Sustained activation of extracellular signal-regulated kinase (ERK) has been detected previously in numerous tumors in the absence of RAS-activating mutations. However, the molecular mechanisms responsible for ERK-unrestrained activity independent of RAS mutations remain unknown. Here, we evaluated the effects of the functional interactions of ERK proteins with dual-specificity phosphatase 1 (DUSP1), a specific inhibitor of ERK, and S-phase kinase-associated protein 2 (SKP2)/CDC28 protein kinase 1b (CKS1) ubiquitin ligase complex in human hepatocellular carcinoma (HCC). Levels of DUSP1, as assessed by real-time reverse transcription–PCR and Western blot analysis, were significantly higher in tumors with better prognosis (as defined by the length of patients' survival) when compared with both normal and nontumorous surrounding livers, whereas DUSP1 protein expression sharply declined in all HCC with poorer prognosis. In the latter HCC subtype, DUSP1 inactivation was due to either ERK/SKP2/CKS1-dependent ubiquitination or promoter hypermethylation associated with loss of heterozygosity at the DUSP1 locus. Noticeably, expression levels of DUSP1 inversely correlated with those of activated ERK, as well as with proliferation index and microvessel density, and directly with apoptosis and survival rate. Subsequent functional studies revealed that DUSP1 reactivation led to suppression of ERK, CKS1, and SKP2 activity, inhibition of proliferation and induction of apoptosis in human hepatoma cell lines. Taken together, the present data indicate that ERK achieves unrestrained activity during HCC progression by triggering ubiquitin-mediated proteolysis of its specific inhibitor DUSP1. Thus, DUSP1 may represent a valuable prognostic marker and ERK, CKS1, or SKP2 potential therapeutic targets for human HCC. [Cancer Res 2008;68(11):4192–200]
- hepatocellular carcinoma
- signal transduction
Human hepatocellular carcinoma (HCC) is one of the most common and deadliest tumors worldwide ( 1, 2). HCC is endemic in certain areas of Southeast Asia and Southern Africa, and its incidence is rapidly rising in Western countries ( 1, 2). Only few patients are amenable to surgery due to the late diagnosis of HCC, and alternative treatments do not significantly improve the prognosis of patients with unresectable HCC ( 1, 2). Thus, the investigation of HCC molecular pathogenesis is needed to identify new targets for its early diagnosis, chemoprevention, and treatment.
Recent studies showed c-myc, cyclin D1, cyclin A, and E2f1 up-regulation, rise in cyclin D1-Cdk4 and E2f1-Dp1 complexes, and pRb hyperphosphorylation in neoplastic liver lesions of c-Myc/Tgf-α transgenic mice ( 3) and Fisher 344 rats ( 4, 5) subjected to the carcinogens according to the resistant hepatocyte protocol ( 6), implying a deregulation of G1 and S phases in these lesions. Deregulation of these cell cycle components also occurs in human hepatocarcinogenesis ( 1– 5, 7). A major player favoring G1-S transition via induction of cyclin D1, CDK4, c-Myc, and pRB hyperphosphorylation is the RAS cascade, whose activation regulates numerous signals involved in cell growth, survival, and migration ( 8). The best-characterized RAS effector promoting cell cycle progression is the mitogen-activated protein kinase (MAPK) pathway ( 9). In this cascade, RAS induction triggers activation of RAF, MAPK kinase kinase (MEK), and extracellular signal-regulated kinase (ERK) proteins, leading to up-regulation of c-FOS, c-JUN, c-MYC, and ETS targets ( 9). Sustained ERK activity is associated with various types of tumors, including lung, ovary, colon, pancreas, and kidney ( 9). This frequently depends on up-regulation of the RAS/MEK cascade. However, constitutive ERK overexpression may also occur independently of the RAS/MEK signaling ( 10, 11). Recent studies ( 12, 13) indicate that prolonged activation of ERK promotes phosphorylation at the Ser296 residue of its inhibitor, dual-specificity phosphatase 1 (DUSP1; also known as MAPK phosphatase-1). Phosphorylation of this specific residue renders the DUSP1 protein susceptible