Poster Competition Winners 2019

Jefte Drijvers 
Sharpe Lab 

Diet-Induced Obesity Causes Metabolic Reprogramming and Suppresses Immunity in the Tumor Microenvironment 

Obesity is a well-known risk factor for many cancers. Various systemic alterations, including nutrient availability and signaling changes, are associated with obesity. However, how these changes impact the anti-cancer immune response is not yet clear. Tumors maintain an immunosuppressive, nutrientpoor microenvironment, and it is unknown whether the systemic metabolic changes associated with obesity are reflected in the tumor microenvironment (TME) locally. We tested the hypothesis that obesity impacts anti-tumor immune function in the TME using the murine obesity model of feeding a high-fat diet (HFD). Our studies show that HFD accelerates tumor growth by inhibiting the anticancer immune response. Moreover, HFD induces different metabolic adaptations in tumor cells and intratumoral immune cells, resulting in an altered nutrient composition in the TME. Genetic manipulations that alter metabolic reprogramming in tumor cells normalize the metabolic milieu in the TME and reduce tumor growth in an immune system-dependent manner. These findings demonstrate how cell-type specific rewiring of metabolism in the TME in response to systemic metabolic changes resulting from diet-induced obesity inhibits the anti-tumor immune response locally. Analysis of publicly available transcriptional data of human cancers from The Human Genome Atlas suggest that similar metabolic reprogramming events correlate with obesity and decreased anti-tumor immune function in human cancers as well. Thus, our studies reveal a novel mechanism linking obesity to cancer through decreased anti-tumor immune function and may inform the development of cancer therapies targeting cancer metabolism and anti-tumor immunity. 

 

Manav Gupta 
Kim Lab 

The Mammalian SWI/SNF Complex Regulates Origin Firing in Lung Cancer 

The SMARCA4 gene encodes the ATP-dependent helicase component of the SWI/SNF complex, BRG1, is involved in chromatin modulation and either mutated or lost in up to 20% of human nonsmall cell lung cancers. To investigate how loss-of-function mutations in SMARCA4 contributes to lung tumorigenesis, we generated murine and human BRG1 knockout cell lines from tumor cells derived from a KrasG12D/+; p53Δ/Δ (KP) mouse model of lung adenocarcinoma, and KRAS/p53 mutant human lung cancer cell lines H460, H2009, and Calu6. RNA-sequencing of murine Brg1 null cells and human SMARCA4-mutant patient data taken from the TCGA cancer datasets revealed the upregulation of the ATR-mediated response to replication stress and activation of the pre-replicative complex as the top cancer pathways in these cancers. We hypothesized that loss of BRG1 contributed to increased replication stress associated DNA damage and genome instability. BRG1-deficient cells had significant more gamma-H2Ax foci and RPA-bound singlestranded DNA compared to isogenic controls. Western blot analysis confirmed activation of the ATR pathway through increased phospho-CHK1 activity in BRG1-deficient cells. Mechanistically, we observed that loss of BRG1 expression leads to increased number of fired origins of replication as measured by DNA fiber assays. The correlation between replication stress and DNA damage was assessed using comet analysis and we observed increased olive tail moments in BRG1-deficient cells, indicative of more DNA damage and genome instability. We then treated our cells with inhibitors that target key DNA damage response kinases, ATM and ATR. While ATM inhibition resulted in no observable change between BRG1 wildtype and mutant cells, BRG1-deficient cells were more sensitive (3-5 fold) to ATR inhibition compared to isogenic wildtype cells. ATR inhibition was also able to significantly increase the total amount of DNA damage in BRG1- deficient cells, suggesting that loss of BRG1 leads to dependence on the ATR pathway to prevent further genome instability. We re-expressed human BRG1in murine/human Brg1/BRG1 knockout cells and observed a reversal in response to ATR inhibition. We also found that combinatorial treatment with replication stress inducing reagents such as topoisomerase I inhibitor irinotecan or hydroxyurea further sensitized BRG1-deficient cells to ATR inhibition in some models. To address the role of the SWI/ SNF complex in origin firing and DNA replication, we examined levels of the early DNA origin licensing proteins and found that loss of BRG1 expression strongly correlated with increased CDC6 presence across all our models of BRG1 loss. To further study the role of Brg1 as a tumor suppressor gene in the lung, we compared KP mice versus KP mice harboring floxed Brg1 (KPB) alleles and found that KPB mice had a significantly higher number of tumor lesions and highergrade tumors after 13-15 weeks of tumor induction. Interestingly, there was a significant correlation in loss of Brg1 and presence of key immune evasion ligand Pd-l1 by immunohistochemistry in Brg1 null tumors. Subcutaneous injections of Brg1 null murine isogenic lines into flanks of immunocompetent mice further showed the increase in Pd-l1 expression only in tumors derived from Brg1 null cells. Taken together, our data suggests that BRG1 or the SW/SNF complex may have a role in regulating DNA replication in lung cancer cells, and it does so by mediating CDC6 expression and controlling origin firing.

