Lorne special: Piecing together the breast cancer puzzle

This feature appeared in the January/February 2010 issue of Australian Life Scientist. To subscribe to the magazine, go here.

Why do some cells in mature tissue or a tumour suddenly break away and initiate a spate of uncontrolled growth and spread into other tissues in a way that harms the whole organism? What are the hallmarks of such a cell that set it apart as a troublemaker? Can we use those marks to identify them, and even more importantly, to stop them before things get out of hand?

These questions underpin many efforts in cancer research worldwide. One such endeavour at the Walter and Eliza Hall Institute (WEHI) for Medical Research in Melbourne has contributed several pieces already to the giant puzzle that is cancer. Recently it found one more piece, although it was a piece that no-one expected to fit where it did, not even the researchers involved.

Jane Visvader’s team is leading this effort, the main focus of which is investigating those cells in breast tissue that are predisposed to becoming tumorigenic and those cells that are responsible for sustaining breast cancer. “To do that, we need to first have a very good grasp of the basic biology underlying normal breast development,” Visvader says.

Breast cancer is a very heterogeneous disease in terms of both what the cells look like and the genetic markers they express, with six distinct subtypes already classified based on gene expression profiling. Thus, isolating and classifying functionally distinct epithelial cell populations in the mammary gland will help to work out the steps and critical factors that govern the differentiation of stem cells and progression of intermediate cells to form the mature tissue hierarchy. Only then can the field really start to reliably define potential cells of origin in breast cancer based on established relationships between the normal cell subsets and tumor types, as well as the key cell changes that could underlie the observed breast cancer heterogeneity.

As part of this global push, WEHI’s Victorian Breast Cancer Research Consortium Laboratory, which Visvader co-heads with scientist and clinical oncologist, Associate Professor Geoff Lindeman, has spent many years systematically identifying and finding out as much as they can about the normal epithelial cells that reside in breast tissue from human and mouse. These sorts of basic studies are critical for the team’s long-term goal of identifying potential targets for breast cancer that may be useful as diagnostic or prognostic markers and for developing new therapeutic strategies.

In 2006, Visvader and Lindeman’s team reported the first identification of a mouse mammary stem cell. “Since then, we have made some progress identifying daughter progenitor cells that lie between the stem cells and the mature cells that make up mammary tissue,” says Visvader. At the same time, the group was also focussing on compiling the same sort of biological data in humans – this culminated in their recent publication in Nature Medicine of the mammary stem cell in normal human tissue.

---PB---

The corner piece

The first thing that struck Visvader during these pivotal studies was the quite analogous hierarchy between human and mouse breast tissue. However, it was another finding from this work that turned out to be particularly significant and exciting, and which provided that unexpected piece of the cancer puzzle.

As part of the tissue hierarchy characterisation, Visvader and Lindeman’s team identified a mammary luminal progenitor cell in the human tissue. However, it was the experiments following this that provided a few surprises, Visvader recalls.

“Firstly, this luminal progenitor cell showed a very similar gene signature to the basal subtype of breast cancer. This was a very unexpected finding which suggested to us that this luminal cell might be a target cell from which basal-type breast tumours originate.”

Next, the researchers looked at some samples of pre-cancerous and normal breast tissue from women carrying the BRCA1 mutation in the context of their newly defined tissue hierarchy including the luminal progenitor cells. And the surprises kept coming.

“BRCA1 gene mutations are carried by 10 to 20 per cent of women with hereditary breast cancer,” says Visvader. “These women often develop the basal-like subtype of breast cancer, which is very heterogeneous and among the most clinically aggressive. What we were trying to do was understand the relationship among normal epithelial cells, cells predisposed to becoming cancerous and those that are propagating the breast cancers.

“What we found was that the precancerous specimens donated from BRCA1 mutation carriers contained a much larger population of luminal progenitor cells than any of the six known molecular subtypes of breast cancer used in our analysis. Furthermore, luminal progenitor cells isolated from the BRCA1 samples displayed aberrant growth properties.

