Dual-specificity phosphatase 6 (DUSP6): a review of its molecular characteristics and clinical relevance in cancer

Overview of MAPKs

The signaling pathway of mitogen-activated protein kinases (MAPKs), which are also called extracellular signal-regulated protein kinases (ERKs), is a highly conserved signal transduction pathway with many important roles. The MAPK pathway is involved in modulating diverse cellular and physiological processes, such as cell growth, survival, apoptosis, inflammation, proliferation, differentiation, migration, development, immunity, and stress responses^1–6. It is also associated with human diseases, such as obesity, diabetes, rheumatoid arthritis, neurodegenerative disorders, and cancer^7–9. The MAPK pathway responds to extracellular stimulation by Gi-coupled receptor (GPCR), ultraviolet irradiation, genotoxic agents, and oxidative stress, and it also can be activated by growth factors and inflammatory and cytokine stimulation¹⁰.

MAPKs are proline-directed kinases that phosphorylate sites containing serine/threonine and tyrosine residues. They are activated by dual specificity MAPK kinases (MAPKKs), which are themselves activated by MAPKK kinases (MAPKKKs)¹¹. The MAPKs at the end of these signaling cascades phosphorylate their target proteins, including transcription factors, translational regulators, MAPK-activated protein kinases (MAPKAP kinases), phosphatases, and other classes of proteins¹². Thus, they regulate multiple biological activities, including cell proliferation, differentiation, cell cycle progression, and survival¹³.

The most intensely studied members of the MAPK pathway include ERK1/2 and ERK 5, the p38 kinases (p38a, b, g, and d), and c-Jun N-terminal kinase (JNK) 1, 2, and 3¹⁴. Activation of an ERK signaling pathway plays a role in mediating cell division, migration, and survival. ERK1 and ERK2 are activated by MAPK/ERK kinase (MEK) 1 and MEK2¹⁵. The p38 kinases are activated by dual MAP kinase kinases (MKKs) termed MKK3 and MKK6. Activation of p38 signal transduction plays an important role in the regulation of inflammation, cell cycle, cell death, development, cell differentiation, senescence, and tumorigenesis^16,17. JNK, which is also known as stress-activated protein kinase (SAPK), is activated by the upstream MKK4 and MKK7 kinases via dual phosphorylation of the Thr-Pro-Tyr motif¹⁸. Activation of JNK thereby mediates the regulation of transcription factors, such as c-Jun, c-Fos, activating transcription factor 2 (ATF-2), activator protein 1 (AP-1), p53, and Elk, and also phosphorylates many cytoplasmic substrates, such as cytoskeletal proteins and mitochondrial proteins like Bcl-2 and Bcl-xl^19,20. Subsequently, a plethora of cellular processes are triggered, including cell proliferation, apoptosis, autophagy, motility, metabolism, and DNA repair²¹, as reviewed by Sui et al.¹⁰.

MAPK signaling pathway in cancer

Aberrant regulation of MAPK cascades contributes to the development of many pathological conditions, including renal cell carcinoma²², pancreatic cancer²³, bladder cancer²⁴, and leukemia²⁵. Regulation of these cellular processes is complex and involves external growth factors and components, such as JNK, ERKs, and the p38 groups²⁶. The JNK signal transduction pathway is associated with the transformation of many growth factor-mediated pathways and oncogenes, such as c-Jun, ATF2, p53, NFAT4, Elk1, and DPC²⁶. Dysregulation of the JNK pathway affects survival signaling and apoptosis in mammalian cells. ERK1 and 2 are involved in the ERK pathway, which is considered to be the best characterized MAPK signaling pathway.

Activation of ERK1 and 2 leads to transactivation and causes changes in the expression of Ets, Elk1, Myc, and estrogen receptor (ER), which are important in cell differentiation, growth, and mitosis²⁶. Activation of JNK and ERK in association with connexin43 (Cx-43) might be involved in promoting bladder tumorigenesis via targeting the MAPK pathway²⁴. In prostate cancer, activation of ERK1/2 by osteopontin (OPN) through c-Raf and MEK1/2 promotes cell cycle arrest in PC3 human prostate cancer cells²⁷. The oncogenesis activity of Raf/MEK and ERK in promoting cell cycle arrest in prostate cancer cells may be regulated by p53 restoration^28,29. Introduction of the wild type p53 into p53 knockout cell lines, such as PC3 and DU145, activates and increases the expression of the RAF/MEK/ERK cascade and improves its sensitivity to chemotherapeutic drugs^28,29.

