NAME: Marcus Peter
POSITION: Professor, cancer research, University of Chicago
Associate professor, cancer research, University of Chicago — 1999-2001
Lecturer, University of Heidelberg — 1997-1999
Staff scientist, tumor immunology, German Cancer Research Center — 1992-1999
Postdoc, Dana-Farber Cancer Institute — 1990-1991
PhD, biochemistry, University of Bayreuth — 1988
MS, biochemistry, University of Bayreuth — 1986
This month, a research team led by the University of Chicago’s Marcus Peter reported in the April 1 issue of Genes & Development that the miR-200 microRNA family plays a major role in determining the epithelial phenotype of cancer cells.
According to the findings, miR-200 directly targets the mRNA of two transcriptional regulators of E-cadherin, a transmembrane protein associated with cancer progression.
“Ectopic expression of miR-200 caused up-regulation of E-cadherin in cancer cell lines and reduced their motility,” according to the paper’s abstract. “Conversely, inhibition of miR-200 reduced E-cadherin expression … and induced” a natural differentiation process similar to the one cancer cells undergo as they metastasize.
This week, RNAi News spoke with Peter about the findings and their implications.
Let’s start with a little about your lab and the research you do there.
I’ve worked for 16 years now in the apoptosis field with death receptors, with an emphasis on the Fas death receptor. Ten years ago, we published a paper [in The EMBO Journal] showing that the Fas receptor can kill cells in two different ways [depending on the cell type]. At the time, we called [the two kinds of] cells type I and type II cells, and I initially thought [the difference between them] was just a difference between the ways the receptor signals in different cells.
Over the years I realized this goes far beyond that; Fas signaling is one of many different properties that change in the cells. I came to believe that we’re looking at two different differentiation stages of cancer, basically.
MicroRNAs … are known to be fundamentally important to both regulating and maintaining differentiation processes. [Last year, we published in the Proceedings of the National Academy of Sciences the results of a miRNA expression analysis] using 10 of the type I cells and 10 of the type II cells [and found that expression of the miRNA let-7 was a marker for two differentiation stages of cancer]. That’s when we realized that what we’ve been looking at are, indeed, different differentiation stages that seem to be regulated by microRNAs.
That sparked my interest in microRNAs and their role in tumor progression.
The Genes & Development paper focuses on miR-200. Had you previously identified this as a microRNA family of interest?
The cells that were different in their expression of let-7 … had been subjected to gene chip analysis in 2000. [The researchers conducting the work] looked at the 60 tumor cell lines of the National Cancer Institute, did an unsupervised, non-hierarchical cluster analysis and found that all the NCI-60 cells fell into two classes. One class had epithelial genes expressed, the other class had preferentially [expressed] what they called stromal genes – basically mesenchymal genes.
That’s when I realized that what we were looking at in the difference in Fas signaling is probably a difference between epithelial cells, or cells that still have some epithelial properties, and mesenchymal types of cancer cells. When we did a screen and identified for let-7, I thought it would be a regulator of epithelial- mesenchymal transition. But it turned out it wasn’t.
I [then] asked myself whether there was another microRNA family that is specifically in charge of this particular function in differentiation. That’s when we interrogated a dataset that came from [our PNAS] study on the expression levels of microRNAs on the NCI60 cells, provided to us by Dr. Mark Israel [at Dartmouth College, using the] classical epithelial and classical mesenchymal markers, E-cadherin and vimentin.
I believed initially [that there might be] a single microRNA … that would be preferentially expressed in either the epithelial (E-cadherin-positive) or mesenchymal (vimentin-positive) cell lines. It was as simple as that.
When I did this analysis one night at home on my computer, I literally fell out of bed … because I had stumbled over a correlation that was not just a correlation — it was more like a rule. MiR-200 was a family that towered above everything else [in that] there was not a single cell line among these established cancer cell lines … that had E-cadherin expression with no vimentin but did not [express] miR-200. On the other hand, there was not a single cell line that had vimentin, but no E-cadherin, and lacked miR-200.
When I plotted the mRNA of … [the E-cadherin transcriptional repressors] ZEB1 and ZEB2 toward any of the miR-200 family members, every single dot in the plot for every single one of the NCI-60 cell lines was on the axes of that graph. It was clear: there is not a single cell line that would even tolerate moderate amounts of miR-200 and express ZEB2 at the same time.
We tried for many years in studying Fas signaling … to induce epithelial-to-mesenchymal transition or the reverse in any of the established cancer cell lines but have never accomplished that with any of the established EMT inducers. MiR-200 did that right away. We introduced miR-200 into … mesenchymal cells [with low expression of the miRNA] and bingo, they underwent MET. We inhibited miR-200 in established epithelial cell lines that had high [expression of the miRNA] and right away the cell underwent beautiful EMT with many of the morphological and biochemical changes.
This is why miR-200 is one of those master regulators for which we have been looking for many years.
Can you explain what epithelial-to-mesenchymal transition is and how it relates to tumor progression?
Epithelial-to-mesenchymal transition is a differentiation process that occurs as a part of embryonic development — that’s where it was first described. During embryonic development, lots of tissues need to migrate around the body to form new organs and limbs, et cetera. Sometimes they start out with an already differentiated tissue and get the instruction to change morphology and travel to distant sites and form new tissues.
Sometimes they have to transition from epithelial tissue to acquire a more mesenchymal phenotype, meaning they lose polarity, they change actin cytoskeleton, they become mobile and start to travel around. It’s quite a fundamental process.
That’s EMT during development. Over that last 20 years or so, it has become clear that a number of other processes that we know under very different names, for instance wound healing and carcinogenesis, actually are genetically quite similar.
[In relation to cancer progression], most carcinomas are derived from … the epithelial layer of a cell that becomes dysplastic, undergoes a chain of mutations, and then [reaches] the moment … that really poses a problem to cancer treatment: the cells change their morphology, undergo an EMT-like process, become invasive and start moving around, and form metastases.
This initial process of metastasis formation is very similar to EMT during embryonic development, and for a reason; these cancer cells re-acquire embryonic properties, turning on a program that is intrinsic to every cell from being an embryonic cell.
MiR-200 is [possibly] the microRNA that regulates this process during embryonic development … and during carcinogenesis, the reverse occurs … and tumor cells for some reason lose [miR-200]. That’s probably the moment they become more invasive and start to metastasize.
What’s the next step to follow up?
Our data and the work from many others suggest that the reason that cancer cells become resistant to drug treatment is that they undergo changes [such as EMT]. One of the new models is that they lose microRNAs. Mir-200 is not the only microRNA [implicated here] — let-7 is another one.
The way we place [the miRNA] in this process of tumor progression is that let-7 is lost first and miR-200 is lost later. So let-7 will cause the cancer cells to acquire a more embryonic phenotype and miR-200 loss will then induce this EMT-like process that causes them to become invasive.
Obviously, what needs to be figured out is how these microRNAs work together and the pathways that regulate the microRNAs in order to be able to devise new therapies. The simplest would be, of course, to re-introduce let-7 and miR-200 — and I mention both because I think one probably would have to combine the two.
You can image there will be … lots of interest now in generating new reagents that will either mimic or represent microRNAs, and introduce them into the cancer cells to reverse this differentiation process that makes them invasive and resistant to therapy.