Expression and functional significance of CDC25B in human pancreatic ductal adenocarcinoma

Article metrics

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of cancer-related deaths. Deregulation of cell-cycle control is thought to be a crucial event in malignant transformation, and CDC25 phosphatases are a family of cyclin-dependent kinase activators, which act at different points of the cell cycle, including G1–S and G2–M transition. Here, we investigated the expression and functional significance of CDC25s in PDAC. CDC25B mRNA expression levels in human pancreatic tissue samples were analysed by cDNA array, quantitative PCR and Northern blotting. Immunohistochemistry was carried out to localize and quantify CDC25B expression. Two specific CDC25B inhibitors were utilized to determine the functional relevance of CDC25B. By quantitative RT–PCR, CDC25B mRNA was overexpressed in pancreatic cancer (7.5-fold) in comparison to the normal pancreas. Strong nuclear CDC25B immunoreactivity was present in both pancreatic and metastatic cancer samples, and there was a marked increase of the percentage of positive cells in primary cancer (48.6±16.3%) and metastatic tissues (71.7±3.1%) compared to normal samples (8.3±1.8%). Two CDC25B inhibitors reduced the growth of pancreatic cancer cell lines, resulting in the accumulation of phosphorylated CDC2 and G2/M arrest. These findings demonstrate an important role of CDC25B in cell-cycle progression, raising the possibility that inhibition of CDC25B may have therapeutic potential in pancreatic cancer.

Introduction

Pancreatic ductal adenocarcinoma (PDAC) is the fourth leading cause of cancer-related deaths in the United States according to the latest statistical analysis (Jemal et al., 2002). In 2002, the estimated number of new pancreatic cancer cases in the US was 30 300, while the estimated number of deaths was 29 700 (Jemal et al., 2002). PDAC is characterized by late diagnosis, early recurrence after potential curative resection, and relative unresponsiveness to standard oncological therapies such as radiotherapy and chemotherapy (DiMagno et al., 1999). The overall 5-year survival rate is less than 5%, and more than 85% of the tumors are diagnosed when the tumor has infiltrated into adjacent organs or when distant metastases are present (Gudjonsson, 1987). Furthermore, even after potential curative resection, the 5-year survival rate is only around 20% in large series (Neoptolemos et al., 2001a, 2001b). Numerous studies have documented different genetic and epigenetic alterations in PDAC, including mutations of the proto-oncogene K-ras, and tumor-suppressor gene mutations such as those of p53, Smad4/DPC4 and p16/CDKN2A (Bardeesy and DePinho, 2002). In addition, overexpression of growth factors and growth factor receptors is present in a significant proportion of these tumors, and PDACs display a number of different alterations in apoptosis, invasion and metastasis-regulating pathways. These changes combine to give PDAC cells a distinct growth advantage (Kleeff et al., 1998; Koliopanos et al., 2001; Liao et al., 2001; Ozawa et al., 2001).

Deregulation of cell-cycle control is thought to be a crucial event in the malignant transformation of cells (Tsai and McKay, 2002). Alterations in cell-cycle regulatory genes are commonly found in human cancers (Sherr, 1996), and cell proliferation that reflects cell-cycle progression is an influential factor contributing to the aggressive growth behavior of cancer cells. Two critical checkpoints have been characterized, and are located at the G1–S phase and G2–M phase transitions. Central components in the cell-cycle machinery are cyclins and their corresponding cyclin-dependent kinases (CDKs). The complexes of CDK/cyclin are activated and inactivated sequentially to regulate different phases of the cell cycle. Several gene mutations are involved in deregulated cell-cycle progression in PDAC, such as mutations of the tumor-suppressor gene p53, which result in antiapoptotic responses and anti-G1–S phase arrest; mutations of K-ras, which result in continued cell proliferation and survival (Bardeesy and DePinho, 2002); and mutations of CDKN2A, which encodes two tumor suppressors – INK4a and ARF – that disrupt retinoblastoma (RB) and p53 pathways, subsequently resulting in uncontrolled cell-cycle progression (Lal et al., 2000).

