To support their accelerated non-homeostatic proliferation, cancer cells depends on access to extracellular nutrients derived from the tumor vasculature. However, dysfunctional tumor vasculature and high interstitial pressure limit tumor perfusion leaving many tumor cells nutrient-deprived. To bypass this nutrient limitation, RAS driven cancer cells recycle their own nutrients via autophagy or scavenge macromolecules from the tumor microenvironment (TME) via macropinocytosis. Oncogenic mutations in RAS promote macropinocytosis, a bulk scavenging process through which components in the TME can be non-specifically acquired. When macropinosomes fuse with lysosomes, degradative enzymes release nutrients that can support cancer cell growth and proliferation. We report macropinocytosis is not just a RAS driven phenotype but a cancer phenotype. Lipid PIP3 generated by the PI3K pathway is necessary for closure of macropinosomes in most contexts. We identified prostate and breast cancers with hyper-activation of the PI3K pathway perform macropinocytosis in an AMPK dependent manner. AMPK mediated macropinocytosis is necessary for proliferation of PTEN-deficient cells in low nutrients but not autophagy. Because scavenging allows cells to acquire nutrients from cell extrinsic sources, it is likely to play a greater role in promoting tumor growth than cell-autonomous autophagy, which leads to atrophy. Necrosis is observed in most solid tumors; necrotic cell debris is selectively scavenged via macropinocytosis. Nutrient-deprived macropinocytic breast cancer cells acquired not just amino acids, but also sugars, lipids, and nucleotides from necrotic cell debris (necrocytosis) suggesting macropinocytosis plays a larger role than previously appreciated in supporting cancer anabolism. In addition, necrocytosis also conferred resistance to various therapies that target de novo nucleotide synthesis, fatty acid synthesis as well standard of care genotoxic therapies resulting in the first demonstration that scavenging confers drug resistance. Together, these studies suggest that macropinocytosis inhibitors could have therapeutic value in many cancer patients, particularly in combination with conventional therapies.

Despite the advances in therapeutics targeting key players that drive cancer metabolism, clinical benefits of these therapies have been limited. As tumors consist of heterogeneous population of cells, targeting a specific oncoprotein may provide selective advantage to cancer cells that are not driven by that oncoprotein. Also, upregulation of anabolic pathways that are not being directly targeted may also contribute to development of resistance to the therapy. Instead of focusing on a specific metabolic protein, our lab is interested in striking cancer metabolism at its apex—cancer cells’ dependence on nutrients. To fuel constitutive anabolism and proliferation, cancer cells must have influx of nutrients. When nutrients become limiting, unlike non-transformed cell that can discontinue anabolic processes and become quiescent, cancer cells undergo metabolic chaos as oncogene-driven growth signals continue to force anabolism. Developing a therapy against this cancer-specific phenotype will have efficacy against broad classes of cancers as all cancers need nutrients to survive. In searching for a compound that can inhibit nutrient uptake, our lab has been focusing on sphingolipids, a class of bioactive lipid that plays important roles in cellular functions including membrane biology. In yeast, stress-induced nutrient transporter down-regulation is regulated by sphingolipid phytosphingosine. Sphingolipid ceramide down-regulates nutrient transporters in mammalian cells. A sphingolipid-derived compound FTY720 is also able to trigger surface amino acid and glucose transporter loss. FTY720, an FDA-approved drug for treatment of multiple sclerosis, has anti-cancer activities in a various cancer models. Unfortunately, FTY720 causes bradycardia at the doses needed for anti-neoplastic effects. However, mechanism by which FTY720 kills cancer cells is completely separate from its effect on heart rate. SH-BC-893 is a constrained analog of FTY720 that lacks its dose-limiting side effect but retains anti-neoplastic activities. SH-BC-893, like ceramide and FTY720, down-regulates surface nutrient transporters. In addition, SH-BC-893 induces cytoplasmic vacuolation secondary to mis-localization of lipid kinase PIKfyve. This blocks lysosomal degradation of autophagosomes and macropinosomes, alternate nutrient acquisition pathways cancer cells can use to adapt to nutrient stress triggered by loss of surface nutrient transporters. LDL receptors are also down-regulated and LDL degradation is blocked by SH-BC-893. Prostate cancers rely heavily on LDL uptake for growth and proliferation and are thereby very sensitive to SH-BC-893. Surprisingly, prostate cancer cells are resistant to amino acid and glucose withdrawal. Our ongoing works have uncovered that this is because PTEN loss in prostate cancers upregulates macropinocytosis. This previously-unappreciated macropinocytosis in prostate cancers requires AMPK activation. Our works demonstrate that prostate cancer cells can use macropinocytosis to consume necrotic debris to build protein biomass and lipid storage to survive and proliferate in low nutrient environment. These findings highlight cancer cells’ addiction to nutrients and offer insights into therapeutic opportunities for targeting this Achilles’ heels of cancer metabolism. SH-BC-893 prevents cancer cells from using nutrients taken up via macropinocytosis by inhibiting lysosomal degradation of macropinosomes. By simultaneously blocking import and lysosomal degradation of nutrients, this water soluble and orally bioavailable compound effectively starves cancer cells to death. Future investigations will help further fine-tune SH-BC-893 into a safe and effective anti-cancer agent that can overcome the problems of tumor heterogeneity and adaptive resistances seen with current metabolic therapies.