to proteasomal degradation by two substrate recognition proteins belonging to a large S-phase kinase-associated protein (SKP)–cullin–F box ubiquitin ligase: the S-phase kinase-associated protein 2 (SKP2) and CDC28 protein kinase b1 (CKS1) complex. Thus, accelerated degradation of DUSP1 may further reinforce ERK activity and its cooperation with SKP2/CKS1 ubiquitin ligase. An additional mechanism leading to ERK up-regulation could be ERK-mediated induction of SKP2/CKS1 ubiquitin ligase, which controls DUSP1 ubiquitination. This effect could be at least partially attributed to induction by ERK of the FOXM1 gene ( 14) which, in turn, up-regulates the SKP2/CKS1 ligase ( 15). In contrast, transient activation of ERK leads to catalytic activation of DUSP1 followed by inactivation of ERK ( 13). This body of evidence indicates that DUSP1 feedback inhibits its activation by ERK and that DUSP1 might be a crucial regulator of ERK activity in the cell. Although much effort has focused on the molecular interactions modulating DUSP1 activity in mammalian cell lines, little is known about the role of DUSP1 in carcinogenesis. DUSP1 inactivation is frequent in prostate and urothelial tumors ( 16, 17), and recent observations indicate that immunohistochemical positivity for DUSP1 in human HCC is associated with longer patients' survival ( 18). However, the interactions of DUSP1 with ERK and the SKP2/CKS1 ligase and the mechanisms responsible for DUSP1 inactivation in the liver and other tumors have not been analyzed to date. Here, we evaluated the effects of these interactions, at molecular and functional levels, in human HCC subtypes with different survival times, in the attempt to evaluate the role of DUSP1 in hepatocarcinogenesis and correlate the effects of its molecular interactions with tumor growth and patients' survival and identify new potential prognostic markers and therapeutic targets.
Materials and Methods
Tissue specimens, cell lines, and treatments. Six normal livers, 42 HCCs, and corresponding surrounding nontumorous livers were used. Tumors were divided in 21 HCCs with poor prognosis (HCCP), characterized by <3 y survival, and 21 HCCs with better prognosis (HCCB) with >3 y survival after liver partial resection. Liver tissues were kindly provided by Dr. Snorri S. Thorgeirsson (National Cancer Institute). Institutional review board approval was obtained at participating hospitals and NIH.
In vitro growing 7703, HuH7, SNU-182, and SNU-387 human HCC cell lines were treated with 20 mmol/L UO126 (MEK inhibitor; EMD Chemicals, Inc.) or small interfering RNA (siRNA) against SKP2, CKS1, or ERK2; 7703 cells with 10 μmol/L 5-aza-2-deoxycytidine (5-Aza-dC; Sigma-Aldrich Corp.); and SNU-182 cells with 5 μmol/L Ro-31-8220 (DUSP1 inhibitor; EMD Chemicals, Inc.) or siRNA against DUSP1. The cells were grown 72 and 144 h in the presence of 5-Aza-dC and 12 and 24 h with all other inhibitors and siRNAs. siRNA experiments were done as reported ( 13, 19– 21). Transient transfection with either Ha-RAS cDNA (wild type), ERK2 (wild type) in a pUSEamp plasmid (Millipore), or SKP2 cDNA in a pCMV6-XL vector (OriGene Technologies) and stable transfection with ERK2 cDNA (wild type; Millipore) were done on SNU-182 cells following manufacturers' protocols. To determine the effect of proteasome-mediated proteolysis, the proteasome inhibitors N-acetyl-Leu-Leu-norleucinal (ALLN; 10 μmol/L) and carbobenzoxy-l-leucyl-l-leucyl-l-leucinal (MG132; 25 μmol/L) were added in the final 3 h of the transient transfection experiment ( 13).
Proliferation and apoptotic indices. Proliferation index was determined by counting proliferating cell nuclear antigen–positive cells, as reported ( 22). Apoptotic index was calculated by counting the apoptotic figures on tumor sections stained with the ApoTag peroxidase in situ apoptosis detection kit (Millipore) and expressed as percentage of the total number of cells counted. Cultured cell viability and apoptosis were determined by the WST-1 cell proliferation reagent and the cell death detection Elisa Plus kit (Roche Diagnostics), respectively.
