Abstracts Keynote & Plenary

Email: jtus@phys.ualberta.caBackground/Problem:
The ultimate goal of cancer research is to
develop a drug or treatment regimen that will target only cancer cells with minimal damage done to
healthy tissues. The significance of microtubules as a molecular target for chemotherapeutic treatments
is outlined in a recent review [1]. Tubulin which makes up microtubules binds numerous small
molecule ligands, which result in the alteration of microtubule dynamics leading to cell cycle arrest and
cell death. Many of these ligands are currently used clinically for the treatment of several types of
cancer and include the drugs paclitaxel, colchicine and vinblastine. These drugs bind to one of three
distinct binding sites within beta tubulin, all of which have been recently identified through electron
crystallography. The drawback of these drugs is their indiscriminate binding to all cells leading to the
death of both cancerous and healthy cells. Hence despite the overall success of the vinca alkaloid and
taxane drug families side effects such as neurodegradation seriously impair the prognosis for many
cancer patients treated with them. Moreover, in many cases drug resistance develops in the course of
chemotherapy.
Tools and Methods:
We have focused on computational searches, optimization and testing new and
re-purposing old molecules that interfere with the formation of mitotic spindles during cell division in
tumors. To build the molecular models of our target tubulin, we used the program Modeller [2] that
uses alignment of the sequences with known related structures to obtain spatial restraints that the output
structure must satisfy. Missing regions are predicted by simulated annealing of a molecular mechanics
model. Only the default parameters of Modeller were used, as simple manual inspection of some of the
output structures suggests that reasonable models were produced.
Results: The existence and distribution of various tubulin isoforms is the basis for novel
chemotherapeutic drug design that can differentiate between different cell types to reduce side effects.
The quality of the resulting models for tubulin isoforms was investigated using two software packages
WHAT_CHECK [3] and PROCHECK [4] followed by an analysis of ten human beta tubulin isoforms
regarding their differences within each of the determined paclitaxel, colchicine and vinblastine binding
sites. New promising compounds representing taxane and colchicine derivatives have been designed
and computationally tested for isoform specificity and will be presented at this conference. They have
been synthesized and are being tested in our lab. The stabilities of these derivatives have been
computationally evaluated using both classical and quantum mechanical methods. In addition, based on
the mechanical properties of microtubules we are working on new treatment modalities that use
ultrasound, laser action and magnetic fields directly on cells.
Conclusions/Discussion: Significant progress has been made in our understanding of the key cellular
target for chemotherapy: tubulin. Based on our computational models, we have designed several dozen
new promising compounds that selectively bind to tubulin isoforms that reflect patient-specific
mutations in tumor cells. Our future plans include a massive computational effort to match every target
protein with an existing chemical entity.
References
[1] M.A. Jordan and L. Wilson (2004) Nature Rev. Cancer 4, 253-265.
[2] R. Sanchez and A. Sali, 2000, Methods Mol Biol, 143, 97-129
[3] R.W.W. Hooft et al., 1996, Nature, 381, 272
[4] R.A. Laskowski et al., 1993, J. Appl. Cryst, 26,
283-291
Acknowledgements: This project has been funded by MITACS,
the Allard Foundation, NSERC,
PIMS, CPCRI, US Department of Defense, Oncovista LLC (San Antonio, TX) and Technology
Innovations LLC (Rochester, NY).