Tracing the roots of breast cancer

A breast cancer cell – but where did it come from?

Our bodies are made of hundreds of different types of cells. And when processes inside them go wrong, and allow them to keep dividing uncontrollably, cancers form.

But individual cells are very small – by the time tumours (which are made of millions of cells) are large enough to be detected, the cells that make them up have evolved and changed along the way.

This makes it difficult to trace a particular tumour’s origins – especially given that we’re now discovering that there are many different types (and subtypes) of cancer. Researchers have long wondered what types of cell different tumours originate from, and what triggers them to ‘go rogue’ in the first place.

A fascinating study from our researchers in Cambridge, published late last year, has begun to help answer these questions – at least for breast cancers.

The team, led by Dr John Stingl, traced the ancestry of different types of cell within normal, healthy breast tissue. And they found clues as to how these diverse cell types might be related to some of the different types of breast cancer.

We’re going to look in detail at what they did – and why it’s an important step forward in understanding the disease. But before leaping into the murky world of biological ‘family trees’, it’s worth a quick pause to look at what breasts are made of.

Back to basics

Most of a breast is made of fat, but it also contains a complex network of lobes and ducts, to make milk and transport it to the nipple. And nearly all breast cancers begin in the cells that line these structures.

The lobes and ducts are made of two layers of cells – called epithelial cells – which are a little like an undercoat and top layer of paint in your living room. The cells in the undercoat layer, or ‘basal’ layer to give it its proper name, are called myoepithelial cells.

And the surface layer of the milk lobes and ducts are made up of cells called luminal epithelial cells.

More pieces to the jigsaw puzzle

Last year, our landmark METABRIC study revealed at least 10 distinct types of breast cancer, each characterised by a unique genetic fingerprint. But why are there so many different types? If the milk ducts are simply made from two types of cell, it’s hard to picture where this variety could spring from.

But – as is usually the case in cancer – the reality is a bit more complex, and over the years, it’s emerged that milk ducts aren’t just made of two types of cell – there are also some subtly different cells with more specialised roles.

This is because epithelial cells don’t last forever: they have a limited shelf-life and eventually they become old and die, meaning that new cells must constantly be made to replace them.

And, as any woman knows, over her life her breasts change quite significantly – first during puberty, then during pregnancy, and after childbirth to produce milk.

The job of making and maintaining all these new cells falls to rare, specialised cells in the breast, known as ‘stem’ and ‘progenitor’ cells. And these are the focus of Dr Stingl’s research.

If we imagine a ‘family tree’ of breast cells, the breast stem cells are right at the top (see diagram below). They’re capable of dividing into two daughter cells, one of which is another stem cell, but the other is more specialised, and will go on to become other types of cell on lower ‘levels’ of the family tree – including the cells of ducts and lobes.

In the family tree’s next layer are the progenitor cells. These are still relatively unspecialised and can divide like stem cells, but are further towards becoming a final luminal, milk-producing, or myoepithelial cell.

Once a cell reaches its final destination (the bottom layer of the family tree) it can’t divide any more or become any other type of cell, its fate is set in stone (it is said to be ‘fully differentiated’).

Stem and progenitor cells have long been on the radar of scientists studying cancer. As these cells can carry on dividing forever, and cancer is a disease caused by uncontrolled cell division, they seem an obvious place to start looking for clues.

And because they live a long time, they can end up accumulating more mistakes in their DNA than shorter lived cells, making them at more risk of becoming cancerous.

Discoveries provide clues but raise more questions

A couple of years ago, researchers led by Dr Matt Smalley made a surprising discovery. His lab, now based in Cardiff, was studying breast cancers that develop due to mistakes in the BRCA1 gene.

These cancers are often described as ‘basal-like’, because they have similar characteristics to myoepithelial cells in the basal layer of the milk ducts. And researchers naturally assumed that basal-like breast cancer originated from the basal myoepithelial cells.

Smalley’s team showed that this wasn’t the case – their experiments demonstrated clearly that these cancers came from faulty luminal progenitor cells with low levels of oestrogen receptor on their surface.