Evaluation of microvessel density. HCCs were stained with mouse monoclonal anti-CD34 antibody (Vector Laboratories). Any brown-stained endothelial cell or endothelial cell cluster was counted as one microvessel, irrespective of the presence of a vessel lumen ( 23). Rare and small necrotic areas were excluded from the analysis. The four highest microvessel density (MVD) areas for each tumor were photographed at high power (200×), and the size of each area was standardized using the ImageJ software. MVD was expressed as the percentage of the total CD34 stained spots per section area (0.94 mm2).
Quantitative reverse transcription–PCR. Primers for DUSP1 and RNR-18 genes were chosen with the assistance of the “Assay-on-Demand Products” (Applied Biosystems). PCR reactions were done with 75 to 300 ng of cDNA, using an ABI Prism 7000 Sequence Detection System and TaqMan Universal PCR Master Mix (Applied Biosystems). Cycling conditions were 10 min of denaturation at 95°C and 40 cycles at 95°C for 15 s and at 60°C for 1 min. Quantitative values were calculated by using the PE Biosystems Analysis software and expressed as N target (NT). NT = 2−ΔCt, wherein ΔCt value of each sample was calculated by subtracting the average Ct value of the target gene from the average Ct value of the RNR-18 gene.
Methylation-specific PCR, bisulfite sequencing, and microsatellite analysis. High molecular weight DNA from human samples was isolated as reported ( 24) and modified with the EZ DNA methylation kit (Zymo Research). The CpGenome Universal Methylated DNA and CpG Universal Unmethylated DNA (Chemicon International) were used as positive and negative control for each reaction, respectively. Primers specific for methylated and unmethylated DUSP1 promoter were designed for methylation-specific PCR with the MethPrimer software ( 25). Results were confirmed by genomic bisulfite sequencing using a set of primers kindly provided by Dr. T. Visakorpi (University of Tampere). Loss of heterozygosity (LOH) of the DUSP1 locus was investigated with D5S677, D5S498, and WI-22759 primer pairs, as published ( 26). LOH was recorded when a 50% or greater reduction in electrophoretic band intensity was detected with silver nitrate staining (Silver Stain Plus, Bio-Rad). The complete list, sequence, annealing temperature, and product size of primers are shown in Supplementary Table S1.
Mutation analyses. Mutations at Ha-RAS, Ki-RAS, and N-RAS (codons 12 and 13), A-RAF (exons 10 and 13), B-RAF (exons 11 and 15), RAF-1 (exons 10 and 14), epidermal growth factor receptor (EGFR; exons 18–21), and DUSP1 (catalytic domain) genes in normal liver, HCC and respective surrounding nontumor liver tissues were assessed by direct DNA sequencing, as previously described ( 27– 30).
Western blots and immunoprecipitation analyses. Protein extracts from liver tissues were evaluated by separating 100 μg of total protein lysate by 10% SDS-PAGE, as reported ( 4). This protein amount allowed quantitative evaluation even in the tissues with low expression levels, although saturating conditions were sometimes reached in high expression tissues. As a consequence, some differences in protein levels may be underestimated. Proteins were then transferred onto nitrocellulose membranes and reacted with the antibodies listed in Supplementary Table S2. Hypoxia-inducible factor-1α (HIF-1α) activation was assessed by determining the HIF-1α/p300 complexes through immunoprecipitation ( 4) with the anti–HIF-1α antibody and probing the membranes with the goat polyclonal anti-p300 antibody. The complexes of DUSP1 with ERK2, SKP2, and CKS1 were determined by immunoprecipitating DUSP1 with anti-DUSP1 antibody and probing the membranes with antibodies against ERK2, SKP2, and CKS1. Immunoprecipitation of SKP2, followed by immunoblotting with CKS1 antibody, and DUSP1, followed by immunoblotting with the mouse monoclonal anti-ubiquitin antibody, was used to assess the SKP2/CKS1 complex and ubiquitinated DUSP1, respectively. Bands were quantified in arbitrary units by Molecular Imager ChemiDoc XRS using the Quantity One 1-D Analysis Software and normalized to β-actin levels.
