Several kinase pathways are involved in phosphorylating FOXK2 and lead to distinct outcomes. Previous studies have shown that FOXK2 phosphorylation fluctuates during the cell cycle, peaking in the M phase. The CDK/cyclin complex catalyzes FOXK2 phosphorylation at S368 and S423, affecting its protein stability and transcriptional activity (35). In response to DNA damage, ATM promotes the phosphorylation of CHK2 at Thr68, enhancing its interaction with FOXKs. CHK2 then phosphorylates FOXK1 at S130 and FOXK2 at S61, facilitating their binding to 14-3-3 proteins. The binding of 14-3-3 sequesters FOXKs in the cytoplasm, releasing their inhibition on autophagy-related gene (ATG) expression and consequently enhancing autophagy (30). Furthermore, GSK3 phosphorylates FOXK2 at S415/S419, resulting in its sequestration within the cytoplasm without insulin stimulation (25). In summary, FOXK2 phosphorylation is tightly regulated by various kinase pathways throughout the cell cycle and in response to DNA damage, leading to distinct cellular outcomes such as transcriptional regulation and modulation of autophagy.

FOXK2 is SUMOylated at K527 and K633 by PIAS4 (24, 43). This modification triggers the translocation of FOXK2 into the nucleus, where it can be recruited to target gene promoters and activate transcription. For instance, binding to the promoter of the tumor suppressor FOXO3 mediates paclitaxel-induced cytotoxicity in breast cancer cells (44), while binding to the promoter of de novo nucleotide biosynthesis genes facilitates hepatocellular carcinoma growth and chemotherapy resistance (24).

FOXK2 undergoes acetylation by cAMP response element binding protein (CBP) and deacetylation by sirtuin 1 (SIRT1) at K223 (45). In the absence of cisplatin stimulation, FOXK2 remains hypoacetylated in the cell nucleus, impeding mitotic catastrophe and diminishing tumor cell apoptosis. However, cisplatin can disrupt the interaction between SIRT1 and FOXK2, leading to an increase in FOXK2 acetylation. Hyperacetylated FOXK2 exhibits reduced nuclear distribution, significantly affecting cell cycle-related genes, leading to cell cycle arrest and promoting apoptosis of cancer cells. Ultimately, this enhances tumor cell sensitivity to cisplatin (45). Acetylation may serve as a general mechanism to regulate FOX protein’s subcellular localization. For instance, FOXO3 can also be deacetylated by SIRT1, and it collaborates with FOXK2 to regulate the apoptosis of cancer cells (46). In contrast to FOXK2, SIRT1 deacetylates FOXO3 and mediates the nuclear export of FOXO3, which relieves the induction of apoptosis by FOXO3.

FOXK2 interacts with multiple co-inhibitory complexes (Figure 5), including NCoR/SMRT, SIN3A, NuRD, and REST/CoREST, to function as a transcriptional repressor. These complexes possess HDAC activity and modulate chromatin status through post-translational modifications, resulting in synergistic inhibition of a series of FOXK2-targeted hypoxia-related genes, including HIF1β and EZH2, effectively suppressing the hypoxic response (23). The recruitment of the Sin3A-HDAC co-inhibitory complexes to FOXK2 occurs under adequate nutrition conditions, resulting in alterations in nucleosome structure through inhibiting histone H4 acetylation and subsequently suppressing autophagy and atrophy-related genes. However, during starvation, dissociation of FOXK2 from chromatin leads to the loss of synergistic activity with Sin3A-HDAC, inducing autophagy (18).

In addition to binding co-inhibitory complexes with acetylation activity, FOXK2 also interacts with Polycomb repressive deubiquitinase (PR-DUB) complexes, which have BRCA1-associated protein 1 (BAP1) as the core component (47). Under physiological conditions, additional sex combs like 1 (ASXL1) can bind to BAP1 which subsequently binds to chromatin by interacting with FOXK2 (20, 47). This interaction modulates the chromatin microenvironment by regulating H2AK119 ubiquitination and controlling the expression of FOXK2 target genes, including Von Hippel-Lindau syndrome (VHL), suppressor of cytokine signaling 1 and 2 (SOCS1/2), thioredoxin interacting protein (TXNIP), and others (20). However, C-terminal truncated ASXL mutants are frequently observed in myeloid neoplasms. Although they can still bind BAP1 normally, they cannot bind FOXK2. Consequently, the PR-DUB complex cannot be recruited onto chromatin (20), leading to improper expression of downstream tumor suppressor genes of FOXK2, thereby promoting tumorigenesis.