But not all the evidence tied in with this. Other researchers had spotted different luminal progenitors that had high levels of oestrogen receptors on their surface – suggesting, conversely, they were more closely related to so-called oestrogen-receptor positive breast cancer (ER+ cancer).

To try to resolve the paradox, Dr Stingl’s team set about studying the progenitor cells in the luminal layer in minute detail, using a more precise technique to isolate them than anyone had previously.

Plucking out the progenitor cells

Scientists wanting to study breast progenitor cells have to isolate them from the rest of the cells in the breast. To do this, they use a technique called ‘fluorescence-activated cell sorting’.

This method uses antibodies – small proteins made naturally by our bodies – to stick to pre-chosen proteins on the surface of human cells. The antibodies have little molecular ‘flags’ on them.

The scientists break down breast tissue into its constituent cells, and mix it with their antibodies so they stick to the cells of interest. Finally, they pass the resulting mixture through a machine which pulls out each flag-bearing antibody and attached cell.

The tricky part is using the exact mixture of antibodies in the right sequence of steps; otherwise you end up contaminating the cells you’re interested in with other cells you don’t want to study. Using a new combination of antibodies, Stingl’s team were able to isolate the different types of cells from both human and mouse breast.

Once isolated, they grew different cells in a laboratory to see whether they could divide and become milk duct cells (i.e. are true progenitor cells), and analysed them to see what proteins they made, and what genes are turned on and off inside them. And these patterns were compared to the different types of breast cancers to look for similarities.

Two types of progenitor

Stingl’s painstaking experiments proved that actually there was no paradox, and all the previous findings were right. There are, in fact, at least two different types of luminal progenitors in mice, some with, and some without, the oestrogen receptor .

Similarly, when the team looked at breast tissue from humans, they also found two comparable groups of luminal progenitors, but with subtly different proteins on their surface. However, the characteristics of these two groups of progenitor cells were similar in both mice and humans, and predicted whether the cells went on to become new luminal epithelial cells or milk-producing cells.

One key finding was that a small group of progenitor cells (those with high levels of the oestrogen receptor that become new luminal cells) were better at surviving in conditions of low oestrogen. The scientists think that this small group of cells may be the ones that become ER+ breast cancers in postmenopausal women (who have low oestrogen levels).

The other main type of luminal progenitor cell was, as predicted by Smalley’s team, more genetically similar to aggressive, basal-like breast cancers.

In a final, perplexing, but completely new observation, they also noticed a third population of ‘progenitor-like’ cells in human breasts, which weren’t able to grow and had very low levels of a protein called Her2 on their surfaces. These seemed to be present in different levels in different tissue samples. What they do, and how they relate to cancer, is still a complete mystery.

So how does this bring us closer to curing breast cancer?

Discovering more about the cellular make-up of breast tissue – including the different types of progenitor cells – will help explain why there are so many different types of breast cancer.  Revealing more about the origins and triggers for the different types may lead researchers to find better treatment strategies.

And if, as many suspect, the root cause of breast cancer is failure to control cell division in progenitor cells, then understanding more about the fundamental workings of these cells, like what processes drive them to divide, could provide new ways to treat cancers in all its forms.

The research may also provide clues to help doctors to predict the behaviour of individual women’s cancers.

This is important. A fundamental question scientists are trying to answer is how to tell the difference between aggressive and slower-growing tumours picked up during breast screening, a topic we’ve covered in some depth. Being able to tell the difference could mean that women with less aggressive tumours could be monitored and potentially avoid over-treatment.

Tracing breast cancer back through its family tree to pinpoint the exact place and time it pops into existence is still a work in progress. But fundamental research like this helps us to understand more about the cells where the disease stems from, why breast cancer is such a diverse disease, and how we can move closer to curing it.

Emma

• More about different types of breast cancer

Reference

• Shehata M., Teschendorff A., Sharp G., Novcic N., Russell I.A., Avril S., Prater M., Eirew P., Caldas C. & Watson C.J. & (2012). Phenotypic and functional characterisation of the luminal cell hierarchy of the mammary gland, Breast Cancer Research, 14 (5) R134. DOI: 10.1186/bcr3334