Dynamic Interconversion and Functional Studies of Heterogeneous CD90 Subpopulations in A549 Lung Adenocarcinoma


Student thesis: Doctoral Thesis

View graph of relations


Related Research Unit(s)


Awarding Institution
Award date6 Jun 2018


Tumors comprise of heterogeneous cell types such as immune cells, endothelial cells, stromal cells and cancer cells. Functional diversity of various cell types and populations is crucial for tumor growth and maintenance. In conventional chemotherapy, while the normal cancer cells are eliminated by chemotherapeutic drugs, the remaining stem-like cancer population can give rise to various heterogeneous cancer cells, leading to tumor relapse. These cancer cells co-exist in different phenotypic states and are maintained in various phenotypic equilibria within the population.

Early therapeutic strategies focused on eliminating the stem-like cancer cells, otherwise known as cancer stem cells (CSCs), to control tumor growth and relapse. However, recent studies have revealed the potential switching between CSCs and non-CSCs, as well as other cancer subpopulations. This increased the therapeutic challenges. However, the interconversion pattern is not yet clear. Despite some studies showing stochastic transitions between phenotypic equilibria of cancer subpopulations, it is controversial whether these transitions are always stochastic.

In this thesis, the aim is to identify different populations in lung cancer cell line and to study the dynamic switching pattern of these populations, in correlation with its function and significance in lung cancer. The content of each chapter is summarized below:
Chapter 1: This part provides a brief introduction to cancer cell heterogeneity and its significance. The rare subpopulation of CSCs in tumor maintenance will be described as an example and its association with the hierarchical model will be introduced. Next, the stochastic model involving switching in cancer subpopulation is discussed.

Chapter 2: Biomarker screening is first performed and Thy-1 (CD90) is selected as a model of interconversion studies. After identification of two enriched CD90 subpopulations in A549 adenocarcinoma cells by fluorescence-activated cell sorting (FACS), it is showed that the subpopulations are capable of interconverting over time and they form a new non-stochastic equilibrium by interconversion. Furthermore, a mathematic model is built to analyze the conversion dynamics of CD90 positive and negative population. Ordinary differential equations (ODEs) are derived to simulate the switching process and various rate constants are obtained from the model by fitting the experimental data to the ODEs. Finally, a probability model is used to illustrate the probability of phenotypic switching.

Chapter 3: Various assays are carried out to characterize the sorted positive and negative populations. With better resolution of single-cell sorting, the presence of committed cells is found to demonstrate very limited or no transition tendency. The functions of CD90 are characterized by both sorted transition and committed cells in different aspects. CD90+ cells are associated with high EMT/Wnt, migration, and proliferation capabilities, while CD90- cells are more tumorigenic and have higher colony formation ability. CD90 knockdown and knockout are performed for functional studies.

Chapter 4: RNA sequencing is performed to reveal additional significant genes and pathways for CD90. For example, ECM-receptor interaction pathway is dominant in CD90+ cells, while complement and coagulation cascades pathway is significant for CD90- ones. Based on all gene expression results, further gene set analysis is performed to reveal potential regulatory genes pathways. HNF6, TP53, JUN, EGR1, and ATF2 are found to regulate various significantly regulated genes of those pathways.

Chapter 5: A conclusion of the major findings of this study with emphasis on future works which could further improve the understanding on the control of interconversion. For future studies, double sortings involving additional markers, CD24 and EpCAM, are also investigated for more precise interconversion trends.

These findings may provide clues on the factors that affect the cell transition outcomes and that potentially change the tumor subpopulation composition. These clues improve the understanding of the control of phenotypic conversion for potential cancer therapy.