FOXK2 not only synergizes with co-repressor proteins, but also collaborates with co-activator proteins to activate downstream target genes (Figure 5). As discovered by Ji et al., the binding regions of FOXK2 commonly include AP-1 binding motifs TGA(G/C)TCA, suggesting that FOXK2 and AP-1 exhibit a synergistic effect on transcription (48). Mechanistically, FOXK2 may facilitate the recruitment of AP-1 major components FOS and JUN to chromatin by opening up the chromatin structure, thereby promoting AP-1-dependent target gene expression. Additionally, FOXK1 and FOXK2 are capable of binding to the PDZ domain-localized region (residues 250 to 355) of phosphorylated DVL through a hydrophobic motif in the FHA domain (L137-F145-F154), facilitating their interaction (17). Upon induction of Wnt signaling, CK1 and MARK kinases catalyze DVL phosphorylation, leading to its translocation into the nucleus where it interacts with FOXKs to promote transcriptional activation of downstream genes. Moreover, DVL can transduce Wnt signaling to the GSK3/β-destruction complex and inhibit its activity, promoting β-catenin stabilization (17). The formation of the β-catenin/TCF/cJun transcription complex is facilitated by FOXK2 and DVL under transcriptional regulation, enhancing Wnt/β-catenin signaling (49).

FOXK1 and FOXK2 have been linked to proliferation and metastasis in human colorectal cancer, which is associated with DVL nuclear localization and activation of the Wnt/β-catenin signaling pathway, the latter plays a pivotal role in regulating cell proliferation and differentiation, self-renewal, tissue homeostasis as well as embryonic development (58, 59). The Wnt signal triggers DVL phosphorylation, leading to its translocation into the nucleus and interaction with FOXK2 to facilitate transcriptional activation of the β-catenin/TCF/cJun complex, thus reactivating Wnt/β-catenin signaling (17).

Protein quality and cellular homeostasis are critical for cancer survival and progression. FOXK2 participates in endoplasmic reticulum (ER) stress control to promote the maintenance of stemness and tumor progression in ovarian cancer stem cells (CSCs) (55). It binds to the distal region of the second intron of the unfolded protein response (UPR) sensor protein IRE1α and directly regulates its expression. IRE1α possesses both endonuclease and kinase activities; it splices the mRNA encoding the transcription factor X-box binding protein 1 (XBP1). The production of an XBP1 splice variant exhibits potent transcriptional activity and confers protection against apoptosis induced by ER stress (60). In cancer cells, the UPR pathway is usually activated, probably related to metabolic stress caused by accelerated nucleotide synthesis and cell proliferation, highlighting the role of FOXK2 metabolic regulation and protein homeostasis for cancer development.

In human colorectal cancer, FOXK2 also exhibits a positive feedback loop as a regulatory factor in the cell cycle and tumor progression. EGF binds to EGFR on the cell surface and activates FOXK2 expression via the extracellular signal-regulated kinase (ERK)/NF-κB pathway. Subsequently, FOXK2 activates the transcription of EGFR and ZEB1, crucial regulators of the epithelial-mesenchymal transition (EMT), invasion, and metastasis in colorectal cancer (41). Notably, FOXK2’s regulation on EGFR can be context-dependent and cell-specific; it functions as a tumor suppressor gene in ccRCC by inhibiting EGFR and inducing apoptosis (29), contrasting its role in colorectal cancer (41).

Furthermore, FOXK2 is significantly upregulated in hepatocellular carcinoma (HCC) cells. Studies have demonstrated that elevated FOXK2 can enhance Snail expression and reduce E-cadherin levels, promoting EMT in HCC cells (56). This process is closely associated with PI3K/AKT signaling pathway activation.