The p38 pathway is another well-studied component of the MAPK pathway. It is activated by cellular stress, including heat shock, high osmotic stress, protein synthesis inhibitors, proinflammatory cytokines, and certain mitogens²⁶. Activation of the p38 MAPK pathway is essential in determination of cell fate, differentiation, and progression of somatic cells, but not during the G2/M transition and mitosis^30,31. The activated p38 MAPK signaling pathway regulates various proteins related to Gaucher’s disease, inflammatory bowel disease, glioblastoma multiforme, and several cancers, such as pancreatic, lung, and breast cancer^23,32–36.

MAPKs are reported to have a significant relationship with various transcription factors, such as reactive oxygen species (ROS)³⁷, epidermal growth factor receptor (EGFR)²³, Cx43²⁴, and others. In somatic cells, ROS-mediated MAPK activation regulates cellular processes, such as cell proliferation, growth, survival, and death^26,38. The accumulation of ROS is essential for prolonged MAPK activation and cell death³⁸. Moreover, aberrant activation of EGFR/p38-MAPK signaling has been shown to control angiogenesis in pancreatic ductal adenocarcinoma by regulating the expression of Sp1-induced cyclooxygenase-2 and vascular endothelial growth factor²³.

Better understanding of the MAPK network will lead to new strategies to design a system to deliver anticancer drugs to cancer cells. Establishment of positive or negative regulatory feedback loop is important to discovering new ways of looking at the role of the MAPK pathway in cancer treatment. Activation of MAPK and its molecules in combination with anticancer drugs can significantly induce apoptosis in PC-3 human prostate cancer³⁷, colon cancer³⁹, and leukemia²⁵ cells. Recent reports have shown that anticancer drugs, such as Toyocamycin, can reduce cell viability and enhance apoptosis in human prostate cancer cells via crosstalk between ROS and the p38/ERK MAPK signaling pathway³⁷. The involvement of p38 activation in inducing apoptosis is in accordance with studies of diosgenin, which is a chemotherapeutic agent used to treat colon cancer³⁹. Understanding of MAPK activities with transcription factors and components in cancer cells with or without anticancer drugs is critical to understanding the involvement of these proteins in cancer regulation.

In this review, we focus on the dual specificity phosphatases (DUSPs), particularly DUSP6. DUSPs negatively regulate MAPK members (MAPK/ERK, SAPK/JNK, and p38) that are associated with cellular proliferation and differentiation⁴⁰. This review demonstrates our depth of understanding of DUSP6 as a natural terminator of MAPK function and regulation, and thereby as a pharmacological inhibitor in cancer disease. Better understanding of how DUSP6 works in carcinogenesis and in the MAPK signaling pathways may provide insights into improved diagnosis, treatment, and prognosis of cancer.

MAPKs and DUSPs

DUSPs and MAP kinase phosphatases (MKPs) have been recognized as key players in the negative regulation of MAPKs by dephosphorylating either the threonine or tyrosine residue, or both, of a conserved signature T-Y-X motif within the activation loop of the kinase^41–43. Ten distinct subfamilies of catalytically active MKPs within the larger family of DUSPs have been characterized from mammalian genomes⁴¹. They are subcategorized into three subgroups based on sequence homology, subcellular localization, and substrate specificity (Table 1)⁴⁴. Class I consists of four inducible nuclear MKPs (DUSP1/MKP-1, DUSP2, DUSP4/MKP-2, and DUSP5). Class II consists of three ERK-specific MKPs (DUSP6/MKP-3, DUSP7/MKP-X, and DUSP9/MKP-4), and Class III contains three MKPs that inactivate p38 and stress-activated JNK MAPKs (DUSP8, DUSP10/MKP-5, and DUSP16/MKP-7).

DUSP6 gene organization and evolution

The subcategorization of DUSPs into three subgroups is also supported by the phylogenetic tree of DUSP sequences with consideration of substrate preferences⁴⁵ (Figure 1). Class II may have diverged earlier, as the central protein is encoded by a single exon⁴¹. Class I and III may be more closely related to each other, as the central portion of the protein is encoded by two exons⁴¹.

• Download figure
• Open in new tab
• Download powerpoint

1

Phylogenetic tree of DUSP sequences. Vector NTI software was used to derive the phylogenetic tree of the mouse DUSP amino acid sequence alignment⁴⁵. Sequence differences between proteins in each DUSP protein are proportionate to the length of the branches of the phylogenetic tree.

