Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Biology (Cell Biology, Microbiology, Molecular Biology)

Major Professor

Robert J. Gillies, Ph.D.

Committee Member

John Cleveland, Ph.D.

Committee Member

Robert Gatenby, M.D.

Committee Member

Conor Lynch, Ph.D.


3D, acid, carbonic anhydrase IX, glycolysis, pH, protons


Cancer is a complex and heterogeneous disease. Not only is there considerable variability between different cancer types, but there is enormous variability between and within patients who have the same type of cancer. Within tumors, there are multiple cell types, including cancer cells, stromal cells, and immune cells. The tumor microenvironment often induces the healthy cells to become pro-tumorigenic. Cell metabolism is exquisitely sensitive to changes in the tumor microenvironment and can be measured to infer the aggressiveness of cancer and predict response to therapy. In this dissertation, we aim to understand how the microenvironment, specifically low pH, affects the phenotype, metabolism, and function of cell types within tumors and ultimately, how this relates to metastasis and therapeutic outcome.

Chapter 2: Unfortunately, measuring metabolism is challenging as metabolism is dynamic and can change depending on the types of models we use. Current studies utilize 2D models of cancer, i.e., cancer cell lines grown in flat, plastic flasks or dishes. 2D cultures are exposed to higher concentrations of nutrients, drugs, and have a reduced cell surface area to volume due to adhering to the plastic; this also causes altered expression of adhesion proteins. The 2D models do not accurately reflect the tumor in vivo, and this can result in discrepancies between in vitro and in vivo results. Growing cancer cell lines in 3D better recapitulates what occurs in vivo. We developed an in vitro live-cell tooling and methodology to metabolically profile 3D cell cultures and directly compare them to 2D cell cultures. Using this, we observed differences in the basal metabolism of 2D and 3D cell cultures in response to metabolic inhibitors and chemotherapeutics. We further expanded this methodology to profile 3D microtissues generated from mouse organs and tumors. The metabolic profiles of microtissues derived from normal organs (heart, kidney) were consistent when comparing microtissues derived from the same organ. Treatment of heart and kidney microtissues with cardio- or nephrotoxins had early and marked effects on tissue metabolism.

In contrast, microtissues derived from different regions of the same tumors exhibited significant metabolic heterogeneity, which correlated to histology. Hence, metabolic profiling of complex microtissues is necessary to understand the effects of morphology and structure on metabolism. This method has the potential to be used as a reproducible, early and sensitive measure of drug toxicity and a potential drug screening methodology. Moreover, this methodology provides an important stepping stone to improve experimental design between in vitro 2D studies and in vivo animal studies.

Chapter 3: Development of novel methodologies and tooling development can drive progress in scientific research. Our method highlights the metabolic heterogeneity within the tumor which is often missed by other means. One under-explored component of tumor heterogeneity is the substantial distinction between the edge and the core. Ongoing intratumoral evolution is apparent in molecular variations among cancer cells from different regions of the same tumor. However, genetic data alone provide little insight into environmental selection forces and cellular phenotypic adaptations that govern the underlying Darwinian dynamics. We have observed in three spontaneous murine cancers (prostate cancers in TRAMP and PTEN mice, pancreatic cancer in KPC mice), that there are two subpopulations with distinct niche-construction adaptive strategies that remained stable in culture. One population is what we call a “pioneer” phenotype, which is found at the invasive edge of tumors. Pioneers are invasive cells that produce an acidic environment through upregulated aerobic glycolysis and upregulation of proton exporting machinery, such as carbonic anhydrase -9 (CA-IX). The other metabolic population observed is what we call an “engineer” phenotype. Engineers are found at the core of tumors; they are non-invasive and have upregulated angiogenesis. These engineer cells are metabolically near-normal with an increased reliance on oxidative metabolism. Intratumoral evolution occurs generating cells with different fitness advantages in different regions of the tumors. These cells may purposefully migrate to their preferred area i.e. edge vs core, or end up there serendipitously, and once there, their phenotype may change with exposure to changing microenvironmental conditions or may be fixed. In our cases, these cells were cultured for many generations and their metabolic profiles were steady, indicating that the selected phenotypes were fixed.

