Lorne Cancer: Architecture and oncology

The way tumour cells talk with their environments is now known to play a major role in cancer, and is a burgeoning area in the field of cancer research.

How do cancer cells develop, maintain their viability, change their motility, dedifferentiate, and exist within the three-dimensional matrix of the body's tissues? All these questions fall under the broad theme of tissue architecture, which is one of the major themes at this year's Lorne cancer conference.

"It's something different," says Dr Ricky Johnstone from the Peter MacCallum Cancer Centre in Melbourne, one of the conveners of the program. "The conference will still cover many of the areas we have covered in the past, such as oncogenesis and angiogenesis, but tissue architecture is an area that is quite new -- it has really taken off in the last couple of years."

Tissue architecture is also about how best to treat cancer cells in the three-dimensional biological space. And there is a growing realisation that researchers need to make things reflect the real-life situation -- for example, breast cancer cells grown in a three-dimensional matrix that better resembles the environment of the body are much more resistant to chemotherapy than if grown in a flat, two-dimensional space.

Tumour-stroma interaction

"Cancer in the lab can be very two-dimensional," says Johnstone. "You often don't have any stromal cells within a sample of tumour cells in the lab, but it is becoming very clear that stromal cell interactions with tumour cells are very important."

The idea that stromal cells interact with tumour cells, and play as much a role in tumourigenesis as tumour cells do, is the focus of the work of invited speaker Assist Prof Valerie Weaver from the University of Pennsylvania in the US.

Weaver is using a model system of breast cancer, which includes a three-dimensional basement membrane assay, to look at structural and behavioural characteristics of breast cancer cells. Tissue polarity is a critical feature of organised tissues and facilitates the regulation of growth and death decisions in cells. One of the focuses of Weaver's lab is the identification of factors that mediate tumour 'dormancy' and tumour resistance to apoptosis.

Compared with normal cells, cancer cells, to a certain extent, lose their cellular organisation and this includes a loss of orientation or polarity. For example, cells that line blood vessels have a luminal and a basal side, enabling them to transport things to and from the blood. Cancer cells can lose this polarity.

Weaver has recently published research that shows that a cancer cell's orientation, or polarity, can be reinstated by molecular manipulation with endocrine or growth factors, after which the cells become resistant to radiation and chemotherapy. This is one of the reasons why normal cells resist these aggressive therapies.

Cell motility and the cytoskeleton

The cross-talk that goes on between cancer cells and their environment also involves the physical mechanisms of cell movement and cell-cell contacts.

Prof Alpha Yap's avid interest in cell motility and the cytoskeleton is what drew the conference organisers to invite him to this year's meeting. Attending the conference for the first time, Yap, from the Institute for Molecular Biosciences in Brisbane, sees himself as a bit of an outsider attending a conference focused on cancer biology, although he does admit that cell motility is "perhaps a little underestimated" within the context of cancer.

Yap says enormous progress has been made in cell motility in the last 10 years or so. This includes insights into the regulation of cell movement by external cues and internal signals, and the translation of these signals into the physical mechanisms of cell movement and cell-cell contact.

"Cell motility is a fundamental cell process -- all cells have the capacity to move, and it is movement in a regulated fashion," says Yap. "In adult life it involves a huge amount of regulation.

"To move, cells need to be able to generate force and they do this via the cytoskeleton; they also need to be able to exert traction and they do this via adhesion molecules; and they need to be able to orientate themselves and this relates to polarity," explains Yap.

Tumours become motile very early in their development, when they are only millimetres in size, and invade local or distant tissues via the lymph and blood vessels. Understanding this abnormal process of cell motility is pertinent to understanding cancer and a lot of work in this area of cancer biology focuses on these early events in the cell as tumourigenesis proceeds.

Yap's research focuses on how cells form contacts with one another, in particular the cadherin superfamily of cell-surface adhesion molecules. E-cadherin, the principal cadherin molecule found in epithelial tissues, is a key mediator of cell-cell recognition. It participates in tissue patterning and its dysfunction contributes to tumour progression and invasion.

"We are studying the molecular and cellular mechanisms by which cadherin cell-adhesion molecules mediate cell-cell recognition," says Yap, whose team has found that E-cadherin functions as an adhesion-activated cell signalling receptor. In particular, upon adhesion, the transmembraneous E-cadherin molecule activates signalling via a number of protein complexes, including the Arp2/3 protein complex, an intracellular protein machine that nucleates the assembly of actin filaments. The actin then generates the forces required to bring cells together.

Prof Denise Montell, of Johns Hopkins University in the US, is another invited speaker who will present her work on cell motility and the cytoskeleton. Montell is studying a model of the developmental process in the Drosophila fruit fly. By looking at the movement of cells in the egg chamber of flies, she has identified genes that serve as switches for cell movement. She is also looking at external cues, which often turn out to be growth factors, that signal when a cell should start moving and which direction it should go in.

Epithelial-mesenchymal development

Epithelial-mesenchymal transition (EMT) goes hand-in-hand with cell motility in allowing cells to move during development. As well as being vital in embryonic development, EMT processes are implicated in the conversion of early-stage tumours into invasive malignancies.