All MKPs share a common N-terminal non-catalytic domain and a more conserved C-terminal catalytic domain. The C-terminal catalytic domain consists of an active site sequence that displays sequence similarity to the prototypic dual-specificity protein phosphatase VH-1 encoded by vaccinia virus⁴⁶. The non-catalytic domain consists of two regions with sequence similarity to the catalytic domain of Cdc25 cell cycle regulatory phosphatase⁴⁶. These findings illustrate that the domain of MKPs and Cdc25 share a common evolutionary origin with the Rhodanese family of sulphotransferases^46,47. In addition, the non-catalytic domain contains sequences that play a role in determining subcellular localization, as it harbors either a nuclear localization signal (NLS) or a nuclear export signal (NES)^48,49. DUSP6 contains a leucine-rich NES that is important for nuclear export of the phosphatase⁵⁰. All Class II MKPs/DUSPs, including DUSP6, share similar domain structure position¹. The non-catalytic domain also contains a conserved cluster of basic amino acid residues, known as the kinase interaction motif (KIM), which is involved in MAPK substrate recognition and binding^44,48,49.

Human DUSP6 is located on chromosome 12q21.33 and contains three exons consisting of 381 amino acids. The first exon encodes for the Cdc25/rhodanese-homology domain, KIM, and ends with the second Cdc25/rhodanese-homology domain⁴¹. Half of exon two and almost all sequences of exon 3 encode the functional phosphatase domain. The catalytic site is located in the third exon (Figure 2).

• Download figure
• Open in new tab
• Download powerpoint

2

The domain structure of MKP-3/DUSP6 is made up of the C-terminal catalytic domain and N-terminal non-catalytic domain consisting of the Cdc25/rhodanese-homology domain, kinase interaction motif (KIM), and nuclear export signal (NES). Domains in the encoded proteins are indicated by the shaded shapes, and the three exons of DUSP6 denoted by roman numerals (rectangles) are connected through lines representing introns^41,44,50,51.

Structural elements of DUSP6

The C-terminal domain contains a highly conserved protein tyrosine phosphatase (PTPase), DX26 (V/L)X(V/I) HCXAG(I/V)-SRSXT(I/V)XXAY(L/I)M, where X is any amino acid⁵². In DUSP6, the highly conserved C-terminal domain is made up of a tyrosine/threonine specific phosphatase designated sequence HCXXXXXR at the active site⁵³. The cysteine plays an important role in the nucleophilic attack of the phosphorus of the substrate ERK2, whereas arginine interacts directly with the phosphate group on phosphotyrosine or phosphothreonine for transition-state stabilization^53,54. This leads to rearrangement of the active residues, which enables the dephosphorylation of Thr183 and Tyr185 in ERK2⁵⁵. Thus, DUSP6 must undergo conformational rearrangement for its activation.

The KIM sequence in the N-terminal domain is conserved in all MKP/DUSP species⁵⁶. KIM has great substrate selectivity. It mediates the binding of MKPs/DUSPs with the conserved MAPK common docking domain without requiring the phosphorylation of MAPK for activation⁴¹. The KIM^DUSP6 docking site consists of negatively charged, highly acidic residues and a hydrophobic groove between the α and β reverse turns of the structure⁵⁶. It is located in a non-catalytic region opposite the kinase catalytic pocket⁵⁶. This leads to the binding of negatively charged residues on MKPs/DUSPs to the positively charged MAPK docking site, suggesting that the MKP/DUSP binding site is critical in determining the binding specificity and catalytic activity⁵⁷. Binding of specific MAPK substrates activates the catalytic activity by causing an allosteric rearrangement of the C-terminal catalytic domain, thus mediating the enzyme-substrate interaction^55,58.

DUSP6 regulation and interaction in cells

DUSP6 is tightly regulated at transcription, after-transcription, or after-translation^59–63. Transcriptional regulation occurs in the presence of putative regulatory sequences for forkhead transcription factors, including Ets-family transcription factors and phosphatidylinositol 3-kinase (PI3K) in the mTOR pathway⁶⁴. Functional analysis of the promoter regions of DUSP6 has demonstrated the involvement of Ets-family transcription factors in controlling the expression of DUSP6 in chick embryo after being activated by FGF8 signaling^64–66. The involvement of Ets-family transcription factors has functional impacts through the ERK pathway, especially via a negative feedback loop of FGF signaling. A luciferase reporter assay showed that the transcriptional activation of DUSP6 by FGF family members was dependent on intron 1 in the phosphatase gene. In addition, other ERK-inducing agonists play roles in the DUSP6 transcription regulation, such as EGF in Drosophila and PDGF^62,67.

