few rogue stem cells might churn out the huge number of abnormal cells that form the bulk of a tumour.
The property of 鈥渟temness鈥� is elusive, though. Definitive proof that cancer stem cells exist came only in 1994. John Dick鈥檚 team at the University of Toronto, Canada, took blood samples from patients with acute myeloid leukaemia and used a machine called a flow cytometer to separate them into two types based on their surface proteins. When the cells were injected into mice with deficient immune systems, only one type caused the animals to develop the cancer.
These 鈥渃ancer stem cells鈥� were incredibly rare. Only one in several hundred thousand cells from the initial blood sample could give rise to a cancer. 鈥淭hese are the real business end of the tumour,鈥� says Dick. The study suggested that although the bulk of cancer cells are what do the damage, the evil masterminds are the tiny subset of stem cells that keep the cancer growing. The implication was huge: to cure a person of cancer, these cancer stem cells must be eliminated, something that most treatments don鈥檛 do.
The reaction was guarded. Many people thought that leukaemia, a cancer of the blood, might be the exception rather than the rule and that cancer stem cells were unlikely to turn up in solid tumours. 鈥淏ut it was the beginning of a paradigm shift,鈥� says Stephen Emerson, who studies blood stem cells at the University of Pennsylvania. This shift was completed in 2003 when cancer stem cells were found in two different solid tumours 鈥� breast and brain cancer.
鈥淭his opened the floodgates,鈥� says Sean Morrison of the University of Michigan in Ann Arbor, one of the leaders of the breast cancer team. 鈥淧eople suddenly thought: 鈥楬oly cow, this model could change the way we think cancer works and change the way it is treated.'鈥�
Previously, most experts had thought that just about any cancer cell could divide like crazy and keep on dividing indefinitely. This means that any cancer cell left in the body could potentially reignite the disease, so the aim of most treatments is to kill all these dividing cells: standard therapies such as chemotherapy and radiation target rapidly dividing cells.
The cancer stem cell theory, however, says that only cancer stem cells can reignite the disease. These cells divide slowly 鈥� unlike the progenitors they give rise to 鈥� so they might survive most therapies. In fact, they may be particularly hard to kill. Last month, Jeremy Rich鈥檚 team at Duke University Medical Center in Durham, North Carolina, showed that cancer stem cells from glioblastoma brain tumours are very good at repairing their own DNA. This helps them survive the DNA damage caused by radiation therapy.
鈥淟ots of people say that a failure to recognise cancer stem cells earlier could explain our inability to treat cancer effectively now,鈥� Morrison says. It鈥檚 certainly a scenario that matches clinical experience. There are numerous cases of chemotherapy or radiation shrinking tumours so much that they become invisible on X-rays, only for the cancer to return later and kill the patient. If the cancer stem cell model is right, these relapses occur when quiescent cancer stem cells regenerate the tumour.
The question then is, how many tumours are like this? Cells with stem-cell-like characteristics have now been found in many other solid tumours, including ovarian, lung and skin cancers, but more work is needed to prove that these really are cancer stem cells. 鈥淭he field is plagued by a lack of understanding of stem cell biology,鈥� says Peter Dirks at The Hospital for Sick Children, Toronto, who was the first to isolate cancer stem cells from human brain tumours. Just because a cell grows in a particular way in culture, or has a particular set of markers, doesn鈥檛 necessarily make it a stem cell, he says.
鈥淭his changes how we think cancer works and how we treat it鈥�
For the moment, the experts are divided. Some, such as Morrison, think that cancer stem cells are involved only in some types of cancers. Others argue that a cancer stem cell of some kind will be found in all cancers. 鈥淚f you鈥檇 asked me even six months ago, I would have been cautious,鈥� says another researcher in the field, Craig Jordan of the University of Rochester, New York. 鈥淏ut I鈥檝e seen a lot of really amazing data presented recently, so now I鈥檓 not so sure.鈥� Two Nature papers out this week, for instance, confirm that rogue stem cells give rise to colon cancers.
Rogue cells
It could be decades before this debate is finally settled. In the meantime, those studying cancers known to involve stem cells are trying to figure out how the rogue cells arise. The parent cell could be a normal stem cell that turns bad when it picks up mutations in the genes that control growth and differentiation, or it could also be a more mature cell that reverts to a more stem-cell-like form.
The scenario is likely to be different for different tumours, but some basic principles will probably hold. For instance, although more than one mutation is needed to send a cell permanently off the rails, the final hit need not always be genetic: with stem cells, their environment or 鈥渘iche鈥� might play a particularly important role. Stem cells constantly 鈥渢alk鈥� with surrounding cells, which normally tell them to remain as they are. If the signals change 鈥� in response to injury, for example 鈥� stem cells may start to pump out more specialised cells. In the case of cancer stem cells, these will be cancerous cells.
When it comes to developing treatments, the problem is that cancer stem cells and normal stem cells are virtually identical, responding to similar signals in similar ways. The big challenge is to find drugs that destroy one without harming the other.
Much of the work to date has focused on leukaemia, as it is one of the easiest cancers to study. Most types are caused by cancerous blood stem cells in the bone marrow pumping out abnormal white blood cells. These pass into the blood and circulate around the body, making it hard for the body to fight infection and causing swelling in the lymph nodes and spleen. Common treatments include chemotherapy and radiation, but these are not designed to kill quiescent cancer stem cells, so relapse is all too common.