“This was all quite unexpected because everybody in the field, including ourselves, had expected the culprit in this BRCA1 mutation-positive group to be the mammary stem cell (although nobody had actually identified a potential culprit cell, stem or otherwise). However, it was not the stem cell, but instead a progenitor cell downstream of the stem cell.”

Then, to make the story even bigger, subsequent gene expression profiling revealed that BRCA1 breast tissue and basal breast tumors are more similar to normal luminal progenitor cells than any other cell type in the breast.

“Together, these results strongly suggested that the luminal progenitor cell could be the target population that eventually gets transformed and gives rise to basal breast cancers in these BRCA1-mutation carriers.”

---PB---

Why is this surprising? “It’s that basal breast cancers tend to express basal markers and it is likely that the stem cell is located in a basal position. This suggests that these cells might give rise to the basal types of breast cancer, which includes the BRCA1-associated cancers,” Visvader explains.

“So, we were a little surprised to see that the stem cell pool was diminished and the luminal progenitor cell pool was expanded in these mutation carriers. We thought ‘what is going on here?’ Subsequently finding that the luminal cell pool was also behaving aberrantly whereas the stem cell pool was not was even more interesting.”

The researchers then took the data into mouse models for confirmation. “Even there, preneoplastic tissue from BRCA1 model tumours had fewer functional stem cells. So, there was certainly no evidence of an enhanced stem cell pool in any of these model systems from human or mouse. In fact, everything was pointing towards the luminal progenitor cell being the culprit.”

One of the final, and perhaps the nicest, findings for Visvader came when the bioinformaticians at WEHI analysed the gene profiles of the different subpopulations, including the luminal cells and the known subtype populations of breast cancer. What they found was a really striking correlation between the luminal progenitor cell gene signature and the basal cancer subtype. For Visvader, this was the clincher.

Notch in the puzzle

At the Lorne meeting this year, Visvader will recap this exciting work and also discuss some new data about potential tumour-initiating cells in humans and mice, including the luminal progenitor cell. She’ll also talk about the potentially key role played in the tumorigenic process by Notch signalling.

“The Notch pathway is, of course, very important in normal development and also in many different types of cancers including breast cancer – where Notch1 and some of the Notch signalling ligands are often activated. We recently found that these luminal progenitor cells also seem to be a target for inappropriate Notch signalling. So, when you have the pathway constitutively switched on (it is normally very tightly regulated), the luminal progenitor cell is targeted and seems to be the cell that eventually gives rise to the mammary tumours – so it acts like a target cell pool for future mutations down the line.”

---PB---

Visvader’s group is now trying to further fractionate the mammary stem cell population in human tissue, which appears to be quite heterogenous. “So far, we have defined gene signatures for all of our different epithelial populations, and in doing so, have identified several potential targets. What we would now like to do with that data – and are making some progress on – is to test the relevance of these potential targets in mouse models.”

This involves growing breast cancers in mice based on these defined cell populations from human tissue. The mice are then treated to try and inhibit the candidate gene targets and to see what happens to the cancer. For example, is growth or metastasis inhibited, or is survival increased? “This part of our work is really about embarking on pre-clinical validation studies with our shopping list of targets and, of course, this will be a long road,” Visvader adds.

Visvader and Lindeman have established strong clinical links through the Victorian Cancer Biobank (established by Lindeman) and other similar national resources, as well as through WEHI’s historically strong ties with nearby oncology and clinical trials units. This situation puts Visvader and colleagues into a stronger position than many for doing pre-clinical studies.

“If we identified a marker on our luminal progenitor cell, for instance, or developed an antibody that could inhibit the growth of these cells in mouse models, then we are in a really good position to taking it this to Phase I trials that could be carried out locally.”

This feature appeared in the January/February 2010 issue of Australian Life Scientist. To subscribe to the magazine, go here.