Post-transcription regulation of DUSP6 can occur in several ways, which determine the stability of DUSP6 mRNA. MEK/ERK signaling and hypoxia have been demonstrated to regulate post-transcription of DUSP6. The half-life of DUSP6 mRNA in a 3’-untranslated region (3’-UTR)-dependent manner was shown to be increased by the luciferase reporter construct, Tk-luc, after Raf: ER was stimulated with 4-OHT7⁶⁸. Hypoxia as a result of basal ERK activity was found to increase the endogenous level of DUSP6 mRNA and the stability of the luciferase reporter construct containing its 3’-UTR in a HIF-1 gene dependent manner. In contrast, two proteins regulating mRNA stability tristetrapolin (TTP), a member of the TISI CCCH zinc finger protein family, and PUM2, a homolog of Drosophila pumilio, were found to decrease the levels of endogenous DUSP6 mRNA and the activity of the DUSP6 in its 3’-UTR luciferase reporter construct in the HEK293 cell line⁶⁸.

In the MAPK signaling pathway, DUSP6 undergoes highly specific interaction with only ERK1 and ERK2 at dual threonine and tyrosine residues of the TEY motif, which leads to DUSP6 inactivation⁶⁹. Arkell et al.⁶⁹ reported that DUSP6 binds to ERK1/2 in both yeast and human cells, but fails to bind to ERK5. The same study demonstrated that recombinant ERK2 can induce catalytic activation of DUSP6, and DUSP6 expression can de-phosphorylate a co-expressed ERK2 construct and block the MEK1-driven activation of GAL4-ELK1, an ERK1/2-regulated transcription factor. However, because ERK5 is closely associated with ERK1/2 in the sense of catalytic domain and presence of the TEY motif for phosphorylation, and because it has similar function in cellular proliferation and survival, previous studies have suggested that ERK5 could be another substrate for DUSP6^70–75. This discordance demands further study.

A previously suggested spatiotemporal model can explain the mechanism of interaction between DUSP6 and ERK1/2^60,64. In this context, phosphorylation of DUSP6 and dephosphorylation of ERK1/2 would create a negative feedback loop to control ERK activity and indirectly the MAPK signal, which is vital for cell proliferation and differentiation (Figure 3)⁴⁸. Two sites, serine 159 and 197, on DUSP6, have been identified as being involved in the phosphorylation that finally leads to degradation by proteasomes. The model clarifies the basis of the mutual interaction between DUSP6 and ERK1/2. In this respect, at the initial time point, active ERK1/2 is found mostly in the nucleus, rather than in the cytosol and ERK signal is maintained. Afterwards, the ERK signal begins to gradually diminish because of dephosphorylation of ERK1/2 by DUSP6. At a later point in time, however, active ERK begins to accumulate in the cytosol to re-sustain the signal. After signal recovery, active ERK begin to re-accumulate in the nucleus until the transmitted signal is diminished again and the same mechanism recurs⁶⁰.

• Download figure
• Open in new tab
• Download powerpoint

3

Inactivation of ERK1/2 by DUSP6. Binding of activated ERK1/2 to the ERK binding (EB) domain of DUSP6 results in conformational change, prompting phosphatase activation of the DUSP6 catalytic domain, leading to dephosphorylation of ERK1/2.

Interaction between DUSP6 and a well-known tumor suppressor gene, p53, occurs in a very unique manner and can determine cell’s fate. Genotoxic stress can spark cellular responses via p53 that result in cell cycle arrest and finally cell death. Piya et al.⁷⁶ reported that p53 inhibits the cell survival pathway by inducing DUSP6 upregulation, which enhances cell death through ERK 1/2 dephosphorylation, subsequently leading to Bcl-2 proteasomal degradation and Bad activation. This results in mitochondrial damage and eventually cell death⁷⁶. These findings suggest that ERK protects cells from apoptosis by phosphorylating Bad, resulting in dissociation of Bad from Bcl-xL and protection of Bcl-2 from proteasomal degradation. Simultaneously, p53 also transactivates pro-death target genes, including Noxa, PUMA, PIG3, Bax, ASC, Fas, and DR5, converging their signals to enhance the cell death pathway⁷⁶, as shown in Figure 4.

• Download figure
• Open in new tab
• Download powerpoint

4

A schematic diagram of DUSP6 regulation and interaction with other associated genes. The diagram shows the multiple effects of these interactions, including synthesis of various functional proteins, mitochondrial damage and cell death, cell cycle and growth arrest, and proteasomal degradation of DUSP6 due to phosphorylation and ubiquitination. These interactions also propagate signalling for cellular proliferation and differentiation.