Statistical analysis. Student's t test and Tukey-Kramer test were used to evaluate statistical significance. Fisher's exact test was used for comparative analysis of the survival of HCC patient subgroups. Linear regression analyses were done by GraphPad Instat 3 software. 1 Values of P < 0.05 were considered to be significant. Data are expressed as means ± SD.
Clinicopathologic features and mutational status. Data in Supplementary Table S3 show the presence of two distinct patients subgroups in our sample collection with mean survival of 67.51 and 9.19 months, respectively, after liver partial resection (P < 0.0001). On the basis of patients' survival length, HCCs were previously classified as tumors with better (HCCB) or poorer (HCCP) prognosis ( 22). No differences between the two subgroups occurred as concerns sex and age of patients, etiology, presence of cirrhosis, tumor size, differentiation (Edmondson and Steiner grade), and α-fetoprotein serum levels. Significant higher proliferation index and microvessel density and lower apoptotic index occurred in the HCCP than HCCB subgroup. As concerns the status of Ha-RAS, Ki-RAS, N-RAS, A-RAF, B-RAF, RAF-1, and EGFR, no mutations were detected in any of the samples analyzed.
ERK activation in human HCC. A progressive up-regulation of active ERK1 and ERK2 proteins was detected from nonneoplastic surrounding tissues to HCC, with the highest levels being detected in HCCP, when compared with normal livers. A similar expression pattern was found for ERK targets, including HIF-1α and HIF-1α/p300 complexes, vascular endothelial growth factor-α (VEGF-α), and HXKII ( Fig. 1 ). These data indicate that induction of ERK proteins is associated with both HCC development and progression, in accordance with a previous report ( 31).
Because the magnitude of ERK activity may be modulated by the specific inhibitor DUSP1 ( 12, 13), we determined DUSP1 expression at mRNA and protein levels. Quantitative reverse transcription–PCR (qRT-PCR) analysis ( Fig. 2A ) showed 2.9-fold increase in DUSP1 mRNA levels in HCCB. In contrast, no increase occurred in surrounding livers and 16 of 21 HCCPs with respect to control liver. In the other five HCCPs (23.8%), however, sharp DUSP1 down-regulation (mean value ± SD, 0.003 ± 0.0017) was found (data not shown in the figure). Subsequent methylation-specific PCR and microsatellite analyses showed concomitant DUSP1 promoter hypermethylation and LOH at the DUSP1 locus only in the five samples exhibiting DUSP1 down-regulation (Supplementary Figs. S1 and S2). In accordance with qRT-PCR results, DUSP1 protein levels ( Fig. 2) increased in HCCB. However, in contrast with DUSP1 mRNA expression patterns, marked down-regulation of DUSP1 protein levels occurred in all HCCPs. Importantly, no mutations were identified in the catalytic domain of DUSP1 in the collection of samples examined in this study, suggesting that posttranscriptional mechanisms may be responsible for DUSP1 down-regulation in HCCP in the absence of DUSP1 somatic mutations or promoter hypermethylation. Because a recent report indicates that cooperation between ERK and the SKP2/CKS1 ligase complex results in phosphorylation at Ser296 and ubiquitination of DUSP1 in lung carcinoma CL3 cell line ( 13), the levels of SKP2 and CKS1 proteins were assessed in the same collection of samples. Both SKP2 and CKS1 were up-regulated in HCC, mainly in HCCP. Furthermore, ubiquitinated DUSP1, as well as DUSP1-SKP2, DUSP1-CKS1, and SKP2-CKS1 complexes, were highest in HCCP, suggesting that DUSP1 down-regulation is achieved by SKP2/CKS1-dependent ubiquitination of DUSP1 in human HCC. These findings suggest the existence of a mutual control of DUSP1 and ERK levels in HCC. This agrees with the existence of a strong negative correlation between active ERK and DUSP1 levels in preneoplastic and neoplastic liver, as shown by linear regression analysis (ΔERK1/2/ΔDUSP1 = −0.8466, r = −0.9292, P < 0.0001).
