Astex Pharma $ASTX at AACR - peak at next potential clinical candidates

Astex Pharma will present a number of abstracts at next week's American Association for Cancer Research meeting. Among them may be a peak into the preclinical drug candidates next poised to enter the clinic in 2012. We may also get details at the March 27th Research and Development day. See the full abstracts below.

Abstract Number: 2018
Discovery of potent dual inhibitors of both XIAP and cIAP1 using fragment based drug discovery

XIAP and cIAP1 are members of the inhibitor of apoptosis (IAP) protein family. Both proteins have the ability to attenuate apoptosis induced through intrinsic and extrinsic stimuli via inhibition of caspase-3, -7, -8 and -9. The defining feature of both XIAP and cIAP1 is the presence in their protein sequence of 3 Baculoviral IAP Repeat (BIR) domains, which are necessary for their antiapoptotic activity. The mitochondrial protein SMAC uses its N-terminal region (AVPI) to interact with BIR domains and deactivate the antiapoptotic function of IAPs.
Several companies and academic groups have active programs developing SMAC peptidomimetic compounds based on the AVPI motif. In general, those compounds have the tendency to be cIAP1 selective like their tetrapeptide progenitor (AVPI IC50 values for XIAP-BIR3 and cIAP1-BIR3 are 0.3 uM and 0.016 uM respectively).

Using our fragment-based screening approach, PyramidTM, we identified a non-peptidomimetic chemotype which binds with similar potency to the BIR3 domain of both XIAP and cIAP1. Hit optimisation using a structure based approach led to the discovery of potent true dual XIAP and cIAP1 antagonists with good in vivo physico-chemical profile and no P450 or hERG liabilities. Dual XIAP/cIAP1 inhibitors have potential for more effective apoptosis and less toxicity associated with cytokine production.

Compounds were initially characterised in fluorescence polarisation binding assays using XIAP-BIR3 or cIAP1-BIR3 domains. Robust induction of apoptosis was observed in two sensitive breast cancer cell lines (EC50s well below 0.1 uM in EVSA-T and MDA-MB-231); whilst HCT116 cells (colon cancer) were insensitive (unless exogenous TNF-α was added). This in vitro cell line killing was demonstrated to correlate closely with cIAP1 antagonism and hence a parallel cell assay was established to measure XIAP antagonism. An engineered HEK293 cell line was stably co-transfected with full length FLAG-tagged human XIAP cDNA and full length (untagged) human caspase-9 cDNA. Inhibition of caspase-9 binding to XIAP was measured in immunoprecipitation assays. This gave us a sensitive read-out for XIAP antagonism in cells which could be plotted against the most sensitive cell killing read-out (from the EVSA-T cell line) to establish relative XIAP vs cIAP1 selectivities and to select dual antagonists of both IAPs. Potent compounds (HEK293-EC50 <0.01 uM and EVSA-T-EC50 <0.01 uM) were further characterised in PKPD studies in mice bearing MDA-MB-231 xenografts. Compounds with good oral exposure achieved high concentration in tumor over 24h periods which ensured excellent inhibition of both XIAP and cIAP1 with consequent reduction of cIAP1 levels and induction of apoptosis markers (PARP, Caspase-3). Finally, dual XIAP/cIAP1 inhibitors have been investigated in xenograft models (melanoma, breast and colorectal cancer) and have achieved significant efficacy at tolerated doses.

Abstract Number: 2315
Pharmacodynamic and antitumor activity of fragment-derived inhibitors of MetAP2 in tumor xenografts

Background: Methionine aminopeptidase (MetAP) 2 is one of the two eukaryotic enzymes responsible for cleaving the N-terminal methionine from newly synthesized polypeptides to allow further post-translational modifications such as myristoylation to take place. The enzyme is the target of the anti-angiogenic natural product, fumagillin and so is believed to play a role in angiogenesis. Fumagillin analogues have shown activity in several disease models including oncology, inflammation and obesity indicating MetAP2 is a promising target in a number of indications. The semi-synthetic fumagillin analogue, TNP470, has shown activity in a Phase I/II cancer trial, suggesting that MetAP2 is a good oncology target. Here we describe the discovery, optimization and anti-tumor activity of fragment-derived MetAP2 inhibitors.

