CCI-001 Scientific Summary

Microtubules are cytoskeletal filaments consisting of α,β- tubulin heterodimers and are involved in a wide range of cellular processes such as organization of cell shape, transportation of vesicles, mitochondria, and other cellular organs, cell signaling, and perhaps most importantly: cell division and mitosis. Microtubules play a key role in mitosis separating the daughter chromosomes and constitute a strategic target in cancer treatment. Over the past several decades, the search for selective inhibitors of tubulin assembly or disassembly has led to the development of some of the most useful antitumor drugs currently in clinical use. These tubulin-targeting drugs are broadly classified as microtubule-stabilizing drugs (e.g. taxanes, epothilone) and microtubule-destabilizing drugs (e.g. vinca alkaloids). There are everal known ligand binding sites in the tubulin α/β-heterodimer: the paclitaxel binding site, the vinblastine binding site, the laulimalide/peloruside binding site, the noscapine binding site and the colchicine binding site.

Microtubule binding drugs are successful in oncological therapy (for example paclitaxel, vinblastine and vincristine) based on at least four types of molecular involvement that disrupt mitotic activity. These drugs have a variety of origins (plants, animals) and numerous synthetic versions have been developed and studied to further understand and improve their therapeutic properties.

In our early studies, we have used computational and bioinformatic tools to identify the most opportune binding site on tubulin, which would provide the greatest scope for the derivatization of a patent compound in order to have distinct binding affinities for specific tubulin isotypes. Tubulin is a protein, which is encoded by at least 20 different genes in the human body. These isotypes of tubulin are very specifically expressed by different tissues offering a design strategy for tubulin-targeting drugs. We have examined all binding sites on tubulin and compared their structure and amino acid composition between tubulin isotypes. Tubulin expression is critical to the assembly/disassembly equilibrium of microtubules during mitosis and hence it is important to know which tubulin isotypes are most actively involved in cell division in cancer cells. It is abundantly clear in the scientific literature on this topic that βIII-tubulin is the isotype of tubulin that is present in virtually all cancer cells but almost never expressed by normal cells with the exception of brain tissue and testes. Therefore, a strategy aiming to target βIII-tubulin, especially by drugs that do not cross the blood-brain barrier, offers promise that other tubulin-targeting drugs do not. We have compared all ligand-binding sites on tubulin and found the colchicine binding site to have the greatest diversity among tubulin isotypes, with a special structural difference in βIII-tubulin where 3 amino acids are different and there is an additional change in its electrostatic charge and polarity.

Colchicine, a well-known bioactive alkaloid derived from plants and a prototype microtubule disrupting drug is not clinically used to treat cancer yet, because of its overbearing systemic toxicity that produces unacceptable side effects when administered intravenously. However, the anti-proliferative effects of colchicine through the inhibition of microtubule formation, leading to mitotic arrest, anti-vascular disruption, and cell death by apoptosis, together with its inherent water solubility, commercial availability and low cost have renewed interest in modifying this molecule and developing less toxic hybrid compounds incorporating the colchicine scaffold. Colchicine and its analogs and derivatives are currently investigated as experimental cancer therapeutics in a number of clinical trials.

Colchicine and its derivatives bind to a well-localized region at the interface between α- and β-tubulin, resulting in a conformational change that prevents the tubulin dimer from polymerizing into microtubules. The new class of colchicine derivatives we have computationally designed and then synthesized and tested was computationally predicted to have increased binding specificity for βIII-tubulin (a tubulin isotype that is over-expressed in most cancer types), with the objective of increased toxicity against tumour cells, but fewer side effects on normal cells, compared to its parent and to standard chemotherapeutic drugs. The in silico drug design was followed by in vitro protein studies, and then cell culture growth inhibition, apoptosis and protein expression assays, which demonstrated significantly increased potency of several of the new derivatives when compared with the parent colchicine. In addition to the growth inhibition effects of the new compounds, they have also been shown to cause the apoptosis of cancer cells, and tests of the derivatives against a normal human fibroblast cell line showed relatively low cytotoxicity effects. We have studied more than 60 colchicine derivatives in cytotoxicity experiments, and defined the most potent ones. CCI-001 had a superior predicted ADMET profile of the two. Therefore, we have proposed the colchicine derivative CCI-001 as a candidate for a detailed evaluation of its in vivo profile to determine its uptake properties in a mouse tumour model. Following this work, we have also examined its toxicity in healthy rat models and found a maximum tolerated dose to be much less than the therapeutic level found in mouse studies.

Below we briefly summarize some pertinent numbers for the lead compound CCI-001.

Predicted binding free energies (kcal/mol) with respect to major tubulin isotypes:

βI: –53.1 βIIa: –34.4 βIIb: –39.1 βIII: –48.4 βIVa: –32.0 βIVb: –44.0 βV: –63.8 βVI: –47.6

Cytotoxicity IC50 (nM): A549: 3.2 NCI-H226: 4.8 CEM: 3.3

Cell survival at highest concentration: A549: 28% NCI-H226: 17% CEM: 0%

Predicted ADMET properties [ADMET Predictor 5.5]: LogP = 2.58

Solubility at pH 7.4 = 0.007 mg/mL
Topological polar surface area = 83.1 Å2
LD50, acute toxicity in rats = 279.4 mg/kg
TD50, carcinogenicity in rats over standard lifetime = 25.7 mg/kg/day

Additional in silico, in vitro and in vivo data are available, including a report from Oncolines, a CRO based in the Netherlands.