Obesity has reached epidemic status, threatening millions of lives and straining national economies and health-care systems. Obesity is a chronic disease with diverse etiologies, arising from complex interactions between genetic, physiologic, psychologic and environmental factors that can undermine one’s ability to maintain a healthy weight. The current standard of care, behavioral and lifestyle interventions, has proven to be inadequate for a majority of obese patients thus use of pharmacotherapies has been proposed as an adjunct to lifestyle changes to overcome barriers to weight loss.

However limited efficacy and poor safety prevent the widespread use of current pharmacotherapies, prompting the demand for new drug targets and novel therapeutics.

893 was previously characterized as a synthetic sphingolipid analog that simultaneously blocks multiple pathways of cellular nutrient uptake. Because chronic nutrient overload underlies much of obesity pathologies, we hypothesized that such a molecule would be an effective anti-obesity agent by putting cells on a diet, rather than a whole patient. To test this, we used 893 in an established mouse model of diet-induced obesity (DIO) that recapitulates the gradual weight gain and metabolic decline that occurs in humans from passive overconsumption of energy dense foods. 893 safely restored normal body weight, reversed hepatic steatosis, and improved glucose homeostasis in obese mice maintained on a high-fat diet (HFD). 893 suppressed food intake and promoted whole body lipid metabolism in vivo. Prolonged HFD feeding induces ceramide-dependent mitochondrial fission that was directly blocked by 893 both in vitro and in vivo. This work demonstrates that 893 and related molecules represent a potential new strategy to tackle the growing obesity epidemic and highlights mitochondrial fission as a potential therapeutic target.

