The overall direction of the Molecular Mechanisms of Tumor Promotion Section is to understand the regulation of the signalling pathways downstream from the lipophilic second messenger diacylglycerol, to elucidate the basis for heterogeneity of response to different ligands which function through this pathway, and to exploit this understanding for developing novel ligands with unique behaviour that function through this pathway. A complementary direction is to understand the regulation and structure activity relations for the vanilloid receptor. The vanilloid receptor is a downstream target of the diacylglycerol signalling pathway, shares partial homology in its ligands to this pathway, and shares with the diacylglycerol signalling pathway an important role in inflammation. Both directions impact both our understanding of biological regulation and the potential development of therapeutic agents. Protein kinase C, the best studied downstream target for diacylglycerol, represents the classic system for tumor promotion and is a therapeutic target for cancer chemotherapy. The vanilloid receptor represents a promising therapeutic target for cancer pain, among other indications, and thus represents an important direction in palliative care for cancer patients. The C1 domain, the interaction domain of diacylglycerol in protein kinase C or RasGRP, forms a complex with ligand and lipid. Studies using combinatorial libraries of diacylglycerol lactones reveal that apparently minor changes in the nature of the lipid interacting groups on the diacylglycerol lactone have substantial effects of the pattern of response selectivity. In collaboration with Victor Marquez and Raz Jelinek, we have characterized the nature of the ligand interactions with lipid bilayers using a range of biophysical methods. We show that there is marked diversity in the how such ligands interact, with self association, surface binding, and bilayer penetration all contributing to variable degrees. These insights provide new guidance for ligand design. The Vav family of Rho-GEFs possess C1 domains which have a homologous 3-dimensional structure to that PKC or RasGRP but which fail to bind diacylglycerol or phorbol ester. Using site directed mutagenesis, we have identified the specific residues responsible for this lack of binding and have designed a variant Vav C1 domain which now does bind. In further work, we have developed ligands which show enhanced selectivity for a C1 domain modified to more approach that of Vav. We conclude that lack of binding results from cumulative changes, none of which alone is sufficient to abrogate recognition. These changes principally alter lipid recognition. Identification of the nature of the critical changes provides a guide for the design of novel ligands targeted to Vav, a protooncogene and a critical signaling regulator. In collaboration with the chemistry group of Gary Keck, we have shown that a close analog of bryostatin 1 fails to show this antagonism on U937 leukemia cells although it retains comparable potency to bryostatin 1 on protein kinase C. Other derivatives retain the unique behavior of bryostatin 1, focusing attention on critical structural features of the molecule responsible for the bryostatin like behavior. A critical structural conclusion is that the A,B ring system in the bryostatin structure is NOT simply a linker region, as had previously been hypothesized. Our studies are further providing insights into which structural features do form the basis for bryostatin like behavior. In further studies, we have shown that, for different responses, the same bryostatin analog may give variable proportions on antagonism, ranging from virtually none to partial. We can thus conclude that the antagonistic behavior is not an all-or-none phenomenon. Our results imply that bryostatin analogs can be designed to antagonize a desired subset of protein kinase C responses. Mechanistic studies are zeroing in on the signaling changes that correlate with the bryostatin like behavior in the LNCaP cell, highlighting the role of PKC delta activity in a nuclear enriched subcompartment. We also find that transient duration of response is a critical aspect of bryostatin action in these cells. RasGRP3 is an activator of the Ras pathway directly stimulated by diacylglycerol and phorbol esters. We find that it is expressed at markedly enhanced levels in melanoma. Inhibition of the expression of RasGRP3 with siRNA in melanoma cell lines inhibits proliferation, growth in soft agar, and tumor formation in mouse xenografts. Reciprocally, overexpression causes enhanced tumor growth. Substantial RasGRP3 expression is also found in a range of other tissue types relevant to cancer such as lung and its role in these systems is currently being investigated. In the development of therapeutics targeted to TRPV1, a major problem is designing sufficient specificity of action. We are currently evaluating resiniferatoxin based ligands which show partial agonism / partial antagonism as a potential strategy for separating the side effect of initial pain induction from the therapeutic effect of desensitization. We have further developed through homology modeling a model of the ligand binding site on TRPV1 and are currently validating that model through site directed mutagenesis and photoaffinity labeling.
Showing the most recent 10 out of 95 publications