 

Adrija Navarro
Toker Lab 

Investigating the Role of ALG3 in the Regulation of N-Glycosylation by the PI3K/AKT Signaling Pathway 

The PI3K/AKT signaling pathway, which is frequently dysregulated in cancer, controls key cellular processes such as survival, proliferation, metabolism, and growth. Protein glycosylation, the process by which carbohydrates are added to amino acids, is essential for proper protein folding and is deregulated in cancer. High proliferation rates in cancer require amplified protein folding. The glycosyltransferase ALG3 catalyzes the addition of a mannose to a glycan precursor once it is flipped into the endoplasmic reticulum lumen during glycan production. ALG3 is required for proper glycan formation and is implicated as a putative AKT substrate. ALG3 resides proximal to PIK3CA in the 3q26 amplicon. Consequently, PIK3CA and ALG3 are co-amplified in 89%, 28% and 76% of lung (SCC), breast and ovarian carcinomas, respectively. Notably, we find that in both lung and breast cancer cells, ALG3 is also phosphorylated downstream of PI3K. This represents, to our knowledge, the first identified link between PI3K oncogene signaling and protein glycosylation in the context of cancer. I hypothesize that ALG3 plays a role in the regulation of protein N-glycosylation by PI3K/AKT signaling, and that aberrant PI3K/ AKT signaling alters glycosylation, leading to functional consequences in cancer. Specifically, I postulate that cells that harbor PIK3CA amplification and ALG3 up-regulation increase glycosylation and protein folding rates, allowing cells to cope with increased protein translation in response to hyperactive PI3K/AKT signaling. This project will advance our understanding of the regulation of glycosylation metabolism by PI3K/AKT signaling and its role in cancer progression; future studies may determine the extent to which combination therapies targeting the PI3K/AKT pathway and proteinglycosylation are effective. 

 

Alexandra Pourzia
Letai Lab 

Cancer Cell Defects in Apoptosis Attenuate Killing by CAR T Cells 

CAR T therapy is now an FDA approved treatment for several hematologic malignancies, yet not all patients respond to this treatment. While some resistance mechanisms have been identified, the possibility of cell death pathways modifying response to CAR T therapy remains unexplored. To assess whether cell death pathways in target cancer cells could impact response to CAR T therapy, we utilized a HeLa in vitro model system. HeLa cells with intact (HeLa-19) and deficient Bak/Bax (HeLa-DKO-19) expressing CD19 were co-cultured with CD19 CAR T cells. We observed that Bak/Bax deficiency, which blocks the intrinsic pathway of apoptosis, conferred resistance to CAR T killing. However, this resistance could be overcome at high E:T ratios. To confirm that the intrinsic pathway of apoptosis contributes to CAR T killing, we forced the expression of Bcl-2 and Bcl-XL in HeLa-19 cells, and observed that both of these anti-apoptotic proteins conferred protection from CAR-T killing in a similar manner to Bak/Bax knockout. Additionally, we wanted to assess whether caspases may be required for CAR T killing, given the role of intrinsic apoptosis in our model system. The caspase inhibitor Z-VAD-FMK protected both HeLa-19 and HeLa-DKO-19 cells from CD19 CAR T effector cells in our in vitro model system. Lastly, we wanted to ascertain the precise mechanism by which CAR T cells eliminate target cancer cells. CAR T cells were co-cultured with HeLa-19 and HeLa-DKO-19 targets in the presence of blocking antibodies against Fas ligand and TRAIL, along with 3,4-dichloroisocoumarin, a granzyme inhibitor. We observed that granzyme inhibition, but not death ligand blocking antibodies, provided protection from CAR T cells. Additionally, a soluble factor contributed to CAR T killing, as demonstrated by conditioned media experiments in which target cancer cells were exposed to filtered supernatant from CAR T coculture experiments. In conclusion, intrinsic apoptosis allows for efficient elimination of target cancer cells by CD19 CAR T cells. This process also requires downstream caspases, and seems to be mediated in part by granzymes. This work implies that agents that promote tumor cell intrinsic apoptosis may be candidates for combination treatment with CAR T therapy; and suggests that tumor cells that are resistant to intrinsic or downstream apoptosis may resist CAR T therapy. Future work will explore whether the intrinsic apoptotic pathway also modifies response to CAR T cells in a mouse model, and identify soluble factors that contribute to CAR T killing of target cells.