Darwinian interactions of these subpopulations were investigated in TRAMP prostate cancers. Computer simulations predicted and experiments confirmed that invasive, acid-producing, proliferative (C2) “pioneer” cells maintained a fitness advantage over non-invasive, angiogenic, quiescent (C3) “engineer” cells. This is most likely managed by promoting invasion and hampering the immune response. Immunohistochemical analysis of untreated tumors confirmed that C2 cells invariably outgrew and were more abundant than C3 cells in TRAMP prostate tumors. However, the C2 adaptive strategy phenotype incurs a high cost due to inefficient energy production (i.e., aerobic glycolysis). Mathematical model simulations predicted that small perturbations of the microenvironmental extracellular pH (pHe) could invert the cost/benefit ratio of the C2 strategy and select for C3 cells. Altering pH using buffer therapy, NaHCO3, increased pH in a TRAMP mouse model, and promoted the growth of C3 engineer cells over C2, enabling the tumor to remain low grade and reducing metastasis when treating established tumors. Mathematical models of the intratumoral Darwinian interactions of environmental selection forces and cancer cell adaptive strategies indicated that the tumor trajectory was steered into a less invasive pathway through the application of small but selective biological forces, such as altering pH.

Pioneer cells found at the invasive edges of tumors, tend to have a high glycolytic metabolism and generate an acidic microenvironment. These cells are necessary for local invasion and cancer progression. Therefore, we were interested to see the impact of this acidic microenvironment on other components of the tumor. We studied how glycolytic metabolism and low pH affects three components of the tumor: the stroma, the immune compartment, and the cancer cells themselves.

Glycolytic metabolism of cancer co-opts vital nutrients needed for normal cell growth and this forces the surrounding cell types to use other metabolites for fuel. One area where this occurs is in bone metastases. We have identified that the metabolic interplay between bone stroma and cancer cells enhances metastases. In our study, we demonstrated that highly glycolytic triple-negative MDA-MB-231 cancer cells, compared to non-metastatic MCF7 cells, release more lactate and form osteolytic bone metastasis in vivo. In vitro, lactate generated by cancer cells was shown to be consumed by osteoclasts as a fuel for oxidative metabolism, ultimately enhancing Type I collagen resorption in the bone. The transport of lactate into osteoclasts was mediated by MCT1, as shown by the significantly upregulated expression during osteoclast differentiation, and using an MCT-1 inhibitor, 7-(N-benzyl-N-methylamino)-2-oxo-2H-chromene-3-carboxylic acid, which impaired Type I collagen resorption in the osteoclasts. Together, these data indicate that lactate released by glycolytic breast carcinoma cells in the bone microenvironment promotes the formation of osteolytic lesions and provide the rationale for using MCT1 inhibitors as a novel therapeutic approach in patients with bone metastases.