CDC25 phosphatases, consisting of CDC25A, B and C, are a novel family of CDK phosphatases, which activate CDK/cyclin complexes by dephosphorylation of CDKs at different checkpoints of the cell cycle (Galaktionov and Beach, 1991; Sebastian et al., 1993). CDC25A is expressed in the late G1 phase and controls G1/S transition by activating the CDK2/Cyclin E complex (Hoffmann et al., 1994), while CDC25B and C are essential for G2 and G2/M transition (Hoffmann et al., 1993; Lammer et al., 1998). CDC25A and CDC25B, but not CDC25C, exhibit oncogenic potential since they can transform primary murine fibroblasts in cooperation with either activated H-ras or loss of RB1 (Galaktionov et al., 1995). Overexpression of CDC25A or CDC25B has been reported in different human tumors, including non-Hodgkin's lymphoma, neuroblastoma, head and neck cancer, lung cancer, esophageal cancer, gastric cancer, colorectal cancer and prostate cancer (Gasparotto et al., 1997; Kudo et al., 1997; Hernandez et al., 2000, 2001; Miyata et al., 2001; Sasaki et al., 2001; Sato et al., 2001; Ito et al., 2002; Ngan et al., 2003). The expression and functional role of CDC25B phosphatases in PDAC has so far not been analysed. Therefore, in the present study we investigated the expression of CDC25s in PDAC in order to elucidate the role of these genes as modulators of cell-cycle progression in this disease. In addition, CDC25B inhibitors were used to study the functional role of these phosphatases in the pathogenesis of PDAC.

Results

CDC25B expression in human pancreatic cancer

We first screened the expression of the three CDC25 family members (CDC25A, B and C) in pancreatic cancer, chronic pancreatitis and the normal pancreas utilizing DNA arrays. This analysis revealed that CDC25B, but not CDC25A or CDC25C, was significantly overexpressed in pancreatic cancer compared to either chronic pancreatitis or normal pancreatic tissues (Table 1a). Thus, CDC25B mRNA expression levels were increased 1.4- and 1.9-fold in pancreatic cancer compared with chronic pancreatitis and the normal pancreas, respectively. Furthermore, in metastatic lesions, CDC25B levels were increased 2.2- and 4.1-fold in comparison with primary cancer and normal pancreatic samples, respectively. In contrast, there was no significant change in CDC25A and CDC25C mRNA levels between pancreatic cancer, chronic pancreatitis and normal pancreatic samples.

Table 1a Expression of CDC25 family members in the normal pancreas, chronic pancreatitis and pancreatic cancer as determined by DNA array

These DNA array results were confirmed by Northern blot analysis and quantitative RT–PCR. Northern blot analysis of total RNA isolated from 48 PDAC and 24 normal pancreatic tissues revealed absent or weak expression of the approximately 3.0 kb CDC25B mRNA transcript (Wu and Wolgemuth, 1995) in the normal samples. In contrast, in 40 of 48 (83%) PDAC samples, the CDC25B mRNA transcript was clearly detectable. In 26 (54%) of these samples there was strong expression, whereas in 14 (29%) samples there was low to moderate expression of this mRNA moiety in comparison to normal pancreatic tissues. An example of a Northern blot depicting eight PDAC and four normal tissue samples is shown in Figure 1a (upper panel). Previously, different splicing variants have been described for CDC25B (Kudo et al., 1997). Therefore, we also characterized splice variants of CDC25B in cancerous and normal pancreatic tissues by RT–PCR using a pair of primers that selectively amplifies four variants yielding fragments of 627 bp (CDC25B4), 522 bp (CDC25B3), 480 bp (CDC25B1) and 399 bp (CDC25B2). The CDC25B3 isoform was detectable in virtually all cases, but at relatively higher levels in cancerous than in normal tissues. The other isoforms were below the level of detection (Figure 1a, lower panel). It is currently not known whether these splice variants are present only at low levels in pancreatic tissues, or alternatively whether there are other less abundant splice variants.

Figure 1
figure1

CDC25B expression analysis in pancreatic tissues and cell lines. Total RNA (15 μg) isolated from the indicated pancreatic tissues (a, upper panel) and cell lines (c, upper panel) was subjected to Northern blot analysis and hybridized with the labeled CDC25B cDNA probe and rehybridized with the 7S cDNA probe to verify equivalent RNA loading as described in the Materials and methods section. Size markers are indicated on the left. RT–PCR analysis of CDC25B splice variants in seven normal and seven cancerous tissues (a, lower panel) as well as seven pancreatic cancer cell lines (c, lower panel) as described in the Materials and methods section. The 522 and 399 bp RT–PCR products representing the CDC25B3 and CDC25B2 splice forms, respectively, are indicated on the right. Size markers are indicated on the left. Quantitative PCR for CDC25B and CDC25B3 (b) in the normal pancreas, chronic pancreatitis and pancreatic cancer tissues was carried out as described in the Materials and methods section. Results are presented as mean±s.e.m. Immunoblot analysis (d) of CDC25B in pancreatic cancer cell lines. Protein lysates (20 μg) were subjected to immunoblot analysis using an anti-CDC25B antibody as described in the Materials and methods section. The molecular size markers (in kDa) are indicated on the left

