C-JUN N-terminal kinases (JNKs) are members of the mitogen-activated protein kinase (MAPK) family. Various stimuli activate them, such as environmental stresses [1
], inflammatory cytokines [2
], and growth factors [3
]. The activated JNKs catalyze the phosphorylation of many substrates, resulting in the alteration of gene expression programs and, ultimately, a variety of cellular signaling processes, such as cell proliferation [4
], migration [5
], and apoptosis [6
]. Additionally, the JNKs play significant roles in programs such as embryonic development [7
], neural functions [8
], wound healing [9
], immunity [10
], metabolic diseases [11
], and tumor progression [12
There are three JNK genes—JNK1
, and JNK3
—from which multiple isoforms are expressed [13
]. While the JNK1 and JNK2 proteins are ubiquitous in human cells, JNK3 is found predominantly in the central nervous system [14
]. Similarly to other MAPK family members, such as the ERKs and p38 MAPKs, the JNKs are activated through multi-tiered phosphorylation cascades (Figure 1
]. External stimuli activate the MAP3Ks, such as MEKK1 [16
], TRAF [17
], ASK1 [18
], TAK1 [19
], HPK1 [20
], and MLK3 [21
]. The MAP3Ks transduce signals to the MAP2Ks, MKK7, and MKK4, which synergistically activate the JNKs through phosphorylation of a Thr and Tyr, respectively [22
]. Once activated, the JNKs phosphorylate a variety of downstream substrates, including the transcription factors C-JUN [23
], JUND [24
], ATF2 [25
], and ELK1 [26
Over the past decade, evidence has emerged for the role of the JNK signaling pathway in certain types of stem cells. Stem cells can maintain themselves through continuous cycles of cell division (self-renewal) or the generation of different types of daughter cells (differentiation). Stem cells have indispensable roles in ontogenesis, embryonic development, and organ homeostasis [27
]. Targeted disruption of JNK signaling-related genes such as JNK1, JNK2, MKK4, MKK7,
causes disorders in embryonic development [28
]. Furthermore, the activity of JNK signaling affects the proliferation and differentiation of tissue-specific stem cells, which mediate tissue homeostasis and regeneration [31
]. As described below, recent evidence has identified several aspects of JNK signaling that regulate self-renewal and differentiation. Here we provide an overview of stem cells and describe the current understanding of the function of JNK signaling within various types of stem cells and between stem cells and their microenvironment.
5. JNK Signaling in Cancer Stem Cells
JNK signaling mediates apoptosis [87
], a process known to suppress tumorigenesis. Thus, it is not surprising that JNK signaling can play a negative role in cancer development. Indeed, loss-of-function mutations in the MAP2K4 gene (which encodes MKK4) exist in approximately 5% of tumors, and multiple lines of evidence indicate that the activity of MKK4 can suppress tumor progression [88
]. Moreover, MKK7 is also known to function as a tumor suppressor. For example, MKK7 can promote senescence through p53 in response to oncogene activation and thereby negatively regulate tumorigenesis [89
]. However, JNK signaling is also involved in cell survival and implicated in pro-tumorigenic function [14
]. For instance, in a mutant Kras-driven lung cancer mouse model, conditional knockout of both Jnk1 and Jnk2 suppressed lung tumor formation, suggesting that JNK signaling plays a positive role in tumorigenicity [12
]. There is also evidence that JNK signaling is involved in the protection of tumor cells from undergoing premature senescence by preventing mitochondrial reactive oxygen species production through activation of BCL-2 [90
]. Thus, the role of JNK signaling in cancer development is likely to vary in a cell type and context-dependent manner.
Accordingly, the association of the JNK signaling pathway with CSCs is also complicated. Some evidence suggests that JNK signaling does not affect CSC activity in specific types of cancer. In contrast, other evidence indicates that JNK signaling contributes to promoting aggressiveness for various kinds of malignancies through maintaining CSC properties such as self-renewal, drug resistance, and tumor-initiating ability.
Girnius et al. [91
] reported that JNK signaling does not regulate CSCs. They showed that in a Jnk-knockout breast cancer mouse model, Wap-Cre+/−
) mice exhibited more rapid breast tumor formation than control Wap-Cre+/−
) mice, implicating the role of JNK signaling as a tumor suppressor. They showed that tumor cells derived from JNKKO
mice exhibited no difference in proliferation, migration, or invasion phenotypes or in stem cell activity, such as sphere formation in vitro compared to tumor cells from JNKWT
mice. Moreover, no differences are apparent between the growth of JNKWT
tumors in orthotopically transplanted mice. These findings implicate the suppressive role of JNK signaling in the initiation of murine breast cancer and the acceleration of tumor development, but only a minor role for JNK signaling in the growth of established tumors [91
]. Ohta et al. [92
] demonstrated that depletion of the histone H3K4 demethylase JARID1B by shRNA knockdown induced cellular senescence associated with JNK phosphorylation and reduced CSC population in colorectal cancer cells [92
]. These results imply that JNK signaling has a suppressive role in CSC proliferation through the induction of premature senescence.