Another well-known target of ERK signaling is E twenty six (Ets) 2 protein, a transcription factor that can bind to the conserved region on the DUSP6 promoter in response to FGF or phorbol 12-myristate 13-acetate (PMA). Ets also can bind to intron 1 of DUSP6 in response to active ERK activity and to Ets-binding sites in the p53 promoter region, causing transactivation of p53 and leading to apoptosis; this interaction was observed in the thymus of Ets2 transgenic mice and in Ets2-overexpressing HeLa cells^64,77–81. These findings establish a feedback regulatory loop among p53, DUSP6, Ets2, and ERK. In this scenario, Ets2 could be an activator for p53 and DUSP6, whereas p53 and DUSP6 may be inhibitors of Ets2 via ERK activation⁷⁶.

Another gene linked to DUSP6 is BAG3, which encodes a multifunctional protein that contains the BAG domain, which can bind to heat shock protein (Hsp) 70 as well as to the other proteins via the WW domain with a proline-rich (PXXP) repeat and IPV (Ile-Pro-Val) motifs. Its functions include controlling levels, localization, and activity of other interacting proteins as well as regulating cellular pathways, such as apoptosis, autophagy, cytoskeleton organization and motility, and mechanotransduction⁸³. Falco et al.⁸⁴ successfully identified a novel role of BAG3 in regulation of neo-angiogenesis and showed that its downregulation decreased angiogenesis. BAG3 was also found to interact with ERK and DUSP6, as removal of BAG3 resulted in reduced binding of DUSP6 to ERK and a subsequent sustained ERK level. This sustained ERK level caused reduction in phosphorylation of retinoblastoma (Rb) by cyclin D (Cdk4/5) and cyclin E (Cdk2), which in turn led to increased p21 and p15 levels and cell cycle arrest in the G1 phase. Thus, DUSP6 needs to form a complex with BAG3 to interact with phospho-ERK1/2⁸⁴.

DUSP6 has also been shown to interact with, and be phosphorylated by, protein kinase CK2α, a ubiquitous kinase. This was illustrated via the yeast two-hybrid system, which showed that DUSP6 can form a complex with CK2α, leading to modulation of ERK2-MAPK signaling and cross-regulation of DUSP6⁵⁹. Serine-arginine protein kinase 1 (SRPK1) is another gene that was found to interact with DUSP6. The signaling mechanism of SRPK1 in promoting gastric cancer progression was explored and it was suggested that DUSP6 might regulate the downstream effectors of SRPK1⁸⁵. Figure 4 shows a summary of the interaction between DUSP6 and associated genes.

Role of DUSP6 and its relevance in cancers

Previous research has reported that DUSP6 is associated with many types of cancer and that it plays a variety of roles. A tumor-suppressive role of DUSP6 via pivotal negative-feedback regulation of the ERK1/2 in the MAPK signaling pathway has been reported in many studies of diverse cancers, including pancreatic, lung, ovarian, esophageal squamous cell, and nasopharyngeal cancers^40,86–88. However, the conventional hypothetical mechanism does not apply in all situations, and instead DUSP6 plays a tumor-promoting or pro-oncogenic role in cancers such as human glioblastoma, thyroid carcinoma, breast cancer, and acute myeloid leukemia^8,89–92. In the case of melanoma, DUSP6 can have both functions depending on the histological subtypes of the cells⁹³. Therefore, DUSP6 definitely influences and determines the fate of the carcinogenic pathway of a given cancer. The functions of DUSP6 are extensively explored below and are summarized in Table 2. Understanding the functions of DUPS6 and its modulation in cancer may lead to potential development of DUSP6 as a molecular target for various types of cancer.

DUSP6 in pancreatic cancer

Furukawa and Horii⁹⁵ showed that there is loss of heterozygosity at 12q21 and 12q22.q23.1, where the DUSP6 gene resides. Molecular pathogenesis studies also revealed gain-of-function mutations of KRAS2, mostly in the pancreatic cell line. Mutation of KRAS2 coupled with DUSP6 loss caused uncontrolled cell growth^94,95. An immunohistochemical investigation of a primary pancreatic cell line showed upregulation of the DUSP6 gene in mildly and severely dysplastic/in situ carcinoma but downregulation in invasive carcinoma, especially in the poorly differentiated type⁸⁶. DUSP6 transcriptional suppression could be due to hypermethylation with histone deacetylation modification in the expressional control region in human pancreatic cancer⁹⁶. Interestingly, reintroduction of DUSP6 via adenovirus and its exogenous expression caused cell growth suppression and apoptosis, thus displaying the tumor-suppressive role of the DUSP6 gene in pancreatic cancer in vitro⁸⁶. The tumor-suppressive role of DUSP6 is also supported by other studies showing that restoration of chromosome 12 caused suppression of pancreatic cancer cell growth⁹⁷ and that exogenous overexpression of DUSP6 resulted in downregulation of MAPK1/ERK2 and AURKA as well as its related genes, such as AURKB, TPX2, and CENPA⁹⁸.