Changes in CDC25B mRNA levels were quantified by real-time QRT–PCR. In pancreatic cancer samples, the median expression level of CDC25B mRNA was 7.5- and 3.9-fold higher than in the normal and chronic pancreatitis tissue samples, respectively (Table 1b). The expression of the CDC25B3 splicing variant displayed a similar pattern. Thus, CDC25B3 mRNA levels were 6.3- and 3.2-fold increased in PDAC in comparison with the normal pancreas and chronic pancreatitis, respectively (Figure 1b). This analysis also revealed that the CDC25B3 isoform accounts for approximately 43% of total CDC25B mRNA expression in pancreatic cancer samples, and approximately 50% in both CP and normal pancreatic samples (Figure 1b). Since different primers might result in distinct efficiencies of the QRT–PCR, these numbers represent only an estimate of the relative abundance of the CDC25B3 splice variant. Nonetheless, our RT–PCR as well as QRT–PCR data both point to a predominance of the CDC25B3 splice variant in pancreatic tissues.

Table 1b Expression of CDC25B and the B3 splice variant in the normal pancreas, chronic pancreatitis and pancreatic cancer as determined by quantitative RT–PCR

CDC25B expression in human pancreatic cancer cell lines

The expression of CDC25B was next analysed in seven human pancreatic cancer cell lines (Figure 1c). Six pancreatic cancer cell lines exhibited high levels of CDC25B mRNA expression, whereas BxPc-3 cells exhibited weak CDC25B expression. RT–PCR analysis revealed predominant expression of the CDC25B3 splice variant in six cell lines. The CDC25B2 splice variant was also detectable, but at lower levels (Figure 1c). Western blot results generally confirmed the mRNA expression data obtained by Northern blot analysis. Thus, Mia-PaCa-2 and PANC-1 had the highest expression levels of CDC25B, while BxPc-3 exhibited absent to weak expression. In contrast, in Capan-1 cells, CDC25B mRNA was present at a moderate level, while protein levels were below the level of detection by Western blot analysis (Figure 1c).

Regulation of CDC25B expression in human pancreatic cancer cell lines

To determine the mechanisms of increased expression of CDC25B in pancreatic cancer cell lines, we next determined the effects of TGF-β on CDC25B mRNA levels, since it has been shown that CDC25B can be transcriptionally regulated by TGF-β (Ulloa and Tabibzadeh, 2001). QRT–PCR analysis revealed that in the TGF-β-responsive cell line Colo-357 (Kleeff and Korc, 1998), TGF-β led to a marked but short-term reduction of CDC25B mRNA expression (Figure 2a), whereas there was a sustained increase in the mRNA expression of the known TGF-β target gene, the CDK inhibitor p21 (Figure 2b) (Kleeff and Korc, 1998). Similarly, in Panc-1 pancreatic cancer cells, which are also growth inhibited by TGF-β although to a lesser extent when compared to Colo-357 cells (Kleeff et al., 1999c), TGF-β reduced CDC25B mRNA levels (Figure 2a) and increased p21 mRNA levels (Figure 2b). In the other pancreatic cancer cell lines that are resistant to the growth inhibitory effects of TGF-β due to Smad4 mutations/deletions (Capan-1, BxPc-3) (Schutte et al., 1996; Kleeff et al., 1999c), or absent TGF-β receptor type II expression (Mia-PaCa-2) (Freeman et al., 1995), TGF-β did not alter CDC25B mRNA expression levels (data not shown).

Figure 2
figure2

Regulation of CDC25B mRNA expression by TGF-β. Colo-357 (solid circle) and Panc-1 (open triangle) cell lines were incubated in the absence (0) or presence of 200 pM TGF-β1 for the indicated time. RNA was extracted as described and QRT-PCR for CDC25B (a) and p21CIP1 (b) was performed. The respective control mRNA expression levels were set to 100%. Results are presented as mean±s.e.m. of three independent experiments

CDC25B localization in human pancreatic tissues

To delineate the significance of CDC25B overexpression in PDAC further, immunohistochemistry was carried out next. In the normal pancreas, nuclear CDC25B staining was observed in some ductal, acinar and islet cells. Calculation of the percentage of positive cells revealed that in the normal pancreas, 8.3±1.8% of cells were CDC25B positive (Figure 3a, b). In contrast, in PDAC, 48.6±16.3% of the cancer cells displayed positive nuclear CDC25B immunoreactivity (Figure 3f–h). Interestingly, in addition to the nuclear localization of CDC25B, some cancer cells also exhibited cytoplasmic staining (Figure 3h). In pancreatic intraepithelial neoplasia (PanIN) lesions that were present in some cancer samples, a comparable percentage of CDC25B-positive cells could be observed (Figure 3c). In addition, in areas surrounding the tumor, consisting of tubular complexes, morphological normal as well as dedifferentiating acinar cells, 20.8±7.5% of these cells displayed positive nuclear CDC25B staining (Figure 3d, e). Fibroblasts in the tissue surrounding the tumor also displayed CDC25B immunoreactivity in the majority of cases (Figure 3h).

