Potential tumor suppressive pathway involving DUSP6/MKP-3 in pancreatic cancer

Abstract

We previously found frequent loss of heterozygosity at 12q21 and 12q22-q23.1 in primary pancreatic cancers, and the DUSP6/MKP-3 gene residing in this region at 12q22 lost its expression in the great majority of pancreatic cancer cell lines. The DUSP6/MKP-3 protein is a dual-specificity phosphatase that dephosphorylates the active form of ERK, making a feedback loop to control ERK activity. Gain-of-function mutations of KRAS2 occur in the great majority of pancreatic cancer cells, and loss of expression of DUSP6/MKP-3 may synergistically promote constitutive activation of ERK and uncontrolled cell growth. To study loss of the feedback pathway and its impact on pancreatic cancer cell growth, we first investigated the expression of DUSP6/MKP-3 in primary pancreatic cancer tissues immunohistochemically; we found up-regulation in mildly as well as severely dysplastic/in situ carcinoma cells and down-regulation in invasive carcinoma, especially in the poorly differentiated type. Adenovirus-mediated reintroduction of DUSP6/MKP-3 into cultured pancreatic cancer cells induced strong expression of recombinant DUSP6/MKP-3 and reduction of phosphorylated ERK in a dose-dependent manner based on the multiplicity of infection and resulted in suppression of cell growth. Moreover, analyses by flow cytometry and immunocytochemistry revealed that the exogenous expression of DUSP6/MKP-3 induced apoptosis. These results show that DUSP6 exerts apparent tumor-suppressive effects in vitro and suggest that DUSP6 is a strong candidate tumor suppressor gene at 12q22 locus.

Figure 1.

Immunohistochemical staining of DUSP6 in pancreatic tissues using the indirect peroxidase method and using DAB as a chromogen. Counterstained with hematoxylin. The immunoreactivity was evaluated with grading and scoring intensely positive as 3+ (A), consistently positive as 2+ (B), focally or weakly positive as 1+ (C), or negative as 0+ (D). In cells of M, mild dysplasia; S, severe dysplasia/carcinoma in situ; W/M, well or moderately differentiated ductal adenocarcinoma; P, poorly differentiated adenocarcinoma. N denotes normal ducts. E shows complete abolishment of the immunoreactivity by neutralization of the antibody.

Figure 2.

A: Expression of ERKs and MKPs in cultured pancreatic cancer cells. The pancreatic cancer cells growing exponentially in culture media containing 10% fetal bovine serum were collected at 70% confluence by scraping. The cells were lysed and processed for immunoblotting performed with antibodies to active/phosphorylated form of ERK (p-ERK), ERK2, ERK1, DUSP6, MKP-1, MKP-2, and β-actin. Lane 1, PK-8; lane 2, PCI-35; lane 3, PK-1; lane 4, SU.86.86; lane 5, MIA PaCa-2; lane 6, BxPC-3. B: Structure of the replication-defective adenoviral vector expressing DUSP6-V5-His. The entire coding region of DUSP6 joined with V5 and His tags in its carboxyl terminus was cloned under cytomegalovirus enhancer (CMV) and chicken β-actin promoter (CA) in the E1 region of an Ad5 fragment which lacked E3. C: AdDUSP6 infection resulted in strong expression of the recombinant DUSP6 and suppression of active ERK in a dose-dependent manner based on multiplicity of infection (MOI). Pancreatic cancer cell lines, PCI-35 and PK-8 were infected at different MOI either with AdDUSP6 and AdLacZ. “Mock” designates no virus infection. Forty-eight hours after the infection, both floating and adherent cells were collected, processed for whole-cell lysate, and assayed by immunoblots with antibodies to V5 tag, DUSP6, active/phosphorylated form of ERK (p-ERK), ERK2, ERK1, and β-actin. Lane 1, mock infection; 2, AdDUSP6 at MOI 10; 3, AdDUSP6 at MOI 50; 4, AdLacZ at MOI 10; 5, AdLacZ at MOI 50.

Figure 3.

Growth of cells as shown by MTT assay. Ten thousand cells in 50 μl of the culture medium containing AdDUSP6 or AdLacZ to infect at MOI 10 and 50 were seeded in 96-well plates. The mock reaction contained no virus. On each day, cells were incubated with 0.05% MTT/PBS (-) for 1 hour and then suspended in 100% ethanol. Absorption at 590 nm was measured. Mean ± SD of actual values of the measurement from 8 independent wells in each experimental group were plotted on a logarithmic scale. Statistical analysis was performed by unpaired Student’s t-test.

Figure 4.

Increase in sub-G1 fraction after infection with AdDUSP6. The cells were infected at MOI 50 with either AdDUSP6 or AdLacZ and collected 3 days later. “Mock” indicates an experiment without infection. The cells were fixed overnight with 70% ethanol at −20°C. The fixed cells were stained with 100 μg/ml propidium iodide in PBS containing 10 μg/ml RNase A after hypotonic treatment with citric buffer. FACS Calibur System (BD Immunocytometry Systems) was used for analysis of DNA content. Bars and numbers indicate the range of the sub-G1 fraction and its rate in total counts, respectively.

Figure 5.

The recombinant DUSP6-V5 was detected directly in apoptotic cells by immunocytochemical staining. Three days after infection with AdDUSP6 at MOI 50, both floating and adherent cells were collected, fixed with 4% paraformaldehyde/PBS, and then spread on coated glass slides. Indirect immunofluorescence staining was performed using the anti-V5 antibody and FITC-conjugated secondary antibody. Nuclei were counterstained with 4′, 6′-diamidino-2-phenylindole dihydrochloride hydrate (DAPI). Non-specific cytoplasmic fluorescence is shown in red. Images represented are nuclei (A, E and I), V5-tag staining (B, F and J), cytoplasm (C, G and K), and those overlayed (D, H and L). Note apoptotic cells with fragmented nuclei in shrunken cytoplasm (arrow) and a cell harboring a deformed nucleus (asterisk) and expressing the recombinant DUSP6 (shown as green in V5 tag-staining).