Feature: Fighting cancer with proteomics

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

One of the most dangerous characteristics of cancer cells is their ability for uncontrolled and rapid multiplication. Anti-cancer chemotherapy drugs that interfere with cell division and cause cell death are therefore highly effective in the treatment of many types of cancer.

Many of these agents target and disrupt the cytoskeletal microtubule network – Paclitaxel (marketed as Taxol), and its derivatives such as Docetaxel (Taxotere) being amongst the most well known. Unfortunately, cancer cells can develop resistance or have differential sensitivities to anti-cancer drugs, with some cancers, such as lung cancer, being notoriously difficult to treat.

Working out how and why this happens by identifying the proteins and pathways involved is a major focus of many cancer research groups, including that of Associate Professor Maria Kavallaris, who heads the Pharmacoproteomics Program at the Children’s Cancer Institute Australia, which is now part of the brand new Lowy Cancer Research Centre at the University of New South Wales.

The ultimate aim of such research is to change either the drug or the cells themselves to get around the resistance and achieve more effective therapies for the cancer patient every time. A long-term goal of her group is to understand the role of cytoskeletal proteins in cancer cells so that improved treatments for cancer can be developed.

At HUPO 2010, Kavallaris will discuss some recently published and unpublished work that brings together models of epithelial cancers, some basic cell biology and a bunch of proteomic data. And what this clever mix is revealing about mechanisms of resistance to anti-cancer drugs and the roles of the tubulin proteins is very exciting indeed.

All living cells rely on the structural architecture provided by the cytoskeleton for a range of critical processes like cell division, cell migration and intracellular transport. The part of the cytoskeleton that has always been of particular interest to Kavallaris is the microtubule system, which comprises a dynamic network of long hollow tubular structures and associated proteins that radiate throughout the cell, doing their thing.

The microtubules themselves are conglomerates of the proteins α- and β-tubulin, which are continually popping on and off the ends to lengthen and shorten the tubules. This assembly and disassembly is particularly evident during cell division when the cell undergoes dramatic structural changes. According to Kavallaris, targeting this dynamic mechanism is a particularly good way to block cancer cell proliferation, and many such agents have been developed over the years to become an important part of chemotherapy regimes for many childhood and adult cancers.

---PB---

Unravelling resistance

Work in recent years by the Kavallaris group is starting to unravel some of microtubule system-associated mechanisms underlying the different responses of cancers to drug therapy.

“My research focuses on the fundamental aspects of the microtubule system in cancer cells,” she says. “In particular, my lab is trying to work out what the tubulin proteins are doing during cell division. It is important to understand what happens when a drug target such as tubulin alters itself in a cancer cell to mediate drug resistance. A number of α- and β-tubulin isotypes have been identified, and we now know from recent work that changes in the levels of tubulin isotypes can influence whether a tumour cell will respond to certain chemotherapies targeted at the microtubule system.

“A decade ago, people were still thinking that the various tubulin isotypes were functionally redundant, and it is really only recently, through a mixture of proteomic and functional analysis, that specific functions have been assigned to some of these isotypes,” says Kavallaris.

One of the major hold-ups in this research has been that tubulin is such an abundant protein, and it is also incredibly tightly regulated. So if you try to overexpress the protein, you can get compensatory changes due to all the extra tubulin hanging around.

“We are actually fortunate in that some of the tubulin isotypes we were working on are not as abundant in the cell. One of these proteins is Class III β-tubulin (βΙΙΙ-tubulin), which is normally expressed at high levels only in neuronal cells and the Sertoli cells of the testis.

“We now know from a bunch of studies by our group and others that high expression of βΙΙΙ-tubulin is associated with a whole range of epithelial tumours, such as lung, breast, ovarian and prostate cancer, as well as with some non-epithelial cancers. In lung cancer, this isotype is clearly with more aggressive and drug-resistant tumours. We were then able to show using gene-silencing that switching off this gene made certain cancer cells more sensitive to microtubule-targeting drugs such as Taxol.”

---PB---

Cancer Research

In her 2007 paper in Cancer Research, Kavallaris showed that depleting lung cancer cells of their high βΙΙΙ-tubulin expression in vitro not only made them more sensitive to tubulin-targeting agents, but it also sensitised the cells to DNA-damaging agents.

“This was really intriguing and started to tell us that this particular tubulin protein was having a broader role in cancer cells – not just in the microtubule system. This has important implications for improving survival in patients with drug-refractory cancers that overexpress this isotype.”

Kavallaris then moved into a mouse model to see if these effects of modifying βΙΙΙ-tubulin were translatable. Would targeting βΙΙΙ-tubulin actually influence how tumours in mice respond to chemotherapy and how does βIII-tubulin mediate this broader role in drug resistance?

Her group generated stable knockdown of βIII-tubulin using a short hairpin RNA (shRNA) construct in mouse models of non-small-cell lung cancer (NSCLC), which is a particularly belligerent and intractable form of lung cancer in humans with a dismal prognosis. Indeed, NSCLC remains the most common cause of cancer-related death worldwide.