DUSP6 in lung cancer

The tumor-suppressive role of DUSP6 was also observed in non-small cell lung cancers (NSCLCs)^40,99–100. In microarray analysis of EGFR inhibition in EGFR-mutant NSCLC cells, several DUSPs, including DUSP6, were among the most highly and immediately regulated genes⁹⁹. DUSP6 expression levels were weaker in most lung cancer cell lines than in human bronchial epithelial cells, and restoration of its expression suppressed cellular growth¹⁰⁰. Screening of multiple NSCLC cell lines and primary human tumor specimens demonstrated that DUSP6 is expressed in tandem with ERK activation. The same study also showed dramatic downregulation of DUSP6 expression via pharmacologic inhibition of ERK activity⁴⁰. Small-interfering RNA (siRNA) knockdown in a highly expressed DUSP6 lung cancer cell line showed a significant increase in ERK activation and cellular proliferation. In contrast, plasmid-driven overexpression in a stably transfected lowly expressed DUSP6 lung cancer cell line displayed significantly reduced ERK activation, cellular proliferation, and finally apoptosis⁴⁰.

In addition to modulating oncogenesis, DUSP6 also influences the sensitivity of cancer cells to chemotherapy drugs either directly or indirectly. In lung adenocarcinoma cells, in which oncogenic rearrangement of the gene encoding for anaplastic lymphoma kinase (ALK), such as EML4-ALK, is observed, up to 40% of ALK+ patients failed to respond to ALK-inhibitor therapy^101,102. Hrustanovic et al.¹⁰³ reported that downregulation of DUSP6 promoted resistance to ALK inhibitors in vitro as well as in individuals with EML4-ALK–positive lung adenocarcinoma by reactivation of the MAPK pathway. A similar study on CAR 2 and CAR 3 cells that have substantial basal phosphorylation of ERK revealed that stable reconstitution of DUSP6 restored sensitivity of the cells to crizotinib and the ability of crizotinib to suppress MAPK signaling. This finding highlights the potential use of DUSP6 as a MAPK kinase regulator in drug design or drug target strategies¹⁰³.

DUSP6 in ovarian cancer

Loss of DUSP6 (MKP3) inflicted by ROS, which plays a tumor-suppressive role, was observed and linked to ovarian cancer progression. Previous research reported that more ROS are produced as ovarian cancer cells progress and develop, resulting in DUSP6 protein degradation via protein ubiquitination-proteosome; however, DUSP6 is restored after cells are treated with antioxidant^87,104,105. The DUSP6 tumor-suppressive role in ovarian cancer was further shown by a knockdown study of endogenous DUSP6/MKP3 by short hairpin RNA (shRNA), which resulted in increased ERK1/2 activity, cell proliferation rate, anchorage-independent growth ability, and resistance to cisplatin in ovarian cancer cells. Conversely, enforced expression of DUSP6/MKP3 in DUSP6/MKP3-deficient ovarian cancer cells significantly reduced ERK1/2 activity and inhibited cell proliferation, anchorage-independent growth ability, and tumor development in nude mice. The enforced DUSP6/MKP3 expression also sensitized ovarian cancer cells to cisplatin-induced apoptosis in vitro and in vivo⁸⁷.

DUSP6 in esophageal squamous cell carcinoma (ESCC) and nasopharyngeal carcinoma (NPC)

Significant loss of DUSP6 gene expression was found in 100% of 11 ESCC and 71% of 7 NPC cell lines. DUSP6 expression was also downregulated in 40% of 30 ESCC biopsies and 75% of 20 NPC biopsies in comparison to their respective normal counterparts. These results illustrate the tumor-suppressive role of DUSP6 in both cancers⁸⁸. DUSP6 overexpression caused impairment of epithelial-mesenchymal transition (EMT)-associated properties, polarized cell spheroid formation, increased expression of E-cadherin as EMT markers, and dramatic suppression of cell migration and invasion, further supporting the anti-tumor effects of DUSP6⁸⁸. DUSP6 expression was downregulated in ESCC cell lines and primary tumor specimens compared to their normal counterparts, and DUSP6 expression was also negatively associated with pathological grade. The same study also showed that exogenous DUSP6 expression significantly induced apoptosis, which illustrates the tumor-suppressive role of DUSP6 in ESCC carcinogenesis¹⁰⁶.

