Funded Grants

Novel lipid targets in the treatment of human glioblastomas

Lipid signaling pathways are essential components in malignant transformation and metastasis. Cancer involves defects in the ability of the cell to accurately proof-read DNA during replication, as well as aberrations in the cell signaling pathways that normally transduce information from outside the cell into intracellular compartments and the nucleus. The normal machinery of programmed cell death is disrupted in cancer cells and constitutive activation of catalytic processes short-circuit the elegant signaling network that normally allows a cell to signal that an important change has occurred in the environment to which it must respond. In this feed forward cycle of constantly responding to a perceived stress the deleterious mutations accumulate until a cell becomes transformed. Among the most intractable of cancers are Glioblastoma Multiforms (GBMs), which typically have a poor prognosis and short median survival times. This is a devastating disease and therapeutic options are currently limited. Several obstacles are associated with successful treatment of GBMs. First, these tumors are not encapsulated and, because they are highly invasive, tend to spread to other regions before they are diagnosed. In addition, many conventional chemotherapeutic agents cannot cross the blood-brain barrier and therefore do not reach the site of malignancy.

Several well characterized oncogenic mutations in GBMs are known to cause constitutive activation in lipid signaling pathways. RAS GTPase is frequently mutated in human glioblastomas, leading to activation of PI3Kinase and subsequent activation of Akt. Phospholipase D (PLD) is intimately associated with this pathway. PLD is activated by oncogenic forms of tyrosine kinase receptors, such as EGFR. Phosphatidic acid (PA) generated by PLD is an essential modulator of pathways that directly regulate levels of PI(3,4,5)P3. PA is a membrane binding site for Raf kinase and has also been shown to modulate sphingosine kinase and the mTOR pathway. We recently discovered that PLD is in complex with Akt and appears to modulate the phosphorylation state of the kinase. PLD catalytic activity is hyperactivated in human GBMs, but it has not previously been possible to interrogate the therapeutic possibilities of this target. In collaboration with the laboratory of Craig Lindsley we have designed and synthesized the first isoenzyme selective inhibitors of PLD1 and PLD2 (Scott et al., 2009 Nature Chemical Biology) and additional compounds with increased potency and efficacy have been developed. The initial description of these compounds focused on effects in breast cancer cells, but subsequent work identified that human glioblastomas frequently have very high levels of PLD catalytic activity and the effects of the inhibitors on blocking invasiveness are striking (Figure, below). Preliminary results show that inhibition of PLD generated product, PA, effectively blocks invasive migration of a glioblastoma derived cell line, U87-MG. Our synthesized lead compounds have high selectivity (>1700 for EVJ) and potencies below 1 nM.

The goal of this proposal is to refine pharmacological properties of the compounds, define the molecular mechanism of these compounds (i.e. how PLD inhibitors block invasive migration), and optimize the medicinal chemistry to develop a preclinical candidate. The long term goal of this research is to determine whether PLD is an effective therapeutic target for the treatment of GBMs alone and/or if it synergizes with other therapeutic targets to effectively block GBM progression. Our hypothesis operates from the premise that no single agent will be entirely successful unto itself in treating brain cancers, but PLD may be a critical node that leads to more successful therapy. If validated we will seek to partner with a pharmaceutical company to develop our lead compound into a suitable candidate for clinical trials. Support from the McDonnell Foundation will be critical at this phase.

The central hypothesis to be tested in this research is that targeted ablation of the signaling pool of PA generated by PLD will be sufficient to significantly block invasive migration of GBMs and in combination with other signaling pathway inhibitors will provide a novel pharmacological agent that can block invasive migration, extend survival times and augment therapeutic treatment. The specific aims include:

  1. Synthesis of isoenzyme selective inhibitors of PLD1 and PLD2 will be followed by testing compounds in parallel cell based and biochemical reconstitution assays for determination of potency and efficacy. In addition to measuring effects on catalytic activity a variety of relevant cellular parameters including cell growth, proliferation, apoptosis, and sensitivity to radiation therapy will be measured. Compounds will be optimized for therapeutic use in glioblastomas by established structure-activity relationship (SAR) approaches (Scott et al., 2009 Nature Chemical Biology).
  2. Test potency and efficacy of inhibitors to block migration using matrigel invasion assay and 3¬dimensional anchorage independent growth (AIG) assays. Examine selected protein and lipid biomarkers following inhibitor treatment alone or as part of a combination therapeutic approach to identify biomarker to assess efficacy.
  3. Ex vivo analysis of drug metabolism will be determined following infusion of selective PLD inhibitors into U87-MG initially with interesting candidates being tested in primary biopsied GBMs derived neurospheres. Compounds and secondary metabolites will be measured using conventional electrospray ionization mass spectrometry (ESI-MS). More detailed metabolic profiles will be determined using an innovative new technology recently developed in the PI’s research group. Alkyne-analogues of PLD inhibitors will be followed in cell lines and tumors. Enhanced detection is achieved by covalent capture and release using cobalt complexation (Milne et al., 2010 Nature Chemical Biology).

The overall goal of this research plan is to develop potent (subnanomolar) and highly selective (isoenzyme specific) that minimize off-target effects. The unique approach in our research plan is that novel small molecules are synthesized and immediately screened in two parallel, and highly complementary, approaches. The compounds are screened in a biochemical assay in which only purified PLD1 or PLD2 are present with chemical defined lipid vesicles. This assures that the chemical series is working directly on the phospholipase. The assay details are described in detail (Brown et al., 1993 Cell; Brown et al., 2007 Methods in Enzymology). The parallel assay utilizes a cell-based approach in systems in which exclusive PLD1 and PLD2 responses have been established. The cellular PLD transphosphatidylation activity is measured by analysis of the formation of phosphatidylbutanol, which an alternate catalytic product of PLD. Other glycerophospholipid species are monitored in parallel. Off-target effects in other lipid signaling pathways are detected in the analysis of the spectrum. In contrast to many drug-discovery approaches we are able to identify immediately if a new series of compounds is ineffective or has substantial off-target effects. This allows us to rapidly assess SAR and focus on more productive chemical modifications. We have multiple series of compounds that inhibit PLD. Preliminary results reveal that the major effects of these inhibitors are to block invasive migration and promote cell death. Given that GBMs have high levels of constitutively produced PA and a major cause of death due to these tumors is uncontrolled invasiveness, PLD is an attractive target for therapeutic intervention. Our results have focused in the matrigel invasion assays, but we propose to extend these studies into 3¬dimensional anchorage independent growth assays. A number of relevant biological parameters will also be measured including effects on cell growth, apoptosis, sensitivity to irradiation and other chemotherapeutic agents, endocytosis of EGFR, and changes in metabolic pathways as biomarkers. This will serve as the primary determination for which compounds should be used for analysis of in vivo tumor models. These studies will also provide important information on entry of the inhibitors into brain and subsequent metabolism. The primary and secondary metabolites will be measured by online HPLC-ESI¬MS. Given the complexity of metabolites in glioblastomas, it may be necessary to utilize chemical probes. We have recently developed a new form of Click chemistry in which alkyne-labelled probes can be covalently complexed to cobalt and resolved from other biomolecules. This is essentially “affinity chromatography” for small molecule metabolites (Milne et al., 2010, Nature Chemical Biology). This approach allows us to detect metabolites after prolong periods of treatments with multiple biotransformation steps. This will facilitate our understanding of the metabolic fate of our small molecule inhibitors as well as assess bioactive compounds within the in vivo model.