Origins of Cancer Symposium Report

New concepts and enduring themes were the topic of conversation at the May 3, 2013, Origins of Cancer meeting. The 1-day symposium, titled “Hallmarks of Cancer,” showcased top scientific and translational discoveries on the origins and causes of cancer.

The symposium got off to a strong start with an opening presentation by Peter Vogt (The Scripps Institute). Using the history of retroviruses as a guide, Vogt revisited the key findings that established the undeniable role of tumor virology in oncogenesis. While this concept is now deeply integrated into the field of cancer biology, it was not initially well received. Early findings in an avian system by Peyton Rous were largely dismissed as irrelevant to mammalian biology. ¹ However, the subsequent discovery of the cellular origins of viral oncogenes having the ability to transform mammalian cells gradually changed opinions and led to the 1989 Nobel Prize for Harold Varmus and J. Michael Bishop. ² This work has since been leveraged to develop effective clinical therapies for melanoma, breast cancer, and chronic myelogenous leukemia by targeting oncogenic proteins.^3-5

Picking up the theme of genetic mutations or alterations as a driving force in cancer, Eric Holland (Memorial Sloan-Kettering Cancer Center) gave an elegant presentation showing how cluster analysis of transcriptional data could be leveraged to identify specific genomic aberrations that define molecular glioblastoma subclasses (proneural, mesenchymal, and classical). Interestingly, he showed that the proneural and classical variants of glioblastoma are commonly associated with aberrant growth factor signaling via amplification of receptors for platelet derived growth factor and epidermal growth factor, respectively. ⁶ Armed with the knowledge of how genetic changes drive glioblastoma, Holland hopes to determine both the order of mutational events as well as the vulnerabilities of a given tumor type in order to develop novel therapies for malignant brain cancer.

The question of which cells contribute to tumor growth was addressed in talks by Tannishtha Reya (University of California, San Diego) and Owen Witte (University of California, Los Angeles). Reya’s talk showed how she is pushing the field of stem cell and cancer stem cell research to new heights by elucidating the roles of Wnt and Hedgehog during asymmetric division of stem cells, and by revealing how cancer cells hijack these developmental pathways during tumorigenesis to create a heterogeneous population of tumor cells. One of the most fascinating aspects of Dr. Reya’s approach is the way she uses fluorescent molecular markers to visualize the way stem cells asymmetrically divide, orient their spindles, and become daughter cells with completely different transcriptomes and fates. Focusing on her work with the Musashi (Msi)–Numb signaling axis in chronic myelogenous leukemia (CML), ⁷ Reya showed that the Msi–Numb pathway can push CML cells into an undifferentiated state, leading to advanced CML and blast crisis. Inhibiting Msi could lead to enhanced differentiation and impaired leukemic growth. This work has interesting implications for the cancer stem cell model in CML, but also in solid tumors, in which Msi is similarly expressed.^8,9

Witte’s presentation centered on his work in prostate cancer and his efforts to decipher the combinatorial effects of multiple genetic alterations on tumor progression by altering distinct genetic pathways, such as p53, phosphatase and tensin homolog (PTEN), or AKT simultaneously with truncated erythroblast transformation-specific-related gene (ERG). ¹⁰ He also talked about the controversy surrounding the cell-of-origin for prostate cancer. ¹¹ While several research groups, Witte’s included, have shown that the basal prostate compartment contains stem cells that act as the cell-of-origin for prostate cancer,^12,13 others have shown that luminal cells are the cell-of-origin. ¹⁴ Witte presented data that challenge the practice of using mouse models to study stemness and tumorigenic properties of murine prostate cells and outlined new approaches he is using to identify the cell-of-origin for prostate cancer.

The topic of tumor metabolism was featured in several talks, notably those of Matt Vander Heiden (Massachusetts Institute of Technology) and Victor Velculescu (Johns Hopkins University). As first noted by Otto Warburg in the early 20th century, cancer cells leverage aerobic glycolysis over oxidative phosphorylation regardless of oxygen availability. ¹⁵ However, the reason tumors chose to use a less efficient means of energy production remains a mystery. One popular theory suggests that aerobic glycolysis is employed to allow more rapid ATP generation and sustain high levels of cell growth, but several key observations suggest otherwise. A marginal difference in ATP is seen between nonproliferating and proliferating cells while an accrual of macromolecules is observed in the proliferative state. Vander Heiden used labeled glucose and gas chromatography–mass spectroscopy to show proliferating tumor cells use glucose alternatives, like glutamine, to synthesize essential macromolecules under specific environmental situations (e.g., hypoxia). ¹⁶ Under hypoxic situations, cells are also glucose- and glutamine-limited. In this context, Vander Heiden hypothesized proteins may be consumed via macropinocytosis to execute macromolecule synthesis. In Ras-transformed cells, macropinocytosis was shown to be up-regulated in response to growth factor signaling and was required to rescue growth defects seen when glutamine was limited. ¹⁷

From the perspective of genomic alterations in cancer, Velculescu presented the discovery of alterations in the key metabolic enzyme isocitrate dehydrogenase (IDH1) in gliomas. He noted only 5% of primary glioblastoma harbor mutations in this gene compared with 80% of secondary glioblastoma, indicating a differential metabolism state in advanced cancer. These talks illustrated the importance of continued research to identify novel ways to treat cancers by exploiting differences in altered metabolism of a tumor cell. Velculescu believes that every tumor, possibly in the world, is different at the genetic level. Therefore, he is improving methods of detection and treatment of cancer using the cancer genome to promote a customized approach over a “one size fits all” approach.^{18, 19} This requires a comprehensive analysis of cancer genomes, which includes identifying novel altered pathways such as IDH.

As a cancer alters its metabolism to adapt to its environment, so too does the environment alter its composition to adapt to a tumor. Jeffrey Pollard (University of Edinburgh) provided an excellent discussion on the contribution of the stromal microenvironment during cancer progression and metastasis. Pollard used intravital multi-photon imaging ²⁰ to graphically demonstrate that macrophages make up large portions of tumors, secrete growth factors that regulate angiogenesis, stimulate invasion, and define a microenvironment that facilitates intravasation.^21-24 His results support the emerging paradigm that a permissive, even contributive, microenvironment is necessary for malignant progression. From a clinical perspective, macrophages, or their unique signaling pathways, represent new therapeutic targets that may be efficacious in reducing cancer mortality.

Yibin Kang (Princeton University) expanded the role of the tumor microenvironment in cancer progression by outlining how the intricate interactions between tumor cells and their stromal microenvironment lead to metastasis. Kang described the role of transcription factor Elf5, a protein that regulates luminal cell fate with known ties to cancer progression in breast cancer metastasis. ²⁵ Using an in vivo murine mammary model, he showed that loss of Elf5 drives an epithelial to mesenchymal transition through loss of adherens junctions and mislocalized vimentin, as well as an increase in overall migration. Animals heterozygous for Elf5 have increased lung metastases and Snail2 expression compared to wild-type Elf5 controls. ²⁶ In contrast, overexpression of Elf5 repressed migration and lung metastasis. Within estrogen receptor-negative breast cancer patient samples, high Elf5 expression reflects a better prognosis. Conversely, high Snail2 levels are associated with a worse prognosis. ²⁶ This work illustrates the functional insights that basic cancer research can provide for potential clinical applications.

The Origins of Cancer symposium highlighted paradigm-shifting work that is being done to challenge the way researchers think about cancer etiology and treatment. Technological advances and multidisciplinary research efforts are driving the convergence of themes from different areas of study into a cohesive understanding of cancer biology emphasizing tumor metabolism and microenvironment in the progression of disease. Our challenge in the coming years is to match this understanding with equally impressive clinical successes.