In contrast, we reported that the JNK/C-JUN signaling pathway promotes the CSC phenotype of triple-negative breast cancer (TNBC) through the upregulation of NOTCH1 [93
]. Inhibiting JNK/C-JUN by siRNA-mediated knockdown or the JNK inhibitor JNK-IN-8 reduced CSC traits such as anchorage-independent growth and ALDH1 activity, as well as the migration and invasion activity of TNBC cells. Suppression of JNK signaling resulted in the downregulation of NOTCH1, which is a critical player in regulating self-renewal and cell fate determination of mammary stem cells (Figure 4
). We also confirmed that C-JUN regulates NOTCH1 expression using a luciferase reporter assay and demonstrated that JNK-IN-8 repressed tumor growth in a TNBC xenograft mouse model via inhibition of CSC properties, including NOTCH1 and ALDH1 expression. These results support the role of JNK signaling in maintaining CSC properties and in crosstalk with NOTCH signaling [93
Nasrazadani and Van Den Berg [94
] also identified a role for JNK signaling in TNBC tumorigenesis that they attributed to the JNK2 isoform. In the metastatic 4T1.2 mammary tumor cell line, the shRNA knockdown of JNK2 reduced invasion by 80%. In vivo, shJNK2-expressing 4T1.2 cells in mice showed impaired tumor growth and reduced lung metastases. Jnk2−/−
mice also exhibited reduced bone metastasis, which was associated with impaired osteoclast differentiation resulting from JNK2 regulation of RANK and RANK ligand expression [94
]. Further studies are needed to similarly evaluate the potential roles of JNK1.
A growing amount of evidence also supports a role for JNK signaling in maintaining glioma stem cells. Six glioblastoma cell lines expressed higher JNK activation in their stem cell populations than in their differentiated cell populations [95
], and JNK activation correlates with glioma histological grade [96
]. Matsuda et al. [95
] found that JNK inactivation via siRNA or SP600125 reduced stemness in glioblastoma cells, as measured by spheroid formation and marker expression (nestin, SOX2, Musashi-1), and induced differentiation. Additionally, transient inactivation of JNK by either pharmacological or genetic methods in a mouse model of glioblastoma generated a sustained loss of tumor-initiating capacity. While some studies indicate that JNK1 and JNK2 both contribute to the CSC phenotype in gliomas [95
], others have implicated JNK2 as the predominant acting isoform [98
In pediatric patients with T-cell acute lymphoblastic leukemia (T-ALL), Shen et al. [99
] showed that KLF4 expression was suppressed, which resulted in activation of MKK7 as well as the downstream JNK pathway, including JNK, ATF2, and C-JUN. They also showed that loss of Klf4 expanded NOTCH1-induced leukemia-initiating cells in mouse T-ALL models, and pharmacological inhibition of JNK using AS602801, JNK-IN-8, and CC401 reduced tumor progression of human T-ALL [99
], implying potential collaboration between JNK and NOTCH signaling for strengthening tumor-initiation ability in CSCs.
There is evidence to support the JNK pathway regulating CSCs in numerous other cancers as well. For example, Okada et al. [100
] used a combination of SP600125 treatment and siRNA knockdown of JNK1 and JNK2 to show that K-RAS/JNK signaling maintains stemness in pancreatic CSCs. JNK signaling also maintains self-renewal and tumor-initiation capabilities of ovarian CSCs. Seino et al. [101
] found that in CSCs cultured from A2780 ovarian cancer cells, transient siRNA knockdown of JNK1 and JNK2 resulted in reduced expression of C-JUN and the stem cell markers NANOG, SOX2, and nestin. In a mouse intestinal model, Sancho et al. [32
] found that transgenic expression of a constitutively active JNK1 fusion protein (JNKK2-JNK1ΔG
) showed an increase in ISC proliferation and villus length. They attributed this result to a positive feedback loop that links JNK and WNT signaling, in which the WNT target Tcf4 is also a gene target of C-JUN, and C-Jun is a target of the TCF4/β-catenin complex. In the same study, JNKK2-JNK1ΔG
showed increased tumorigenesis in a model of colitis-induced intestinal carcinogenesis. Interestingly, this result was not observed in mice with Apc mutant–induced intestinal tumorigenesis, likely indicating that different signaling outcomes arise from different endogenous JNK pathway stimuli [32
Further studies are needed to elucidate such paradoxical aspects of JNK signaling in cancer development and CSCs. Much of the evidence so far is based on experiments using pan-JNK inhibitors or compound disruption of JNK expression. The relative expression levels of the JNK proteins in each type of tumor cell is poorly understood. Each isoform of JNK likely has many non-redundant functions; for example, the catalytic activity of the individual JNK isoforms toward different substrates may vary. Indeed, some studies imply different roles of JNK isoforms in particular types of malignant tumors. For example, JNK1, but not JNK2, has been reported to be required for lymphoblast transformation by BCR-ABL [102
] and carcinogen-induced hepatocellular carcinoma [103
], while some studies implicated JNK2 as the primary isoform involved in glioma cell proliferation [98
] and others indicated both JNK1 and JNK2 contribute to maintaining self-renewal and tumorigenicity of glioma CSCs [95
]. Thus, isoform-specific studies can be vital in assessing the roles of JNK signaling in tumorigenesis as well as in the regulation of CSC phenotypes.