DUSP1 levels inversely correlate with human HCC growth. Due to the different behavior of DUSP1 expression in HCCs with different survival rate, we evaluated the relationship of DUSP1 with clinicopathologic features to explore the prognostic role of DUSP1 in human HCC. HCCB exhibited proliferation index and MVD at 2-fold lower (P < 0.0001) and apoptotic index at 1.5-fold higher (P < 0.001) than HCCP (Supplementary Table S3). A significant inverse correlation of DUSP1 with proliferation index (r = −0.837, P < 0.0001) and MVD (r = −0.961, P < 0.0001) and a direct correlation with apoptosis (r = 0.626, P < 0.0001) and patients' survival length (r = 0.943, P < 0.0001) were found ( Fig. 3 ). No other clinicopathologic features, including etiology, presence of cirrhosis, α-fetoprotein levels, and tumor grading was detected. These results indicate that DUSP1 down-regulation is associated with HCC progression and aggressiveness regardless of the etiologic agent.
SKP2, CKS1, and ERK cooperation suppresses DUSP1 in human HCC cell lines. Functional consequences of DUSP1, ERK, SKP2, and CKS1 interactions were assessed in 7703, HuH7, SNU-387, and SNU-182 human HCC cell lines. High DUSP1 levels were detected by Western blot analysis only in the SNU-182 cell line ( Fig. 4 ). Methylation-specific PCR showed that DUSP1 down-regulation depended on DUSP1 promoter methylation in 7703 cells (data not shown), whereas DUSP1 low levels were associated with SKP2 and CKS1 overexpression in HuH7 and SNU-387 cells. Different approaches to modulate DUSP1 and related molecule levels in the HCC cell lines were applied ( Fig. 4, with quantitative analysis in Supplementary Figs. S3 and S4). Treatment with the demethylating agent 5-Aza-dC caused DUSP1 up-regulation only in 7703 ( Fig. 4A and Supplementary Fig. S3A), confirming silencing of DUSP1 by promoter hypermethylation. As expected, no effect of 5-Aza-dC on DUSP1 levels was detected in the other cell lines where no promoter hypermethylation occurred (not shown). In 7703 cells, DUSP1 induction by 5-Aza-dC was paralleled by down-regulation of SKP2, CKS1, pERK1/2, and its targets, HIF-1α and VEGF-α. Suppression of ERK activity, either via treatment with UO126 (MEK-ERK inhibitor; Fig. 4B and Supplementary Fig. S3B) or siRNA against ERK2 ( Fig. 4C and Supplementary Fig. S3C) increased DUSP1 expression, whereas diminishing DUSP1-Erk2 complex, ubiquitinated DUSP1, SKP2, and CKS1 levels, as well as SKP2-DUSP1 and CKS1-DUSP1 complexes in the SNU-387 cell line. Equivalent results were obtained in the HuH7 cell line (data not shown). Together, these results point to a role of ERK proteins in DUSP1 down-regulation and SKP2 and CKS1 transactivation.