DUSP6 in human glioblastoma

DUSP6 has been reported to have a tumor-promoting role in certain cases. Messina et al.⁸⁹ reported that upregulation of DUSP6 gene transcription in primary and long-term cultures of human glioblastoma plays a tumor-promoting role and exacerbates the malignant phenotype. DUSP6 overexpression in glioblastoma cultures through replication-deficient adenovirus-mediated expression of DUSP6 was found to inhibit cell growth by inducing G1-phase delay and increased anchorage dependent, as well as clonogenic potential in vitro. The study also showed that cisplatin treatment failed to sensitize DUSP6-overexpressing glioblastoma cells in vitro and in vivo, indicating chemoresistance, which normally causes poor treatment results in patients with glioblastoma. However, depletion by antisense DUSP6-AS reversed the effect by lowering the threshold sensitivity to drug treatment⁸⁹. Another study revealed the relationship between glioblastoma multiforme and the EGFR gene. An EGFR mutant, which was likely to mediate high carcinogenic activity, caused upregulation of a small group of genes, including DUSP6, which influenced signaling pathways known to play a key role in oncogenesis and function in interconnected networks¹⁰⁷.

DUSP6 in thyroid carcinoma

Papillary thyroid carcinoma (PTC) cell lines and the majority of PTC and poorly differentiated thyroid carcinoma (PDTC) specimens have demonstrated overexpression of DUSP6, which has as a tumor-promoting effect⁹⁰. Activated BRAF, RET/PTC, and TRK oncogenes in thyroid carcinoma were able to upregulate DUSP6 expression and indirectly activate both the ERK1/2 pathway and its negative feedback mechanism. Levels of DUSP6 mRNA and protein were significantly higher and overexpressed in PTC and PDTC tissues, both in expression profile datasets and in patients’ surgical samples, reflecting a positive correlation between DUSP6 expression and activation of the ERK1/2 pathway⁹⁰. DUSP6 silencing resulted in reduction of branched morphogenesis in TPC1, reduced proliferation and enhanced apoptosis in the NIM-1 cell line, and reduced invasive ability of all four PTC cell lines, suggesting that DUSP6 plays a vital pro-tumorigenic role in thyroid carcinogenesis⁹⁰.

DUSP6 in breast cancer

DUSP6 expression was reported to play an oncogenic role in breast cancers^78,82,91. The specific role of the gene has been explored in triple-negative human breast cancer cells (MDA-MB-231 cell), in which small interfering RNA (DUSP6-siRNA) and microRNA (miR-145) were employed to suppress the DUSP6 gene through transfection into the human breast cancer cells. This resulted in reduced DUSP6 mRNA and protein expression, cell proliferation, migration and invasion of the cells, and arrest at the G0/G1 phase⁹¹. In another study, DUSP6 was upregulated in the MCF-7 breast cancer cell line after being treated with PMA, which later inactivated ERK1/2 as a response to the sustained ERK 1/2 activation and played a positive role in proliferation and migration of the cells⁷⁸. In contrast, silencing of DUSP6 accelerated cyclin-dependent kinase inhibitor upregulation (p21^Waf1/Cip1), prevented cell growth arrest and inhibition of colony growth in soft agar, and stimulated the invasive phenotype. These findings illustrate the tumor-promoting role of DUSP6⁷⁸. DUSP6 was also found to be strongly upregulated in HER2-overexpressed breast cancer that promotes malignancy¹⁰⁸. Another study reported that upregulation of DUSP6 rendered estrogen receptor-positive breast cancer cells resistant to the growth inhibitory effect of tamoxifen, illustrating the pro-oncogenic role of the phosphatase gene¹⁰⁹.

DUSP6 in acute myeloid leukemia (AML)

The tumor-promoting role of DUSP6 has been demonstrated in AML⁹². The protein-tyrosine phosphatase (PTP) expression profile was established and the role of Fms-like tyrosine kinase 3 with internal tandem duplication (FLT3 ITD) was explored in AML cells from patients and cell lines that showed three transmembrane classical PTPs and six non-receptor classical PTPs, including several DUSPs, such as DUSP6. Pharmacological inhibition and silencing technology (si/shRNA) caused downregulation but increased constitutive ERK1/2 activation⁹². Attenuation of DUSP6 in 32D cell pools resulted in moderate impairment of proliferation, suggesting its positive role in FLT3 ITD-dependent cell growth. Forced exogenous DUSP6 overexpression in stable transfectants reduced ERK1/2 phosphorylation but did not inhibit cell proliferation, illustrating the DUSP6 pro-oncogenic role in FLT3 ITD-positive AML⁹².

