3

That's not supposed to happen, or is it?

When I began my post-doc I was full of excitement and anticipation. Okay, possibly also a generous fixing of fear with a side of self-doubt. I remember being especially careful in everything that I did, from simple tasks like making new buffers to spending hours learning how to culture previously unfamiliar breast cancer cell lines. I remember the urgency of finding and starting a project. I made lists and short-lists (and shorter lists) of project ideas. I talked to everyone I could about their projects, seeking advice and inspiration for my own. Despite all that planning, it turned out that it was instead a very simple observation that took me in a new direction, and ended up forming the basis of my current work and plans for my future lab…

So let’s talk about some science. Breast cancer cell lines are adherent cell populations (ie, they stick to and grow on the surface of cell culture plates) and, like most adherent cell lines, they do not survive in suspension. Or at least they weren’t supposed to. When I first began culturing them I noticed that there were floating cells in the media. Normally, floating cells should be dead or dying, but these ones looked alive and healthy. Mildly curious, I decided to collect the cells and re-plate them in new dishes. I was surprised a couple of days later to see the floating cells had adhered to the tissue culture plates and were proliferating happily. I investigated further. As it happened, the floating cells turned out to be enriched for two specialized cell types. I first found that they are markedly enriched for mammary stem-like cells. This method of isolating floating cells became a simple way to enrich for putative stem cell populations.

The more important finding lay in the nature of the second cell type, the non-stem cells enriched in the floating population. Using the best available cell-surface markers to identify mammary stem cells, I purified the stem-like fraction and the remaining non-stem fraction. My initial idea was to demonstrate that the stem-cell-enriched fraction could recreate the non-stem fraction (which is a required feature of naming any cell a ‘stem cell’). Instead, I observed over time that stem cells arose in the purified non-stem fraction. This suggested that non-stem cells could recreate the stem-cell fraction.

Fair enough, you say? The hiccup was that the then widely accepted stem cell model featured stem cells residing at the top of a hierarchy and undergoing unidirectional differentiation into non-stem progeny. That is, stem cells could become other cells, but other cells could not reverse the process to become stem cells. My data disagreed. Rather, it suggested that there are bi-directional interconversions between stem cells and non-stem cells. I was further able to demonstrate that oncogenic non-stem cells could create cancer stem cells (CSCs), which are highly aggressive tumor cells thought to be responsible for tumor initiation and metastasis.

And so, after many repeats of these and other experiments, I presented my findings to my colleagues, first at our weekly lab meetings and then more broadly at conferences. Let’s just say that not everyone was immediately convinced that the paradigm warranted shifting. I did have encouragement and support from my amazing lab head, Bob, who is not unfamiliar with the efforts required to budge a paradigm. After up-mountain battles with several reviewers, editors and journals, I went on to publish the work, with the take-home message being that non-stem cancer cells can create and continually replenish the pool of CSCs within a tumor.

That work then led me to question how common non-CSC to CSC conversions were in the cancer setting, forming the basis of my current work (and most recent paper).

To address this, I analyzed a series of breast cancer cell lines representing two broad groups of clinical breast cancer, luminal breast cancer - which generally has a good clinical prognosis, and basal breast cancer – which is generally more aggressive and has a poor clinical prognosis.

To cut a longish story shortish, I first noted that basal non-CSCs, but not luminal non-CSCs, could generate CSCs. I surmised that the ability to generate CSCs was an important indicator of a cancer’s potential to progress to a more aggressive state. In search of a mechanism, my team and I analysed a series of factors, followed a hunch around the epithelial-to-mesenchymal transition (EMT), and identified induction of the EMT transcription factor ZEB1 to be critical for basal non-CSC-to-CSC plasticity. Unsure as to why basal but not luminal breast cancer cells could turn ZEB1 on, we looked at the state of the chromatin surrounding the ZEB1 gene. These experiments led to an important finding: basal non-CSCs maintain the ZEB1 promoter in a state poised for activation, whereas luminal non-CSCs maintain ZEB1 in a repressed state.

Our work offers an explanation for why basal breast cancer is generally more aggressive and has a poorer clinical prognosis than luminal breast cancer. It also suggests that therapies aimed at inhibiting this type of cell plasticity may help to reduce tumor metastasis and recurrence.

So, lots has happened to this Australian girl since my first wintry day in Boston. Then-boyfriend is now-husband. Then-twinkles-in-our-eyes are now two beautiful daughters, what was a new city now feels like a home, and I’ve found that with patience and robust data, paradigms can sometimes be budged.

I see now that even for the most accomplished and confident post-docs, the first six months in a new lab are challenging. I still remain cautious, but now of the science rather than of myself. I am grateful to have had the opportunity to work with and learn from Bob, as well as some wonderful colleagues that make me laugh every day, and for the unwavering support of my family.

And to my daughters I say follow your curiosity and believe in yourself, and that the simplest observations can lead to important discoveries.