During the EMT process, epithelial cells lose polarity and cell-cell contacts and undergo dramatic cytoskeletal remodelling. Concurrent with this loss of epithelial cell adhesion and cytoskeletal components, cells undergoing EMT acquire mesenchymal cell expression profiles and a migratory phenotype.

A regular attendee at the Lorne cancer conference, Assoc Prof Rik Thompson, at St Vincent's Institute in Melbourne, says this year's line up of talks on plasticity, polarity and tumour-stromal interactions in cancer is an exciting development.

EMT has been a research focus of Thompson's since the late 1980s and he is pleased the area is finally becoming better accepted. "Ten years ago you could hardly convince anyone this was real," says Thompson. "But there is a growing level of acceptance now. Even Bob Weinberg [at Massachusetts Institute of Technology], one of the original discoverers of oncogenes, is talking about EMT!"

Thompson collaborates with Dr Don Newgreen of Murdoch Children's Research Institute and Assoc Prof Leigh Ackland of Deakin University, in looking at EMT processes in a novel breast cancer system that exploits the PMC42 cell-line -- a cell line derived from work initially done by Dr Bob Whitehead, who is now based at the Ludwig Institute for Cancer Research in Melbourne.

The epithelial PMC42 cells can be driven by epidermal growth factor (EGF) to become mesenchymal and invasive.

Newgreen, who will also be presenting at Lorne, has been studying EMT from the developmental angle for many years. "Don brings a wealth of experience from a developmental perspective -- he was the first to show that the best described embryonic EMT, that of neural crest cells, is started by reducing cadherin cell-cell adhesion function," says Thompson.

Thompson says breast cancer cells appear either epithelial or mesenchymal -- epithelial cells are relatively benign, whereas mesenchymal cancer cells are aggressive and undergo metastasis. Since they all come from an epithelial mammary gland, the researchers reasoned that the mesenchymal cells must have come from epithelial cells. However, getting evidence for this in the in vivo situation is difficult. The best functional data was recently published in a paper in which researchers caught cancer cells in the EMT act, and showed that FSP-1, a mesenchymal gene product, has a gatekeeper effect in EMT.

Husband and wife team Dr Carmen and Prof Walter Birchmeier, at the Max Delbruck Centre for Molecular Medicine in Germany, have worked together in the EMT area for the last 20 years and are among the few who have produced some hard data on EMT. The Birchmeiers, both invited speakers at Lorne this year, made the seminal discovery that the structural link protein at cadherin junctions, beta-catenin, has another role as a transcription factor regulating key genes in EMT. As well as showing that beta catenin is involved in the cell-adhesion signalling pathway, they have also shown that mutations in the beta-catenin gene correlate some types of cancer.

Last year, the Epithelial Mesenchymal Transition International Association (TEMTIA) was created at an EMT meeting held in Port Douglas - a meeting in which the Birchmeiers and Betty Hay, who coined the phrase EMT and played a significant role in establishing the field, participated. TEMTIA will be holding their next conference in Vancouver in 2005 (www.magicdatabases.com/TEMTIA.temptia.html). The formation of TEMTIA, along with the program at this year's Lorne conference, suggests that EMT research may be an idea whose time has arrived.

Mesenchymal cell markers

One explanation for the EMT process is that clonal evolution occurs in the primary tumour that, through genetic changes, leads to epithelial cells adopting a mesenchymal profile enabling the spread of the cancer. However, Rik Thompson, Don Newgreen and Leigh Ackland are beginning to think differently. "By comparing the primary tumour with tumours at metastatic sites, we are starting to appreciate that they look very similar," says Thompson.

Thompson proposes that more transient changes may be occurring. That is, the primary tumour may adopt changes temporarily and release aggressive mesenchymal cells into the blood that recolonise at metastatic sites, such as bone or brain, and then, once the cells have established the changes are reversed and cells begin to grow in the same way as the primary tumour.

"EMT is a temporary change in development so it fits with this idea," says Thompson. "And EMT may well be an epigenetic environmental change rather than hard-coded changes in these cells."

Most work on EMT is done on cell-lines, which can be misleading since they can behave very differently to cells in tumours. However, cell lines have shed significant light on the specifics of the EMT process. Mesenchymal cells are also much more able to survive as single cells in culture, which fits with their ability to be transported in the blood and tissue during the metastatic process.

Thompson and colleagues are also using mesenchymal cell markers to look for evidence of tumour epithelial cells turning into mesenchymal cells.

Vimentin, a marker that appears quite late in the transition to mesenchymal cells, is one that they have been employing to track these cells. But they hope to find more unique markers that are expressed earlier in the EMT transition period from benign to aggressive cells.

Thompson says they have some potential leads from gene profiling analyses -- snail and slug are two genes that are switched on early in the EMT process by EGF, and are also central to embryonic EMTs.

EMT is certainly more widely implicated than breast cancer, as it also occurs in the kidney and lung during development. It is also very prominent in colon cancer and fibrosis of the kidney. "Cells at the invasive edge of colon cancers look clearly mesenchymal," says Thompson.

Although more research is needed to target these mobile cells, they could provide a potential future diagnostic in acting as biomarkers for the detection of cancer.