Conversely, DUSP1 suppression in SNU-182 cells by both the DUSP1 inhibitor RO-31-8220 ( Fig. 4D and Supplementary Fig. S4A) and siRNA against DUSP1 ( Fig. 4E and Supplementary Fig. S4B) triggered up-regulation of SKP2, CKS1, pERK1/2, and ERK targets VEGF-α and HIF-1α, further substantiating the inhibitory role of DUSP1 on the ERK cascade. Moreover, suppression of SKP2 by siRNA in SNU-387 ( Fig. 4F and Supplementary Fig. S3C) and HuH7 (not shown) cell lines induced DUSP1 up-regulation and down-regulation of the complexes of DUSP1 with SKP2, CKS1, and ERK2, ubiquitinated DUSP1 and pERK1/2, and HIF-1α and VEGF-α, without affecting CKS1 levels. The decline in DUSP1-CKS1 complex, apparently contradictory to the absence of variations in CKS1 expression in cells treated with siRNA against SKP2, reflects the decrease in at least one component of the trimeric complex formed by DUSP1, SKP2, and CKS1 proteins. As expected, SKP2 inhibition had no effects on DUSP1 expression in 7703 and SNU-182 HCC cell lines (data not shown) due to the absence of posttranscriptional regulation in 7703 cells and the elevated DUSP1 activity in SNU-182 cells. Finally, down-regulation of CKS1 by siRNA resulted in DUSP1 overexpression and decrease in ubiquitinated DUSP1, without affecting SKP2 and pERK1/2 levels in SNU-387 ( Fig. 4G and Supplementary Fig. S3D) and HuH7 (not shown) cell lines. Taken together, these results assign a role to DUSP1 in the modulation of the ERK cascade and show that ERK, SKP2, and CKS1 are involved in DUSP1 degradation in human HCC.
The effect of the modulation of DUSP1 and related molecules on cell growth and apoptosis are shown in Fig. 5 . Induction of DUSP1 expression in 7703 and SNU-387 cells treated with 5-Aza-dC and UO126, respectively, was associated with a sharp inhibition of growth and increase in cell death. Equivalent results were obtained in the same cell lines treated with the siRNA against ERK2 (not shown). The same effect on cell growth and apoptosis followed the treatment of SNU-387 cells with SKP2 or CKS1 siRNAs. Finally, treatment of SNU-182 cells with either siRNA against DUSP1 or the DUSP1 inhibitor RO-31-8220 (not shown) resulted in a 50% increase in growth rate without affecting cell death.
The above results suggest a role of ERK and SKP2/CKS1 ligase in proteasomal degradation of DUSP1 in HCC. To further substantiate this hypothesis, we investigated the effect of the proteasomal inhibitors ALLN and MG132 on DUSP1 levels in SNU-182 cells transiently transfected with Ha-RAS, ERK2, or SKP2 ( Fig. 6 ). Transfection of pUSEamp/Ha-RAS, pUSEamp/ERK2, and pCMV6-XL/SKP2 constructs led to equivalent results, determining a remarkable increase in Ha-RAS, ERK2, and SKP2 expression with respect to cells receiving the plasmid alone ( Fig. 6A–C). These changes were not affected by ALLN, whereas MG132 induced further increase in ERK2, Ha-RAS, and SKP2 levels ( Fig. 6A, B, and D). Both Ha-RAS and ERK2 transfection resulted in DUSP1 down-regulation and increase in ubiquitinated DUSP1, but these effects were prevented by ALLN and MG132 ( Fig. 6A and B). Moreover, increase in SKP2 and CKS1 levels occurred in ERK2-transfected cells, which was not/poorly affected by proteasomal inhibitors ( Fig. 6B). SKP2 transfection led to sharp decrease in DUSP1 levels, associated with increase in ubiquitinated DUSP1, ERK2, DUSP1-ERK2 complex, and CKS1 and DUSP1 complexes with CKS1 and SKP2 ( Fig. 6C). Similar to that described for Ha-RAS–transfected and ERK2-transfected cells, DUSP1 down-regulation was prevented by proteasomal inhibitors in SKP2-transfected cells ( Fig. 6D). Finally, overexpression of ERK2 was reversed by siRNA against either SKP2 or CKS1 in ERK2-transfected cells ( Fig. 6E). In the latter cells, reduction of ERK2 levels via SKP2 or CKS1 siRNA was paralleled by DUSP1 up-regulation and decrease in ubiquitinated DUSP1. No complexes of ERK with SKP2 and/or CKS1 were found (data not shown).
A number of observations support a prominent role of the RAS-MAPK cascade in hepatocarcinogenesis ( 31– 34). This signaling pathway leads to ERK activation, which may favor HCC progression by promoting tumor growth and angiogenesis (Supplementary Fig. S4). Recent researches on the molecular interactions between ERK and DUSP1 have shown a reciprocal regulation between these two proteins in mammalian cell lines. The physical interaction of transiently activated ERK2 with DUSP1 induces the catalytic activation of the latter and the subsequent inactivation of MAPKs ( 12, 13). In contrast, sustained ERK2 activation triggers DUSP1 degradation via the ubiquitination proteasomal pathway ( 13).