The good news is that researchers have spotted a couple of genetic differences between normal and cancerous blood stem cells that are already leading to therapies aimed specifically at cancer stem cells. Jordan has discovered that a cell signalling system called the NF-kappaB pathway is permanently switched on in the cancer stem cells of people with acute myeloid leukaemia (AML) but not in normal blood stem cells. 鈥淚t was an exciting observation,鈥� Jordan says, 鈥渂ecause it was the first clear difference between leukaemia stem cells and normal stem cells that might have therapeutic relevance.鈥�
What鈥檚 more, his team has found that parthenolide 鈥� the main active ingredient in feverfew, a medicinal plant used for treating aches and pains 鈥� blocks a key part of this system. Treating AML cells with parthenolide caused cancer stem cells to commit suicide but left the normal stem cells intact (Blood, vol 105, p 4163). Jordan hopes that trials of a parthenolide-like molecule will start sometime in 2007.
The second difference involves a gene called Pten that usually prevents tumours arising. Mutations in this gene help turn blood stem cells cancerous: if the gene is deleted in mice, the number of normal blood stem cells falls while the number of leukaemic stem cells rises. Most of the animals develop leukaemia and die within six weeks, says Morrison, who led the work (Nature, vol 441, p 475).
Here too there is hope, however. Treating the animals with a drug called rapamycin reverses the effect of deleting the Pten gene, slowing the proliferation of cancer stem cells and reducing their numbers. 鈥淕iven early it鈥檚 fabulously effective, and mice never get sick,鈥� says Morrison.
Rapamycin, originally approved for use as an immunosuppressant, is already being tested in cancer trials because its target, Pten, is frequently missing in human tumours. Morrison thinks the therapy will work best when given early and in combination with other treatments such as Glivec (Gleevec in the US). He is helping to plan a trial for chronic myeloid leukaemia along these lines.
Contrary to expectation, however, it might not be necessary to kill cancer stem cells in order to cure a cancer. Forcing them to differentiate normally might work just as well.
Blood stem cells, for instance, normally give rise to more specialised cells called myeloid stem cells. These go on to form all the fully differentiated cells of the blood system, including platelets and red and white blood cells. In patients with leukaemia, this pathway becomes blocked: rogue myeloid stem cells instead produce large numbers of abnormal, undifferentiated cells called myeloid blasts.
Coaxing these rogue myeloid stem cells to differentiate into normal cells, however, would cut off the supply of myeloid blasts. It鈥檚 an idea that has already yielded some success. The standard treatment for acute promyeloid leukaemia (APL), a rare subtype of AML, is now retinoic acid alongside chemotherapy. The retinoic acid appears to force the myeloid stem cells to differentiate into normal cells, while chemotherapy kills the rest of the cancerous cells. The result: around 85 per cent of patients go into remission.
Glivec works in a similar fashion. The treatment, which is routinely used to treat chronic myeloid leukaemia, was designed to inhibit an abnormal enzyme that is found in most CML patients. 鈥淧eople call it a targeted therapy,鈥� says cancer stem cell researcher Stewart Sell of the Wadsworth Center and Ordway Research Institute in New York, 鈥渂ut it鈥檚 really targeted differentiation.鈥� By blocking the enzyme, Glivec forces cancerous blood cells to differentiate and die within a few days, just as their normal, non-cancerous counterparts would do.
Like other existing differentiation therapies, however, Glivec targets progenitor cells rather than stem cells. This means it doesn鈥檛 eliminate the cancer stem cells, so patients have to keep taking it to prevent the cancer from recurring, which allows resistance to evolve. If you could instead make the cancer stem cells differentiate into normal, specialised cells instead of self-renewing, then you could effectively eliminate them 鈥� and the disease 鈥� altogether.
This is what Dick is trying to do. His team have shown that if mice with leukaemia caused by human AML cells are injected with an antibody that binds to a key protein on the leukaemia stem cells鈥� surface, the animals鈥� cancer disappears (Nature Medicine, vol 12, p 1167). Critically, when bone marrow taken from these treated animals is injected into other mice, the recipient animals do not develop leukaemia. Bone marrow from untreated animals with AML cells, however, does trigger leukaemia in other mice.
Dick thinks that this antibody, which binds to the surface protein CD44, robs cancer stem cells of their ability to self-renew, perhaps by interfering with the relationship between the cells and their niche. Healthy blood stem cells spend most of their time in the bone marrow, where signals from surrounding niche cells help keep them in their stem cell-like state. Once in a while, the stem cells enter the bloodstream for a time before returning to their bone-marrow base. The exact reason for this is unclear, but it鈥檚 a wanderlust shared by some cancer stem cells. 鈥淟eukaemia stem cells need to be able to home back to their supportive microenvironment,鈥� says Dick. 鈥淭he antibody treatment makes them less able to find their way from the bloodstream to the bone marrow.鈥�
It鈥檚 not clear if disrupting the homing process makes leukaemia stem cells lose their ability to self-renew or vice versa, but CD44 is clearly a key player in the process, as it helps stem cells maintain their ability to self-renew and, in essence, remain as stem cells. 鈥淚f we can identify the factors responsible for keeping these cells in their stem-cell-like state, we can target these pathways,鈥� says Sell.
As promising as all this research is, it鈥檚 unlikely to lead to a quick fix or a one-drug-treats-all scenario. Different cancers have different underlying mechanisms, Sell says. 鈥淭he problem is the concept of a cure for cancer. There is no cure for cancer. There are many cures for cancer.鈥�
