Obesity increases the risk for colorectal cancer. In mice, a pro-obesity high-fat-diet (HFD) leads to an intestinal phenotype characterized by enhanced proliferation, numbers, function and tumor-initiating capacity of stem cells, the cell-of-origin for many intestinal cancers. This phenotype is driven by a lipid metabolism program facilitated by an intrinsic Peroxisome Proliferator-Activated Receptor/Fatty Acid Oxidation (PPAR/FAO) axis that senses and utilizes cellular lipids. However, the microbiome is a known regulator of lipid metabolism in the gut, but little is understood about how the gut commensals affect access to the lipids and alter stem cell function. Here, we use the long term HFD-fed mouse model to analyze the phenotypic changes in the intestinal stem cells (ISCs) after depletion of the gut microbiota. We find that the loss of the gut microbiome after four weeks of antibiotic treatment imposes significant changes in ISC function leading to reduced HFD ISC regenerative potential. These results indicate that the gut microbiome plays a crucial role in the lipid metabolic process which regulates and maintains the HFD ISC phenotype, and further suggests that the gut microbiome may augment the diet-induced tumor initiating capacity by altering the stem cell function.

Cooperative cellular phenotypes are universal across multicellular life. Division of labor, regulated proliferation, and controlled cell death are essential in the maintenance of a multicellular body. Breakdowns in these cooperative phenotypes are foundational in understanding the initiation and progression of neoplastic diseases, such as cancer. Cooperative cellular phenotypes are straightforward to characterize in extant species but the selective pressures that drove their emergence at the transition(s) to multicellularity have yet to be fully characterized. Here we seek to understand how a dynamic environment shaped the emergence of two mechanisms of regulated cell survival: apoptosis and senescence. We developed an agent-based model to test the time to extinction or stability in each of these phenotypes across three levels of stochastic environments.

Early detection of disease is essential for alleviating disease burden, increasing success rate and decreasing mortality rate especially for cancer. To improve disease diagnostics, many candidate biomarkers have been suggested using molecular biology or image analysis techniques over the past decade. The receiver operating characteristics (ROC) curve is a standard technique to evaluate a diagnostic accuracy of biomarkers, but it has some limitations especially for heterogeneous diseases. As an alternative of the ROC curve analysis, we suggest a jittered dot plot (JDP) and JDP-based evaluation measures, above mean difference (AMD) and averaged above mean difference (AAMD). We demonstrate how JDP and AMD or AAMD together better evaluate biomarkers than the standard ROC curve. We analyze real and heterogeneous basal-like breast cancer data.