The present results show a progressive unrestrained activation of ERK proteins from nontumorous surrounding liver to HCC, with the highest increase characteristic of HCCP, suggesting a role for DUSP1, whose expression is significantly higher in HCCB than HCCP, as a putative tumor suppressor which negatively regulates ERK in the progression stage of hepatocarcinogenesis. Mutations of the RAS, RAF, and EGFR genes, supporting active RAS-MAPK pathway, are rare in HCC (refs. 31– 35 and present study). RAS mutations and silencing of RAS inhibitors are mutually exclusive events in pancreatic ( 36), colorectal ( 37), and non–small cell lung cancer ( 38). Our results suggest a prominent role of ERK up-regulation in reducing DUSP1 expression in HCCP. Indeed, suppression of active ERK1, ERK2, SKP2, or CKS1 expression in cultured HCC cells resulted in DUSP1 up-regulation. Furthermore, ERK proteins were active in surrounding liver and HCCB (although at a significantly lower extent than in HCCP), where DUSP1 expression did not change (surrounding liver) or even increased (HCCB). These findings indicate that although DUSP1 inhibition theoretically could contribute to ERK overactivity, this is presumably not the main causative event responsible for pERK1/2 up-regulation in human HCC. This agrees with the notion that DUSP1 acts as a feedback inhibitor by limiting the duration or magnitude of pERK1/2 activity rather than preventing pERK1/2 expression ( 13). DUSP1 suppression by ERK proteins may instead further strengthen their effects on HCC growth by prolonging the half-life of active ERK.
DUSP1 down-regulation was significantly correlated to increased tumor aggressiveness and reduced patient's survival, strongly suggesting its prognostic role in HCC regardless of the etiologic agent, in accordance with previous reports in urothelial, prostate, and liver cancer ( 16– 18). In contrast, DUSP1 is overexpressed in breast, gastric, and lung cancer ( 39– 41), and its inhibition reduces pancreatic tumor development in nude mice ( 42). The means by which DUSP1 exerts such opposing effect on growth of different cancer types remain elusive. It has been shown that elevated DUSP1 protects cancer cells against Cisplatin-induced apoptosis by suppressing c-Jun-NH2-kinase (JNK) activity ( 43). However, the lack of JNK activation after DUSP1 inhibition by siRNA in SNU-182 HCC cell line (data not shown) argues against JNK regulation by DUSP1 in liver cancer. It has been shown that the protein IEX-1, an ERK substrate, plays a role in prolonging ERK activation ( 44), thus contributing to DUSP1 inactivation ( 13). IEX-1, upon phosphorylation by ERK, prevents cell death and favors cell proliferation ( 44). Therefore, relative amounts of IEX-1 protein and duration of ERK activation in different tumors should modulate DUSP1 effect. Interestingly, preliminary results indicate the presence of elevated amounts of IEX-1 protein in HCCP (data not presented), substantiating the role of low DUSP1 levels in these tumors.
