The metabolic profiles of cancer stem cells

Abstract

Cancer stem cells (CSCs) represent a minor but highly adaptable subpopulation within tumors that drives long-term growth, metastasis, and therapy resistance. Their ability to survive and regenerate under metabolic and therapeutic stress relies on a unique integration of energy flexibility, redox balance, and proteostatic programs. While bulk tumor cells typically favor aerobic glycolysis and high protein turnover, CSCs often exhibit elevated mitochondrial activity, fatty acid oxidation, and selective suppression of proteasome function. These metabolic features support quiescence, stress tolerance, and self-renewal. Beyond energy production, metabolic intermediates such as acetyl-CoA, succinate, and lactate serve as epigenetic cofactors, linking nutrient availability to chromatin remodeling and transcriptional plasticity. Reactive oxygen species and antioxidant responses further tune this balance, shaping the transition between glycolytic and oxidative CSC states. These intrinsic programs are continuously influenced by the tumor microenvironment, where hypoxia, cytokine-driven signaling, and metabolic coupling with stromal and immune cells modulate CSC metabolism and reinforce stemness. Despite rapid progress, major conceptual and methodological gaps still limit our understanding of CSC metabolism and this review highlights these unresolved issues and further outline key contextual factors—including tumor-intrinsic, microenvironmental, systemic, and metastatic cues—that shape CSC metabolism and help explain the divergent observations reported across studies. Understanding this network will be essential for designing combinatorial therapies that target CSC metabolism while accounting for their heterogeneity and plasticity.

1. Introduction

Tumors are comprised of a mixture of non-malignant (stromal) cells that form the tumor microenvironment, together with the malignant population of tumor cells. Within the malignant cell population, a subset of cells with stem-like properties is commonly referred to as cancer stem cells (CSCs). These cells are characterized by their capacity for self-renewal and their ability to generate progenitor cells that proliferate rapidly and give rise to the heterogeneous tumor cell populations. Two conceptual models have been proposed to explain the organization of CSCs within tumors: the hierarchical model, in which a defined CSC population drives tumor growth, and the plasticity model, which proposes that stemness represents a dynamic state rather than a strictly defined cellular hierarchy (Fig. 1) [1, 2]. Although early CSC models proposed a rigid hierarchical organization, accumulating evidence has revealed substantial plasticity, whereby non-CSC tumor cells can acquire stem-like properties in response to microenvironmental or intrinsic signals [3–5]. These stemness properties are lost during differentiation and are governed by pathways such as STAT3, NANOG, NOTCH, WNT, and HEDGEHOG, which are highly dysregulated in CSCs due to genetic and epigenetic changes [6]. Similar to normal tissue stem cells, CSCs often represent only a small fraction of the tumor, and are regulated by signals from their microenvironment (the CSC niche) [7, 8]. These cells exhibit relatively slow turnover rate compared to rapidly proliferating progenitor and tumor cells, which contributes to their resistance to therapies targeting fast-growing cells [9]. CSCs exhibit multiple mechanisms of therapy resistance, including efficient DNA damage response systems and elevated expression of drug-efflux transporters that reduce intracellular accumulation of chemotherapeutic agents [10]. Additional CSC-associated traits include altered signaling activity, epithelial–mesenchymal transition (EMT), increased anti-apoptotic and autophagy programs, and epigenetic modifications such as DNA methylation and histone remodeling [11, 12].

Identifying CSCs—markers and functional assays: CSCs were first identified in leukemia via their CD34+/CD38− phenotype in 1994 [13]. The first reports of CSCs in solid tumors were published in 2003, demonstrating that CD44+/CD24− breast cancer cells were able to self-renew and re-establish tumors with a range of partially differentiated cell types when transplanted into immunodeficient mice [14]. Subsequently, CSCs from solid tumors have been identified using a variety of assays [15]. These include the presence of specific cell surface proteins, such as Lgr5 [16], CD24/CD44 [17], and CD133 [18], high aldehyde dehydrogenase (ALDH) activity [19], the ability to efflux dyes to produce a side population in flow cytometry [20], the ability to form spheres in suspension culture [21], the production of colonies with holoclone morphology [22, 23], or by forming xenograft tumors in immunodeficient mice [24]. However, none of these is a universal marker of the CSC population in all tumor types, or in different tumors of the same type. For example, CD44+/CD24− and ALDH+ cells are widely used as CSC markers in breast cancer. However, luminal breast cancers are enriched in CD44−/CD24+ cells, whereas tumor cells in basal/mesenchymal breast cancers are enriched in the CD44+/CD24− phenotype, and the remaining basal/epithelial tumor types are often positive for both markers. ALDH is found mainly in HER2-overexpressing and basal breast cancers [25]. Moreover, markers identify a distinct sub-population of CSCs within a single tumor [11, 26], and ALDH+ “epithelial-like” CSCs may transition into “basal-like” CD44+ CSCs [27, 28], and vice versa. In addition, CD44/CD24 and ALDH are associated with different tumor characteristics: A high CD44/CD24 ratio is related to cell proliferation and tumorigenesis, while ALDH is an indicator of metastasis [29]. Within these CSC subtypes, breast adenocarcinoma CSCs may express the stem cell marker ΔNp63, which associates with the basal CSC type, while ALDH in these tumors is associated with a lack of ΔNp63 [30] and prostate adenocarcinomas may also show a ΔNp63/basal-like CSC subpopulation [31]. Importantly, ΔNp63 transcriptionally regulates CD44 [32], providing at least one mechanism for inducing specific CSC phenotypes within an individual tumor. However, the defining feature of CSCs is not marker expression but functional competence—the ability to self-renew and generate heterogeneous progeny. The plastic nature of these cells limits the reliability of marker-based identification alone.