Results: We identified fragment hits to MetAP2 using our fragment-based screening approach, Pyramid™. These were optimized by structure-based drug design to novel, potent lead compounds with sub-100 nM potency against the isolated MetAP2 enzyme. The two most advanced compounds inhibited proliferation of HUVECs with potencies of 130 nM and 300 nM. Levels of the MetAP2 substrate, Met-14-3-3, were shown to increase on treatment of HUVECs with these compounds indicating that Met-14-3-3 was not being processed and hence MetAP2 was being inhibited. The compounds also inhibited HUVEC tubule formation demonstrating their anti-angiogenic properties.
The two lead compounds were further tested in vivo. Both compounds were well tolerated at doses up to 200 mg/kg bid. Levels of Met-14-3-3 were seen to increase in the thymus and spleen (where high expression of MetAP2 has been reported) of mice treated with the compounds, indicating again that MetAP2 was being inhibited in these tissues. Compound 1 was tested in a mouse HCT116 xenograft model. Mice were subcutaneously inoculated with HCT116 cells and oral dosing at 200 mg/kg bid started one day later. Tumor growth was inhibited in treated mice compared with control and growth inhibition was greater in Compound 1 treated mice (T/C 46%) compared with mice treated with 30 mg/kg TNP470 subcutaneously q2d (T/C 61%).

Conclusions: The compounds described here are novel small-molecule inhibitors of MetAP2. Their promising tumor growth inhibitory properties merit their testing in further tumor models and potentially other indications.

Note: the two abstracts above are products of the Astex Therapeutics legacy fragment-based drug discovery platform, while the two abstracts below are products of the Supergen legacy "CLIMB" drug discovery platform.


Thanks to a helpful ASTX scientist who pointed out my mistake in the prior version of this post, which did not categorize the PKM2 program correctly.

Abstract Number: 3762
Second generation PIM inhibitors exhibit improved activity in solid tumor models

The proto-oncogene PIM kinase family (PIM-1, -2 and -3) includes constitutively active serine/threonine kinases upregulated in multiple cancer indications, including lymphomas, leukemias, multiple myeloma, prostate and bladder cancers. Overexpression of one or more PIM family members in patient tumors frequently correlates with poor prognosis. The PIM kinases function by inhibiting apoptosis in MYC-driven tumors, and promoting tumor cell survival and proliferation. In the HEK-293T cell line, enhanced PIM kinase substrate BAD phosphorylation (pBAD) was observed following PIM and BAD overexpression. Enhancement of pBAD was inhibited by SGI-1776, a well-described PIM inhibitor, and more effectively by second generation PIM inhibitors exhibiting 4-10 fold improved potency against the PIM kinase family. The current PIM inhibitors display sub-µM activity in pharmacodynamic marker, proliferation and 2D colony formation assays using the UM-UC-3 human bladder cancer cell line. PIM1 and PIM2 overexpression models were established in the human prostate cancer cell line 22RV-1 and the non-tumorigenic mouse NIH-3T3 cell background. Overexpression of PIM kinases led to enhanced cell growth and tumorigenicity in both NIH-3T3 and 22RV-1 cell lines. In vivo xenograft studies using both PIM overexpression models and a clinically relevant solid tumor model facilitated identification of a lead candidate with demonstrated efficacy and favorable toxicity. IND-enabling studies with a lead candidate are underway.

Abstract Number: 3226
PKM2 metabolic activator slows lung cancer xenograft growth.

Tumor cells primarily utilize aerobic glycolysis, rather than oxidative phosphorylation, to metabolize glucose (the Warburg effect). The M2 splice form of pyruvate kinase (PKM2), the enzyme catalyzing the rate-limiting final step of glycolysis, is highly upregulated in tumors. Unlike the M1 splice form (PKM1), a constitutively active tetramer found predominantly in non-cancerous tissues, PKM2 is an inactive dimer under normal physiological conditions. Tetramerization of PKM2 requires binding of the allosteric activator fructose-1,6-bisphosphate (FBP), an upstream glycolytic intermediate, resulting in a fully active enzyme. Inactivation of PKM2 by cancer cells may allow glycolytic intermediates to be diverted into other biosynthetic pathways necessary for biomass production. The finding that PKM2 rather than PKM1 expression enhances tumorigenicity suggests that activators of PKM2 may have anti-tumor properties. We have identified and developed a series of small molecule PKM2 activators that exhibit low nM activation activity in biochemical and cell-based assays that measure pyruvate and ATP production. The extent of activation of these compounds is equal to or greater than that of FBP in biochemical assays. In addition, preliminary studies show that PKM2 activators inhibit the growth of lung cancer cell lines in vitro. The current lead compound was tested in established subcutaneously implanted A549 lung adenocarcinoma xenografts, where we observed a statistically significant decrease in tumor growth, with no observable toxicity. These data suggest that this class of PKM2 activators is effective as tumor cell metabolic regulators with anti-tumor activity for lung cancer and potentially other malignancies.