DUSP6 in endometrial carcinoma

Endometrial carcinoma is a prevalent cancer in females, and epigenetic alterations have been found to play critical roles in the development and metastasis of endometrial cancer^110,111. A DUSP6 gene is expressed in normal human uterine tissue¹¹², and its expression is higher at stage I than stage II of endometrial cancer¹¹³. Recent intensive methylation profiling in 8 hyperplasia, 33 primary, and 53 metastatic endometrial cancers revealed the presence of hypomethylated CpG islands of the DUSP6 gene in primary and metastatic endometrial cancer, which supports the presence of epigenetic involvement in controlling the expression of DUSP6 in endometrial cancer¹¹³. Another study showed that DUSP6 is expressed at high levels in endometrial carcinoma, and an in vitro study using Ishikawa cells showed that overexpression of DUSP6 enhanced the growth-promoting effect of estrogen in the endometrial adenocarcinoma cells¹¹⁴. These results revealed the involvement of DUSP6 in the estrogen and estrogen receptor signaling pathway as well as the participation of DUSP6 as a negative feedback regulator of the MAPK/ERK1/2 signaling pathway in the development of endometrial adenocarcinoma¹¹⁴.

DUSP6 in malignant melanoma

The role of DUSP6 in malignant melanoma is very unique because it is histological subtype dependent⁹³. Microarray data showed that 10% of the thick human primary melanoma had high DUSP6 expression and similar poor melanoma-specific survival, as the majority of thick primaries with low DUSP6 expression levels exhibited molecularly distinct melanoma subtype characteristics⁹³. In another study, the tumor-suppressive role of DUSP6 in human malignant melanoma was observed when overexpression of DUSP6 reduced ERK activation, reduced basal or cisplatin-induced expression of the DNA repair gene ERCC1, and increased sensitivity of melanoma to cisplatin¹¹⁵. In tumorigenic mouse melanocyte cell lines of a mouse xenograft, a low level of ERK phosphorylation, high expression of DUSP6, low expression of p-MEK and p-AKT, and retention of p16 and PTEN were observed⁹³. DUSP6 overexpression decreased ERK phosphorylation but increased anchorage-independent growth and invasive ability of their immortal mouse melanocytes. Ectopic expression of active MEK increased p-ERK levels, which suppress tumorigenic melanocyte invasiveness in a Transwell migration assay. These results illustrate the pro-oncogenic role of DUSP6 in mouse melanocyte cell lines⁹³. In contrast, the same study also showed that a human malignant melanoma cell line that contains the BRAF V600E mutation and elevated ERK activity has smaller and fewer anchorage-independent colonies and decreased invasive ability in the DUSP6-transfected cells, illustrating the DUSP6 tumor-suppressive role.

Wittig-Blaich et al.¹¹⁶ recently highlighted the oncogenic role of DUSP6 via ERK1/2-mediated phosphorylation. They analyzed four selected cancer cell lines, including melanoma, and constructed an isogenic single recombinant (ISR) of 108 genes in the melanoma cell line. The generated ISRs were subjected to cell cycle analyses, which demonstrated strong growth and a suppressive effect of five genes (DUSP6, RPS6KA2, MAPRE3, STARD3, and EMD). Further analysis showed consistent elevated levels of only DUSP6 mRNA expression in primary melanoma and melanoma cell lines compared to primary normal human epidermal melanocytes (NHEMs). The mRNA expression of the other four genes was not consistently elevated in primary melanoma and melanoma cell lines compared to NHEMs¹¹⁶. Analysis using siRNA-mediated knockdown of DUSP6 in BRAF wild-type melanoma resulted in either promotion of cell growth or no effect. However, knockdown of DUSP6 in BRAF V600E melanoma with high DUSP6 expression caused cell death via induction of apoptosis, which further illustrates the tumor-suppressive role of DUSP6¹¹⁶. BRAF V600E has been implicated in melanoma progression in several different mechanisms, such as by activation of the downstream MEK/ERK pathway, angiogenesis, evasion of senescence, and apoptosis¹¹⁷. In BRAF V600E melanoma, elevated expression of DUSP6 levels is required to compensate for BRAF V600E hyperactivation, which might otherwise trigger apoptosis via ERK1/2 downstream targets. This finding using BRAF V600E melanoma highlights the oncogenic role of DUSP6 via ERK1/2-mediated phosphorylation, and it successfully proves the concept that tumor suppressor DUSP6 could serve as a putative synthetic lethal target in melanoma with BRAF V600E mutations¹¹⁶.