Additionally, one common symptom of bone metastasis in patients is pain (cancer-induced bone pain -CIBP). We have shown bone metastatic cancer cells are highly glycolytic, which generates high lactate levels and an acidic microenvironment. We postulated the acid in the tumor microenvironment may be a cause of CIBP which warranted further investigation. We found breast carcinoma cells that prefer bone as a metastatic site have very high extracellular proton efflux and express pumps/ion transporters associated with acid-base balance (MCT4, CA9, and V-ATPase). Acidosis can stimulate and sensitize the nociceptors in bone and result in hyperalgesia. Exposing cancer-associated fibroblasts, mesenchymal stem cells, and osteoblasts to acidic pH promoted the expression of inflammatory and nociceptive mediators (NGF, BDNF, IL6, IL8, IL1b, and CCL5). Further, the impairment of intratumoral acidification via V-ATPase targeting in bone metastases models significantly reduced CIBP. Current treatments for patients with CIBP are often ineffective, but our in vivo results correlate with patients clinically providing a potential new treatment avenue. Pain in patients, as measured by a questionnaire, correlated with higher levels of inflammatory and nociceptive mediators’ production, e.g. IL6 and IL8 by the stromal cells, suggesting tumor microenvironments generate pro-tumor stroma. Notably, a clinical trial (NCT01350583) treated 9 patients in a phase I trial of Na-bicarbonate (to raise pH) as an adjuvant to reduce bone pain. While the study failed to escalate, there was a significant reduction in patient reported CIBP. In summary, intratumoral acidification in the bone marrow may activate the tumor-associated stroma promoting CIBP and this finding offers a new target for palliative treatments in advanced cancer.

Chapter 4: Immune cells are another component of the tumor that are exposed to the stressful microenvironmental conditions generated by tumor cells. In our studies, we have focused on effector T cells of the immune system; i.e. CD8+ lymphocytes that have the potential to kill tumor cells through release of granzyme B. CD8+ T cells are regulated under normal conditions through expression of checkpoint inhibitor proteins that reduce effective lifetime or prevent T cell activity. Immunotherapy, such as checkpoint blockade inhibitors, increases T cell persistence, and has induced durable responses in many patients. However, many tumor types and many patients have shown minimal or no response, and it is unknown why. As we have discussed, aggressive tumors are often acidic, and we have seen effects of low pH on both cancer cells and the stroma. However, the impact of acid on T cells is unclear. Therefore, we aimed to understand the effect of acidity on T cell function. We found that activated murine T cells primarily rely on glycolytic metabolism, have reduced effector function and reduced glycolysis in low pH conditions. Specifically, acid inhibits glycolysis by preventing lactate transport out of the cell, which increases intracellular lactate and reduces intracellular pH resulting in downregulation of glycolysis. Altering the pH of solid tumors using buffer therapy improved the efficacy of immunotherapies. These findings have shown that acidity in solid tumors inhibits T cell effector function, which is vital for response to immunotherapy. By reducing intratumoral acidosis, we may be able to improve immunotherapy response in tumors and patients which have so far seen little benefit.

Chapter 5: In our studies, we assume low pH is generated by the highly glycolytic metabolism of cancer cells. However, expression of proteins involved in cellular acid export, such as carbonic anhydrase IX, are often concurrently expressed in cells which exhibit high levels of glycolysis such as in pioneer cells at the invasive edges. What if acid export drove the uptake and metabolism of glucose? Prior work strongly correlates tumor acidity with increased invasion and metastasis, and that neutralization of tumor acidity can prevent the formation of metastases in many systems. However, correlation is not causation. We tested the hypothesis that acidosis can systemically cause metastasis by induction of acid production through the expression of plasma membrane proton exporters. We designed two “gain-of-function” systems to generate acid production independent of metabolic acid production. We engineered acid-producing cells by overexpressing a yeast proton pump, H-ATPase (PMA1), or a proton-exporting exofacial human carbonic anhydrase 9 (CA-IX). In vitro studies showed expression of either proton exporter into non-metastatic human cancer cell lines enhanced aerobic glycolysis, migration, and invasion. In vivo, the acid-producing cells formed higher-grade tumors, which were associated with increased spontaneous and experimental metastases. Neutralizing acidity with buffer therapy reduced metastasis. Therefore, increased rates of H+ export can drive increased aerobic glycolysis (the Warburg Effect) and convert non-metastatic cells into those that form extensive metastasis. Thus, we propose that variables associated with metastasis may be activated and enhanced by increased acid production, and thus acidosis is systemically causal of metastasis.

Overall, we have been able to understand the impact of low pH on multiple different cell types in tumors, how this impacts their function and the resultant implications for cancer aggressiveness and response to therapy.