In summary, FOXK2 emerges as a potential biomarker for certain tumors in which it plays a promoting role, indicating its utility in guiding tailored tumor treatment strategies. By identifying FOXK2 expression patterns, clinicians can better understand the underlying mechanisms driving tumor progression and select more personalized therapeutic approaches for patients. This highlights the importance of FOXK2 as not only a diagnostic marker but also a potential therapeutic target in cancer management.

FOXK2 plays a regulatory role in hypoxia pathway, prevalently in locally advanced solid tumors, and holds significant pathological and physiological effects in tumor progression and invasion (61). FOXK2 interacts with various co-suppressor transcriptional complexes including NCoR/SMRT, SIN3A, NuRD, and REST/CoREST, to inhibit multiple hypoxia response genes such as HIF1β and zeste homolog 2 (EZH2) (23). In breast cancer cells, FOXK2 is positively regulated by estrogen receptor alpha (ERα). However, FOXK2 also plays a role in the degradation of ERα through the ubiquitin E3 ligase BRCA1/BARD1 complex (50). As breast cancer progresses, there is a gradual loss of FOXK2 expression, leading to the activation of the hypoxia pathway and an increase in the expression of enhancer of EZH2. Notably, the hypoxia signal further increases the level of EZH2, down-regulating FOXK2. This intensifies the activation of the hypoxia pathway, promoting EMT and metastasis in breast cancer (23). Elevated expression of EZH2 serves as a significant hallmark of invasive breast cancer (62). In general, low FOXK2 expression correlates with higher histological grade, positive lymph nodes, and ERα^-/PR^-/HER2^- negativity.

In myeloid tumors, ASXL1 C-terminal truncated mutants are common (63). The C-terminal truncated ASXL1 mutant protein retains its binding ability to BAP1, but loses its interaction with multiple DNA-binding transcriptional regulators, including FOXK1 and FOXK2 (20). As a result, multiple tumor suppressor genes, such as zinc finger protein 516 (ZNF516), membrane-associated guanylate kinase 1(MAGI1), SOCS1/2, TXNIP and VHL, are down-regulated, ultimately promoting cancer development and disrupting cellular metabolism and homeostasis (20).

In NSCLC cells, FOXK2 inhibits EMT through the PI3K-Akt pathway, contrasting with its promoting role in hepatocellular carcinoma cells (54). It suppresses CDK4 and cyclin D1 expression, arresting cells in the S phase, thus impeding NSCLC proliferation (54). Furthermore, FOXK2 expression is inversely correlated with gastric cancer (53) and glioma (52) grade as well as prognosis. However, further exploration is required to understand the regulatory mechanisms of FOXK2 in gastric cancer and glioma. Additionally, FOXK2 exerts its tumor suppressive effects through various circRNAs that sponge miRNAs and upregulate FOXK2 expression. For instance, Circ-ITCH (51) and circ-UBAP2 (57) function as tumor suppressors in cervical cancer and ccRCC, respectively, by sequestering miR-93-5p51 and miR-148a-3p52, thereby enhancing the expression of FOXK2. Remarkably, hypermethylation near the upstream promoter of miR-602 (40) negatively regulates its expression, upregulating FOXK2 to inhibit the invasion and metastasis of ESCC.

The dual role of FOXK2 reveals the intricate complexity and specificity of its involvement in tumor development and progression. Understanding these nuances is crucial for elucidating the precise role of FOXK2 in different cancer types and for developing targeted therapeutic strategies. Moreover, it underscores the need for further research to unravel the intricate molecular mechanisms underlying FOXK2’s function in cancer, ultimately leading to more effective treatments and improved patient outcomes.