A previous report suggested epigenetic regulation of DUSP1 expression in prostate cancer ( 16). We show here that DUSP1 down-regulation, similarly to that of RASSF1A oncosuppressor gene ( 34, 45), may occur through two mutually exclusive mechanisms, namely promoter methylation associated with LOH at DUSP1 locus or, most frequently, posttranscriptional events. The latter results from the combined activity of ERK, SKP2, and CKS1, leading to proteasomal degradation of DUSP1. This is shown as follows: (a) DUSP1 induction by 5-Aza-dC is paralleled by down-regulation of SKP2, CKS1, pERK1/2, and its targets (VEGF-A and HIF-1α); (b) SKP2 or CKS1 suppression by specific siRNAs is associated with DUSP1 up-regulation and down-regulation of ubiquitinated DUSP1, as well as of pERK1/2 and its targets; (c) siRNAs targeting SKP2 affects the formation of the SKP2/CKS1 complex (this does not occur by siRNA silencing of CKS1, suggesting that SKP2 is a rate-limiting step in ERK-mediated control of DUSP1 ubiquitination); (d) ERK inhibition by UO126 or siRNA rescues DUSP1 expression and induces a decrease in ubiquitinated DUSP1; (e) suppression of either SKP2 or CKS1 by specific siRNA markedly decreases ERK-dependent proteolysis of DUSP1 in ERK-overexpressing cells; (f) inhibition of DUSP1 by RO-31-8220 or DUSP1 siRNA is associated with up-regulation of SKP2, CKS1, pERK1/2, and its targets. Furthermore, the role of ERK and SKP2 in DUSP1 proteolysis is substantiated by the inhibition of DUSP1 degradation in SNU-182 cells overexpressing SKP2, ERK2, or Ha-RAS, in which proteasomal inhibitors were added to the culture medium. These findings confirm recent observations in embryonic kidney cells, where down-regulation of DUSP1 was achieved by ERK/SKP2/CKS1 cooperation, with sustained ERK activity promoting phosphorylation at the Ser296 residue of DUSP1, followed by SKP2/CKS1-dependent degradation of DUSP1 ( 13). Moreover, our results show a positive regulatory role of ERK on SKP2 and CKS1 expression in ERK-transfected cells. The mechanism underlying this effect is not clarified by the present data. However, it may be envisaged the possibility that ERK-induced up-regulation of FOXM1 gene ( 14) is followed by up-regulation of SKP2/CKS1 ligase ( 15). The induction of SKP1/CKS1 ligase by inducing DUSP1 ubiquitination should contribute to ERK sustained overactivity, which was indeed found in SKP2-transfected cells. Because unrestrained activity of ERK and overexpression of CKS1 and SKP2 are frequently observed in many tumor types ( 9, 20, 46), the present results may have broader implications in carcinogenesis. The cooperation of ERK and SKP2 has been shown in breast cancer, where the synergistic activity of the two oncogenes results in p27 degradation ( 47), and it has been suggested in cervical cancer ( 48). Furthermore, these findings are consistent with the observation that SKP2 accelerates the abilities of either activated N-RAS or Ha-RAS (both ERK upstream activators) to induce cell transformation and tumor development ( 45, 49). According to recent findings, the Ataxia telangiectasia mutated kinase gene, which plays a crucial role in the cellular response to DNA damage, inhibits ERK activity via increased expression of DUSP1 ( 50). This observation suggests a new mechanism of the anticancer action of DUSP1 based on the enhancement of DNA repair. Intriguingly, HCCPs characterized by low DUSP1 expression exhibit elevated genomic instability, differently from HCCBs with relatively high DUSP1 activity ( 31). However, further work is needed to clarify the relationships between variations in DUSP1 expression, DNA repair, and genomic instability in HCC.
In summary, the present findings show the disruption of the mutual control between DUSP1 and RAS-ERK pathway during hepatocarcinogenesis progression and emphasize a potential role of DUSP1 in determining the prognosis of human HCC. DUSP1 suppression is associated with unrestrained activation of ERK and its downstream effectors in HCCP (Supplementary Fig. S5), resulting in increased proliferation, survival, and angiogenic properties of neoplastic hepatocytes. Furthermore, these results point to a major role played by the cooperation between ERK, SKP2, and CKS1 in transmitting a signal to destroy DUSP1 and envisage the possibility that these oncogenes may inactivate other tumor suppressor genes. Also, the data suggest that therapeutic strategies aimed at reactivating DUSP1 and/or suppressing ERK, SKP2, and CKS1 might be highly beneficial for the treatment of human HCC.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Grant support: Associazione Italiana Ricerche sul Cancro, Ministero dell'Istruzione, Università e Ricerca, and Assessorato Igiene e Sanità RAS.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
- Received November 8, 2007.
- Revision received January 24, 2008.
- Accepted February 26, 2008.
- ©2008 American Association for Cancer Research.