Overwhelming evidence indicates that virtually all treatment-naive tumors contain a subpopulation of cancer cells that possess some stem cell traits and properties and are operationally defined as cancer cell stem cells (CSCs). CSCs manifest inherent heterogeneity in that they may exist in an epithelial and proliferative state or a mesenchymal non-proliferative and invasive state. Spontaneous tumor progression, therapeutic treatments, and (epi)genetic mutations may also induce plasticity in non-CSCs and reprogram them into stem-like cancer cells. Intrinsic cancer cell heterogeneity and induced cancer cell plasticity, constantly and dynamically, generate a pool of CSC subpopulations with varying levels of epigenomic stability and stemness. Despite the dynamic and transient nature of CSCs, they play fundamental roles in mediating therapy resistance and tumor relapse. It is now clear that the stemness of CSCs is coordinately regulated by genetic factors and epigenetic mechanisms. Here, in this perspective, we first provide a brief updated overview of CSCs. We then focus on microRNA-34a (miR-34a), a tumor-suppressive microRNA (miRNA) devoid in many CSCs and advanced tumors. Being a member of the miR-34 family, miR-34a was identified as a p53 target in 2007. It is a bona fide tumor suppressor, and its expression is dysregulated and downregulated in various human cancers. By targeting stemness factors such as NOTCH, MYC, BCL-2, and CD44, miR-34a epigenetically and negatively regulates the functional properties of CSCs. We shall briefly discuss potential reasons behind the failure of the first-in-class clinical trial of MRX34, a liposomal miR-34a mimic. Finally, we offer several clinical settings where miR-34a can potentially be deployed to therapeutically target CSCs and advanced, therapy-resistant, and p53-mutant tumors in order to overcome therapy resistance and curb tumor relapse.

Prostate cancer (PCa) lethality is driven by its progression to a metastatic castration-resistant state, yet the signaling mechanisms underlying metastatic spread remain unknown. Here we show that STAT3 converges with the LKB1/mTORC1 and CREB to control metastatic disease in PCa mouse models. Unexpectedly, STAT3 was found to be upregulated in diabetic PCa patients undergoing metformin therapy with a concomitant reduction in mTORC1 expression. In preclinical mouse models of PCa, genetic ablation or activation of STAT3 had opposing effects on LKB1/AMPK/mTORC1-dependent tumorigenesis. Using genetic and pharmacological approaches, we identified LKB1 as a direct STAT3 target while repressing CREB. Furthermore, PCa patients with high CREB expression had inferior clinical outcome with significantly increased risk of disease and metastatic recurrence. We observe that castration state lowers STAT3 abundance and increases AR and CREB levels, leading to castration-resistant PCa (CRPC). Our findings revealed that STAT3 controls mTORC1 and CREB in metastatic disease, suggesting CREB as a promising target for lethal CRPC.