Cancer cells often develop drug resistance through protective mechanisms, such as damage-induced autophagy (30). Autophagy is a cellular process that maintains homeostasis by removing misfolded proteins and damaged organelles in response to various stressors (64). FOXKs act as transcriptional repressors in autophagy (18), contributing to chemotherapy resistance in cancer cells. Specifically, the ATM enzyme promotes phosphorylation of CHK2 at T68, enhancing interactions between CHK2 and FOXKs. Subsequently, CHK2 phosphorylates FOXK1 and FOXK2 at S130 and S61, respectively, sequestering them in the cytoplasm through binding with 14-3-3 proteins (Figure 4). This cytoplasmic localization alleviates ATGs inhibition, promoting autophagy (30). Hyperphosphorylation of FOXKs, caused by cancer-derived mutations, promotes autophagy and chemoresistance. However, combining the autophagy inhibitor chloroquine (CQ) with chemotherapy drugs, such as cisplatin (30), effectively overcomes chemoresistance induced by FOXKs mutants.

Another factor contributing to chemotherapy resistance is the abnormal activation of the de novo nucleotide synthesis pathway, crucial for rapid tumor proliferation (Figure 4) (24). A FOXK2-specific binding motif and a SUMO-related signal exist within the promoter region of genes governing nucleotide biosynthesis. PIAS4 catalyzes FOXK2 SUMOylation at K527 and K633 (24, 43), facilitating nuclear translocation of FOXK2 and promoting de novo nucleotide synthesis, leading to hepatocellular carcinoma resistance against the chemotherapeutic drug 5-FU (24). Notably, FOXK2 SUMOylation can be inhibited by DNA damage induced by chemotherapy and radiation, presenting an opportunity to improve treatment efficacy by combing chemotherapy drugs with de novo nucleotide synthesis inhibitors, such as mycophenolate mofetil, MMF (24).

FOXK2 SUMOylation also plays a role in breast cancer treatment. SUMOylation facilitates its recruitment to the forkhead response elements (FHRE) region the FOXO3 promoter, activating downstream apoptotic target genes and enhancing paclitaxel cytotoxicity (44). However, in paclitaxel-resistant cells, FOXK2 accumulates in the nucleus but lacks SUMOylation, hindering recruitment FOXO3 for downstream gene activation, contributing to cellular resistance. Both PIAS4-mediated SUMOylation and CHK2-mediated phosphorylation of FOXK2 can modulate its nuclear localization and function. The two modification sites are spatially distant and do not interfere with each other during the regulatory process, thus they synergistically control the perception of DNA damage signals.

Additionally, SIRT1 catalyzes FOXK2 deacetylation, reducing tumor cell apoptosis and decreasing sensitivity to the chemotherapy drug cisplatin (Figure 4) (45). However, cisplatin stimulation inhibits SIRT1 and FOXK2 interaction, leading to increased acetylation levels at K223 of FOXK2, enhancing sensitivity to chemotherapy drugs. Therefore, the combined application of cisplatin and SIRT1 inhibitors offers promising prospects for developing novel cancer chemotherapy strategies (45).

FOXK2 emerges as a potent regulator of lipid oxidation and aerobic glycolysis (25, 32). It undergoes translocation from the cytoplasm to the nucleus via the AKT/mTOR pathway, where it participates in the regulation of lipid oxidation related gene expression (25). Moreover, FOXK2 regulates the expression of genes involved in aerobic glycolysis by binding to their promoters, including phosphofructokinase, hexokinase-2, lactate dehydrogenase, and pyruvate kinase (32). Through regulation of glucose and lipid metabolism, FOXK2 is closely linked to tumor initiation and progression. It is worth noting that tumor cells significantly contribute to remodeling the tumor microenvironment (TME) by producing large amounts of lactate through aerobic glycolysis (65) and enhancing lipid uptake and oxidation (66). This TME remodeling, in turn, influences the metabolic profile and immunophenotype of immune cells, leading to their immunosuppression (67). For instance, regulatory T (Treg) cells metabolize lactate to support their proliferative and suppressive functions, and tumor-infiltrating Treg cells require lactate uptake to maintain their heightened suppressive function (68). Additionally, up-regulation of lipid uptake and FAO elevates lipid metabolism in tumor-associated macrophages (TAMs), Tregs and myeloid-derived suppressor cells (MDSCs), thereby promoting their immunosuppressive function (66). However, no direct association between FOXK2 and the tumor microenvironment or immune response has been reported thus far; further investigation is warranted.