Endolysosomal trafficking determines the ultimate destination of molecules that enter cells and therefore has great impact on multiple cellular processes. Our lab has established that bioactive lipids called sphingolipids can rewire endolysosomal trafficking to result in accumulation of endocytosed material in pre-lysosomal compartments. In this thesis research, the impact of disrupted endolysosomal trafficking caused by sphingolipids on the activity of oligonucleotide therapeutics will be discussed. Therapeutic oligonucleotides include large RNA molecules such as mRNA vaccines and those that target RNA such as single-stranded antisense oligonucleotides (ASOs) or the double-stranded small interfering RNAs (siRNAs). ASOs base pair with a target RNA to elicit RNaseH-dependent degradation, inhibition of translation, or changes to splicing. siRNAs can produce prolonged target RNA silencing after being loaded into Ago2 to form the RNA-induced silencing complex (RISC). Given that all of these oligonucleotide therapeutic platforms require cytoplasmic entry for activity, inefficient endosomal escape remains the primary barrier to the broad application of oligonucleotide therapeutics. Liver uptake after systemic administration is sufficiently robust that a therapeutic effect can be achieved but targeting extrahepatic tissues remains challenging. Prior attempts to improve oligonucleotide activity using small molecules that increase the leakiness of endosomes have failed due to unacceptable toxicity. This thesis will discuss the development and characterization of a well-tolerated and orally bioavailable synthetic sphingolipid analog called SH-BC-893 that increases the activity of antisense oligonucleotides (ASOs) and small interfering RNAs (siRNAs) up to 200-fold in vitro without permeabilizing endosomes. SH-BC-893 treatment trapped endocytosed oligonucleotides within extra-lysosomal compartments thought to be more permeable due to frequent membrane fission and fusion events. Simultaneous disruption of ARF6-dependent endocytic recycling and PIKfyve-dependent lysosomal fusion was necessary and sufficient for SH-BC-893 to increase non-lysosomal oligonucleotide levels and enhance activity. In mice, oral administration of SH-BC-893 increased ASO potency in the liver by 15-fold. More importantly, SH-BC-893 enabled target RNA knockdown in the CNS and lungs of mice treated subcutaneously with cholesterol-functionalized duplexed oligonucleotides or unmodified ASOs, respectively. Together, these results establish the feasibility of using a small molecule that disrupts multiple endolysosomal trafficking steps to improve the activity of oligonucleotide therapeutics in extrahepatic tissues.

Sphingolipids are found in all kingdoms of life. In mammals, sphingolipids regulate multiple cellular processes. Ceramide is a sphingolipid of interest to cancer biologists due to its regulation of cell cycle, membrane morphology, cell migration, differentiation, and cell death. Disruptions in ceramide metabolism are associated with disease. Reduced ceramide in cancer cells leads to increased malignancy and metastasis. Conversely, increasing cellular ceramide leads to differentiation, proliferative arrest, and cell death. Our lab has developed synthetic sphingolipids that effectively and selectively induce cancer cell death in vivo. We have recently described one such synthetic sphingolipid, SH-BC-893 (893), that is selective and effective in murine cancer models without notable toxicity to the animal.

Our publications show that the selectivity of sphingolipids for cancer cells is due to oncogene-driven constitutive anabolism of cancer cells. Highly anabolic TSC2 knockout cells are resistant to sphingolipid-induced death but are sensitized by expression of oncogenic KRAS.

893 inhibits multiple, parallel cellular nutrient acquisition mechanisms. One mechanism is by decreasing surface nutrient transporter expression, an activity shared with the endogenous sphingolipid ceramide. However, 893 further limits nutrient acquisition by reducing lysosomal fusion to prevent access to intracellular metabolites from LDL, autophagosomes, and macropinosomes. These phenotypes were not observed with ceramide.

The molecular mechanism through which ceramide and synthetic sphingolipids induce loss of surface nutrient transporters is not yet known. The work presented here suggests that recycling of nutrient transporter proteins is inhibited. We show that actin remodeling, ARF6 inactivation, PKC and Rac1 activation all appear to contribute to sphingolipid-induced nutrient transporter loss. ARF6 inactivation may be downstream of sphingolipid-induced PP2A activation similar to other effects. Preliminary data suggests ARF6 inactivation may result from PP2A-mediated de-phosphorylation of the ARF6 guanine nucleotide exchange factor Grp1, which activates ARF6. Both endogenous sphingolipids and 893 inactivate ARF6. Interestingly, ARF6 activation is associated with tumor progression, grade, and metastasis in multiple cancer classes. We found that combining sphingolipid with ARF6 inhibitors synergistically and selectively kill cancer cells. These findings clarify the mechanism of sphingolipid-induced cell death and identify inactivation of ARF6 playing a role in nutrient transporter loss.