Cancer-Associated Fibroblast Subpopulations With Diverse and Dynamic Roles in the Tumor Microenvironment

Abstract

Close interactions between cancer cells and cancer-associated fibroblasts (CAF) have repeatedly been reported to support tumor progression. Yet, targeting CAFs has so far failed to show a real benefit in cancer treatment, as preclinical studies have shown that such a strategy can enhance tumor growth. Accordingly, recent paradigm-shifting data suggest that certain CAF subpopulations could also show tumor-inhibitory capabilities. The present review aims to provide an in-depth description of the cellular heterogeneity of the CAF compartment in tumors. Through combining information from different cancer types, here we define 4 main CAF subpopulations that might cohabitate in any tumor microenvironment (TME). In addition, a model for the evolution of CAFs during tumor development is introduced. Moreover, the presence of tumor-inhibitory CAFs in the TME as well as their molecular characteristics are extensively discussed. Finally, the potential cellular origins of these distinct CAF subpopulations are reviewed. To our knowledge, this is the first attempt at establishing a broad but comprehensive classification of CAF subpopulations. Altogether, the present manuscript aims to provide the latest developments and innovative insights that could help refine future therapeutic targeting of CAFs for cancer treatment.

Figure 1.

Overview of the TME. In the bulk tumor, cancer cells interact with a complex meshwork of ECM proteins, including collagen and fibronectin. Cancer cells also interact with a milieu of stromal cells, including endothelial cells, pericytes, immune cells, and CAFs. As the tumor develops, cancer cells communicate with the TME through cytokines, growth factors, and extracellular vesicles.

Figure 2.

Potential cellular origins of CAFs. Multiple stromal cell types have been proposed as potential origins of CAFs, including resident-tissue normal fibroblasts, adipocytes, pericytes, MSCs, hematopoietic stem cells, epithelial cells, and endothelial cells. How different subpopulations of CAFs are formed in the growing tumor is yet to be fully understood but two main models of CAF transformation and lineage are recognized in the literature. The first of these purports that CAF subpopulations are derived from one type of normal stromal cell that undergoes CAF reprogramming followed by further differentiation that would lead to different CAF varieties and subpopulations (Model A). The second model proposes that CAF subpopulations are derived from different precursor stromal cells in the host tissue (Model B). In both models, CAFs subpopulations can further differentiate into more specialized subpopulations.

Figure 3.

Proposed subpopulations of CAFs. The figure displays 4 broad subpopulations of CAFs in the TME as extracted from the literature: Immune, Desmoplastic, Contractile, and Aggressive. The immune and desmoplastic populations tend to be tumor inhibitory while the contractile and aggressive are more tumor supportive. The ‘immune’ subpopulation is associated with C3, ENG, IL6, PDGF-Rα, and PDPN expression. The ‘desmoplastic’ subpopulation is specifically characterized by a high expression of DCN, LUM, and POSTN as well as ECM components, such as elastin and collagen. The ‘contractile’ and ‘aggressive’ subpopulations are respectively defined by a high expression of factors involved in contraction of actin stress fibers or cell cycle (‘contractile’) and high expression of markers associated with EMT, such as vimentin or VEGF-A, or the TGF-β pathways (‘aggressive’). Both show the highest expression of α-SMA and are linked to poor patient survival/outcome. The studies used to highlight the existence of these broad CAF types are listed on top.

Figure 4.

Model for the evolution of CAFs with tumor progression. CAFs are heterogenous and dynamic in nature. Tumor progression could be associated with a decrease of the tumor-inhibitory CAF/tumor-supporting CAF ratio, with highest levels seen at earlier tumor stages as a reaction to tumor emergence followed by a progressive conversion towards a tumor-supporting CAF that overtakes the CAF compartment.