In very recent work using these models, published this year in Cancer Research, Kavallaris and her group found that xenografts of the NSCLC tumour cells with reduced βIII-tubulin expression were more responsive to the DNA-damaging drug, cisplatin, and that the mice carrying these grafts survived much longer than control mice engrafted with tumours expressing high levels of βIII-tubulin.

“Although this result would seem partially predictable based on the in vitro data, it was nevertheless important to show this in a mouse tumour model,” Kavallaris says. “Usually when we do these drug resistance studies in the mice, we wait until all the tumours reach a given size to start the treatment so we are dealing with responses to the same tumour volumes. In these experiments, the tumours in the knockdown mice were actually taking much longer to reach that starting size.

“So we then asked whether βIII-tubulin was involved in tumour growth, based on number of clinical studies had showed that tumours in NSCLC patients not even given the option of chemotherapy that had high levels of βIII-tubulin had tumours that progressed, or grew, more rapidly,” Kavallaris says.

Her group found their NSCLC cells with suppressed βIII-tubulin expression had reduced anchorage-independent growth, suggesting a role for βIII-tubulin in tumourigenesis. Further experiments showed this effect to also be specific for this isotype.

“We then found that if you xenografted 20 mice, the βIII-tubulin knockdown cell grafts only produced tumours in 14 out of the 20 mice compared to 100 per cent growth of control tumours. This reduced incidence really surprised us because we know how aggressive these NSCLC cells are, but it seems that once they get into the tumour environment they start slowing down. It is also the first direct evidence that a β-tubulin isotype is linked to tumour formation.”

---PB---

Proteomics to the rescue

With this clear indication that βIII-tubulin is indeed a multifunctional protein in cancer cells, Kavallaris was very keen to find out exactly how this protein was working outside of its effect on microtubule structure in their βIII-tubulin-suppressed NSCLC cells. So this is where proteomics joined the cause to work out what was going on in terms of the tumorigenic phenotype.

Using the fluorescence-based 2-D DIGE system (difference in gel electrophoresis), Kavallaris’ group analysed the NSCLC cells with βIII-tubulin suppressed – making nuclear and cytosol fractions and separating proteins based on size and pH. This proteomics approach identified a number of protein changes in the cancer cells associated with various cell processes – cytoskeleton, the cell cycle, signalling and more – but what really caught their attention was a few tumour suppressor genes whose expression at both the gene and protein level went up in expression in the βIII-tubulin-knockdown cells.

The final piece in this series of experiments was then to ‘rescue’ the effect caused by the suppressed βIII-tubulin. “We generated some rescue clones whereby we took the knockdown cells and expressed another vector so that we basically restored the βIII-tubulin levels back to control levels in the knockdown cells. We found that the expression levels of one of the tumour suppressor genes that had gone up returned to normal levels, that is, normal levels of βIII-tubulin suppressed that particular gene again.”

Besides further validating their shRNA stable knockdown models for in vivo work, these latest results show for the first time that βIII-tubulin has a specific and very important role in tumour development, progression and drug sensitivity in human lung cancer. Kavallaris will talk about this work in her presentation at HUPO, and is currently writing up the results for publication.

“Whether what we have found from our proteomic data is a direct effect of βIII-tubulin on tumour formation, we don’t know. However, it really is quite exciting, and even though we have a long way to go to work out exactly what it happening, we have a starting point,” Kavallaris says.

“Normally, an epithelial cell from that site in the lung would not express βIII-tubulin and so these NSCLC cells have obviously upregulated this protein. It is intriguing to speculate that βIII-tubulin is directly involved in regulating certain tumour suppressor proteins, and the next stage of our research is aimed at unravelling this pathway.

“In terms of potential clinical strategies we are developing, we plan to take a multipronged approach to target the pathway that is regulating βIII-tubulin in the tumour cells, and to develop new therapeutics based on modulating or silencing βIII-tubulin in cancer cells to reverse the phenotype induced by these proteins in tumours.

“Based on our preclinical data, this would be expected to sensitise the tumours to chemotherapy, as we have shown our in vivo models. In addition, we are also hoping that it may actually reduce tumour growth. How we proceed all depends on the work we are doing right now.”

Kavallaris is already in discussions with clinicians and Australian biotech company Benitec, who are collaborating with Maria Kavallaris for the development of a DNA-directed RNAi therapeutic for lung cancer.

“Based on what we have shown so far, I believe this work has great potential as a treatment for lung cancer, especially in these cases where treatment is not even an option as it is.”

Maria Kavallaris did her PhD at UNSW on drug resistance in childhood leukaemia and was an original staff member of the Children’s Cancer Institute Australia when it opened in 1984. In 1996, she received an international cancer research fellowship to the Albert Einstein College of Medicine, New York. There she worked with Susan Horwith, a pioneer in the microtubule-targeted drug field who discovered how taxol works. She has received a number of awards and prizes, including an AACR Women in Cancer Research Award, a Young Tall Poppy Award, and an Australian Museum Eureka Prize.