7. Concluding Remarks
Research on stem cell biology offers new insight into potential applications, such as regenerative medicine and treatments for congenital diseases and malignancies [123
]. However, the mechanisms for regulating stem cell traits are still mostly unknown. For example, it is unclear why some types of stem cells tend to be dormant and maintain an undifferentiated status, while others actively proliferate and differentiate. The transcriptional programs for proliferation and differentiation may not be the same among stem cell types, and the signaling pathways which regulate the program also can vary for different types of stem cells. As discussed in this review, the JNK signaling pathway regulates both normal stem cell and CSC properties. The role of JNK signaling in maintaining a self-renewal/undifferentiated state or promoting differentiation is not straightforward, however, and further studies to elucidate the cell context-dependent role of JNK signaling should provide for a better understanding of the characteristics of each stem cell type.
Significant questions to address are how cells balance the pro-apoptotic and pro-survival/tumorigenic roles of JNK signaling and how differences in the regulation, enzymatic activities, and expression of the JNK isoforms impact their control of stemness. In addition to genetic approaches (e.g., RNA interference, CRISPR-Cas9, or genetically engineered mouse models), there needs to be an emphasis on a better understanding of the differential mechanisms of isoform regulation, the development of isoform-specific JNK inhibitors, as well as the utilization of chemical biology tools that allow for rapid downregulation of specific isoforms using pharmacological approaches [124
]. Moreover, the crosstalk of the JNK signaling pathway with stemness-related pathways such as WNT and NOTCH, as described in several contexts above, is also largely unexplored and may be vital to unveiling the role of JNK signaling in regulating stemness.
Despite several preclinical studies supporting therapeutic strategies for targeting CSCs, the evidence for clinical efficacy is still limited [126
]. Although the classical CSC model assumes an irreversible and straightforward hierarchy of cancer cells, recent studies suggest subclonal genetic diversity among CSCs [127
]. Also, some studies using newer techniques have found that cells can manifest plasticity between CSC and non-CSC characteristics [129
]. This genetic diversity and flexibility of CSCs gives rise to complex intratumoral heterogeneity and may make CSC-targeted therapy challenging. One possible solution is targeting the signaling between the CSCs and the TME or CSC niche, which supports the maintenance of the stemness of CSCs. This notion follows from the premise that if the CSC population is dynamic and plastic, CSCs will always re-emerge after targeted ablation if the niche remains intact. JNK signaling, as described above, plays several roles in regulating CSC self-renewal in both CSCs themselves and the CSC niche. Therefore, targeting the JNK signaling pathway has the potential to be an attractive therapeutic strategy in terms of multiple mechanisms, which may eliminate CSCs as a whole by both inhibiting CSC stemness and suppressing the CSC niche. The roles of JNK signaling in normal stem cells and CSCs are still mostly unclear and controversial because of the conflicting oncogene/tumor suppressor results. However, growing evidence suggests a crucial role of JNK signaling in regulating stemness in some types of stem cells or organs/tissues. For example, even though tumor-suppressive roles of JNK signaling are suggested by some reports, including in breast cancer, our group has shown that high JNK activity correlates with poor outcome of TNBC patients and that JNK inhibition reduced CSC proliferation, NOTCH1 expression, and ALDH1 activity in multiple TNBC cell lines as well as in a preclinical mouse model [93
]. The next step should be to identify the JNK isoform contributing to CSC and CSC niche regulation in each type of cancer as well as biomarkers to predict the subpopulation of cancer patients who will benefit from JNK-targeted therapy.
Overall, further studies of JNK signaling and stem cells will be essential for a better understanding of the mechanisms for regulating stemness of normal stem cells and developing novel therapeutic approaches targeting CSCs as promising strategies against malignant tumors.