Project 1: Phytochemicals and Metabolism in Pancreatic Diseases
Principle Investigator: Stephen J. Pandol, MD
Co-Principle Investigator: Anna S. Gukovskaya, PhD
Co-Investigator: Ilya Gukovsky, PhD
Co-Investigator: Aurelia Lugea, PhD
Co-Investigator: HongXiang Hui, MD PhD
Postdoctoral Fellow: Mouad Edderkaoui, PhD
Hypothesis: We propose that certain phytochemical decrease proliferation of stellate cells and cancer cells at least in part by inhibiting ROS-regulated pentose cycle pathways, resulting in decreased nucleic acid synthesis, decreased proliferation and increased apoptosis.
Roles of Pancreatic Stellate Cells in Carcinogenesis
Specific Aims for Stellate and Cancer Cells in Vitro
- Determine the effects of selected phytochemicals on ROS production, glucose metabolism, DNA synthesis, proliferation and apoptosis.
- Determine the effects of pharmacologic and genetic inhibition of ROS production on glucose metabolism, DNA synthesis, proliferation and apoptosis.
Specific Aims for Experimental Animal Models
- Determine the effects of selected phytochemicals on progression of the pathologic processes in chronic pancreatitis.
- Determine the ability of selected phytochemicals to prevent the initiation and/or progression of pancreatic cancer.
We have completed the experiments for aims 1, 2, 4 and partially aim 3. We found that phytochemicals ellagic acid, curcumin, rottlerin, embelin and lycopene dose dependently increase apoptosis and inhibit proliferation in both human pancreatic cancer cell lines MIA PaCa-2 and HPAF-II, and in stellate cells from the rat and mouse pancreas. We found that combinations of ellagic acid and embelin, or rottlerin and embelin had synergistic effects on proliferation and apoptosis in these cells. Interestingly, the combinations that showed synergism were at very low concentrations (between 0.5 and 3 uM) indicating possible use of these combinations in preventive and treatment strategies. Mechanistic studies showed that the effects of ellagic acid are due to down-regulation of the NF-kB pathway; embelin are due to the inhibition of STAT3 and its downstream effects on the expression of the inhibitors of apoptosis, survivin and XIAP; and rottlerin due to interruption of the interaction between pro-survival Bcl-2 and Bcl-xl with pro-apoptotic Bim and PUMA. In vivo experiments using subcutaneous and orthotopic models of pancreatic cancer showed inhibition of tumor growth when feeding mice with each of the compounds or combinations of ellagic acid and embelin or embelin and rottlerin. Inhibition of growth was dramatic in the 60-80% range.
We are currently making great progress in defining the molecular steps involved in regulating the metabolic phenotype of the pancreatic cancer cell through approaches of in silico modeling of the atomic interactions between a phytochemical and its direct target complemented by metabolic flux studies. The in silico approach started with our experimental findings that rottlerin promotes apoptosis by directly interacting with Bcl-2 and Bcl-xl resulting in dissociation of Bim and PUMA from their binding sites on Bcl-2 and Bcl-xl. The dissociated Bim and PUMA are necessary for the apoptosis because genetic deletions resulted in prevention of apoptosis. These findings led us to collaborate with experts in in silico modeling of biological molecules-Bill Goddard and Ravi Abrol from Caltech in Pasadena. We worked to develop an atomistic dynamic model of the interactions between Bcl-xl, Bim and rottlerin. We confirmed the model by experimental approaches in two ways. First, we used the model to predict how specific amino acid changes in Bcl-xl would change its interaction with Bim. Our predictions were tested by creating mutations in Bcl-xl predicted to increase or decrease binding of Bim and expressing the mutated forms in cancer cells. We then measured Bim binding to the mutated Bcl-xl expressed in the cancer cells. We not only found that the model correctly predicted how changes in specific amino acids would increase or decrease Bim binding but that increased binding of Bim was associated with increased proliferation whereas decreased Bim binding was associated with decreased proliferation and apoptosis. Furthermore, the changes in proliferation were due to the Bcl-xl-Bim interaction because genetic deletion of Bim prevented the effects on proliferation.
We have also been using our in silico model to screen large libraries of plant-derived and synthetic small molecules to identify others that interact with Bcl-2 proteins in a manner similar to rottlerin at low concentrations (i.e. 1-5 uM). We have identified several and have tested 3. To our satisfaction, the ones we have tested perform as predicted.
As described above, our work has led to the discovery of a novel and major role for Bcl-2 family protein interactions in regulating proliferation. That is, increases in the interaction of pro-survival Bcl-xl and pro-apoptotic protein Bim lead to proliferation while decreases lead to less proliferation and apoptosis. In order to determine how these interactions regulate proliferation we performed metabolic flux studies with Dr. Paul Lee using both rottlerin and the mutations that increased and decreased proliferation. The key finding was that the regulation of DNA synthesis was due to the effects of the treatments on flux of glucose through the G6PD pathway. This finding not only fits our hypothesis but demonstrates that Bcl-2 proteins are key regulators of the metabolic pathway involved in proliferation. These findings suggest that plant derived compounds may in fact have major effects on the cancer phenotype through regulation of a key cancer signal.
We have also investigated mechanisms underling rottlerin-induced cell death in the pancreatic stellate cell, a central mediator of desmoplasia in pancreatic cancer and pancreatitis. We found that rottlerin induces in a dose-dependent manner a sustained activation of the metabolic regulator AMPK and marked inhibition of mTOR. These effects were associated with rapid induction of autophagy, blockade in protein translation, endoplasmic reticulum (ER) stress, and activation of ER stress pro-apoptotic signals. These data indicate that phytochemicals may have a beneficial effect on both pancreatic cancer and pancreatitis by dysregulating metabolic signaling in the pancreatic stellate cell.
Plans for this coming year
We are now deeply invested in determining the steps in the pathways involved in regulation of the metabolic steps by Bcl-2 proteins as well as finishing the experiments proposed in aim 3. For the pathways we are considering regulation of NADPH oxidoreductases by the Bcl-2 proteins. We hypothesize that upregulation of NADPH oxidases would increase the ratio of NADP to NADPH.
This increase would result in promotion of flux through the G6PD pathway leading the nucleotide synthesis. We have significant experience in the mechanisms of regulation of NADPH oxidases so we will be successful in providing a detailed and atomic based model to describe how phytochemicals regulate the metabolic phenotype of a cancer through specific effects on key Bcl-2 proteins.
Future plans and translational potential
The work described above sets that stage for future projects with enormous potential for translation to human health. In brief, our work over the next several months will show the key molecular steps and targets involved in sustaining the proliferative metabolic phenotype of both pancreatic cancer and stellate cells. This work involves both atomistic modeling of phytochemical interactions with key molecular targets starting with the Bcl-2 proteins and including others such as components of NADPH oxidase; and metabolic profiling to show how phytochemical interactions with each of the targets disrupt the proliferative metabolic phenotype. The models we develop allow us to screen large databases of molecules to identify other agents that potentially interact with each of the targets. Also, by combining agents interacting with different and key steps in the regulation of proliferation, we will be able to create formulations with additive and synergistic effects. Overall, the findings will lead to formulations that can be used for treatment and prevention strategies.