Bioengineered 3D collagen-alginate interpenetrating network (IPN) to elucidate phenotypic switching of cancer associated fibroblasts

• Main Author: Cao, Huan
• Other Authors: Dalton Tay Chor Yong
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Bioengineering; Engineering::Materials::Biomaterials
• Online Access: https://hdl.handle.net/10356/148935
• Institution: Nanyang Technological University

Three-dimensional (3D) biomaterials with physiologically relevant and experimentally tractable biomechanical features are important platforms to advance the understanding on the influence of tissue mechanics in cancer progression. Cancer is the second leading cause of death worldwide, which conservatively estimated to have 13.1 million cases of cancer-related death by the year 2030. It has been shown that cancer-associated fibroblasts (CAF), the most abundant cell type within the tumor stroma, plays a significant role in promoting cancer progression via multiple mechanisms such as paracrine crosstalk, extracellular matrix (ECM) remodeling, etc. However, recent studies have also shown that certain subtypes of CAF may also possess anti-tumorigenic functions. Factors underpinning the existence of the different CAF subtypes with functionally diametrically different (anti and/or pro) tumorigenic properties remain elusive to date. Delineating the spatiotemporal phenotypic switching of CAF phenotype is important as it may uncover novel and specific targeted depletion of certain CAF subsets to effectively impede tumorigenesis. However, current conventional in vitro biological toolkits are inadequate to recapitulate the biochemical and biophysical tumor stroma microenvironment that is necessary to systematically investigate CAF heterogeneity. In this thesis, an interpenetrating polymer network (IPN) hydrogel system composed of rat tail type I collagen and alginate (CoAl-IPN) was used as an in vitro tumor stroma surrogate. The distinct crosslinking chemistries of each network offer precise spatiotemporal control over the biochemical and biomechanical properties of the hydrogel. Hydrogels with Young’s modulus (E) of ~100 Pa and ~1000 Pa were utilized as “soft” and “hard” cancerous stroma respectively. The rheological, mechanical, mass transport and microarchitectural properties of the hydrogel were systematically characterized. Using this platform, the effects of 3-dimensional (3D) matrix stiffness and network density on CAF morphology, adhesion, phenotype and paracrine activity were examined. In human breast CAF (b-CAF), it was observed that a soft 3D microenvironment promoted cell adhesion and spreading. In contrast, in a 3D stiff matrix (E ~ 1000 Pa) which is characterized by a highly cross-linked and dense network, spreading of b-CAF was severely constrained. Moreover, nuclear translocation of the YAP/TAZ was also inhibited in b-CAF grown in a stiff matrix, indicative that the intracellular tension is in a “relaxed” state. Importantly, using a 3D MDA-MB-231 tumoroid invasive model, the pro-tumorigenic paracrine activity of b-CAF in the stiff matrix group was significantly muted. Therefore, using the CoAl-IPN hydrogel platform it was shown that the biophysical parameters of the tumor stroma could either promote or restrain the pro-tumorigenic activity of b-CAF. Interestingly, the matrix stiffness dependent event was recapitulated in human colon-CAF (c-CAF), suggesting that the observed phenomenon was not specific to CAF of a particular tissue origin. Further mechanistic insights were sought via whole transcriptome profiling of the encapsulated CAF using RNA Seq. A total of 1,623 differentially expressed genes (DEGs) (FDR < 0.05) were identified between the c-CAF in the hard and soft matrix group. Further gene enrichment analysis of the DEGs against the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways in combination of reverse transcription polymerase chain reaction (RT-PCR) and immunoassays revealed that a soft 3D matrix favors the adoption of a myofibroblastic (m-state) while a stiff 3D matrix favors an inflammatory (i-state) c-CAF phenotype. Specifically, m-state CAF displayed a higher expression level of intracellular α-SMA expression and downregulated IL-6 expression. On the other hand, i-state CAF has a higher secretory of inflammatory phenotype (i.e., increased IL-6) together with downregulated intracellular α-SMA expression. Importantly, phenotypic switching between both cellular states is reversible. Hub gene analysis also unveil reactive oxygen species (ROS)-hypoxia inducible factor 1 alpha (HIF1-α) as a novel mechano-redox transduction axis that can be conditionally switched “on” and “off” between i- and m- state CAF, respectively. The transition from the m- to i-state was accompanied by a significantly attenuated capacity to induce epithelial-mesenchymal transition (EMT) in human DLD-1 colorectal adenocarcinoma by the c-CAF-derived conditioned medium (CM). Consistently, upregulated cytokine/ chemokines genes in the i-state CAF were associated with good overall colorectal cancer patient survival based on the gene expression profiling interactive analysis (GEPIA) platform. Overall, these findings showed that 3D network density and spatial cellular confinements can profoundly determine CAF states, paracrine signaling, and metastatic potential. Deciphering the interplay between ECM mechanics and CAF states (deleterious vs. beneficial) may reveal critical vulnerabilities of cancer and open-up new avenues of effective combinatorial anti-cancer therapeutic strategies.