Figure 3
figure3

CDC25B localization in pancreatic tissues. Immunohistochemistry using a CDC25B-specific antibody was carried out in the normal pancreas (a, b), pancreatic cancer (c–h) and liver metastasis (i). C depicts a PanIN-Ia lesion observed in a PDAC sample. D and E depict areas surrounding the tumor mass containing degenerating acinar cells as well as tubular complexes. The inset in (i) depicts normal liver tissue next to the metastatic lesion. Percentage of positive cells in normal, next to cancer, cancer, metastasis as well as in pancreatic cancer according to stage and grade (k). Data are expressed as mean±s.e.m.

We also examined the expression of CDC25B in liver and lymph node metastases of PDAC samples. Approximately 71.7±3.1% of the tumor cells in the metastatic lesions were CDC25B positive (Figure 3i). In contrast, liver or lymph node tissues only occasionally exhibited CDC25B-positive cells (Figure 3i, inset). Statistical analysis revealed a significant difference of the percentage of CDC25B-positive cell between normal tissues, areas next to cancer, pancreatic cancer and metastatic lesions. In contrast, there was no difference in the percentage of CDC25B-positive cells between stage I/II and III/IV tumors, between grade 1/2 and 3 PDACs (Figure 3k) as well as between lymph node-negative and -positive tumors. In addition, there was no correlation between the percentage of CDC25B-positive cells and patient age and gender.

Functional role of CDC25B in human pancreatic cancer cells

To assess the functional role of CDC25B further, two CDC25B-specific inhibitors were utilized (Lazo et al., 2001). These inhibitors exerted a dose-dependent growth-inhibitory effect in six of seven pancreatic cancer cell lines (Figure 4a). In BxPc-3 cells, there was a slight effect at low concentrations that did not increase with higher dosages. Calculation of the growth inhibition of 50% value (GI50) revealed GI50 values of 2.0±0.5 μ M to 5.4±0.8 μ M for NSC663284 and 1.6±0.4 μ M to 6.2±1.8 μ M for NSC668394 in the six responsive cell lines. In contrast, GI50 values for both inhibitors were more than 10 μ M in BxPc-3 cells, which exhibited the lowest expression levels of CDC25B (Table 2a).

Figure 4
figure4

Effects of CDC25B inhibitors in human pancreatic cancer cells. Pancreatic cancer cells (5000/well) were seeded in 96-well plates and incubated for 48 h with different concentrations of the CDC25B inhibitors. Cell growth was determined by the MTT assay (a). Data are expressed as mean±s.e.m. of triplicate determinations of three independent experiments. (b) Exemplary FACS analysis of Mia-PaCa-2 (left) and BxPc-3 (right). Measurement of DNA content for 20 000 cells in each sample was performed and the DNA histogram was interpreted using CellQuest Pro software as described in the Materials and methods section. Control (violet), NSC663284 (pink) and NSC668394 (green). (c, d) Two pancreatic cancer cell lines, BxPc-3 and Mia-PC-2, which showed the lowest and highest expression levels of CDC25B, respectively, were treated with NSC663284 and NSC668394 at the GI50 dose for the indicated time. Immunoblot analysis was carried out with the anti-phospho-cdc2 and the control cdc2 antibody as described in the Materials and methods section. Representative blots from two independent experiments are shown

Table 2a Summary of GI50 values for two specific CDC25B inhibitors in cultured human pancreatic cancer lines

To determine the effects of CDC25B inhibitors on cell-cycle progression in pancreatic cancer cells, FACS analysis was carried out next. Cells were treated with the GI50 dose of the respective CDC25B inhibitor. After 48 h, the percentage of cells in the G0/G1, S and G2/M phases was determined by portioning the DNA histogram based on the control G1 and G2/M peaks. The results revealed a shift to the G2 phase and G2/M accumulation in Capan-I, Colo-357, Mia-PaCa-2, Panc-1 and T3M4 (Table 2b). In Mia-PaCa-2 cells, for example, the population of G2/M increased about 130% following addition of the CDC25B inhibitor NSC668394. In contrast, in BxPc-3 cells, no significant change in cell-cycle distribution was observed after incubation with both CDC25B inhibitors (Table 2b). An exemplary FACS analysis of Mia-PaCa-2 and BxPc-3 cells is shown in Figure 4b.

Table 2b Changes in G2/M population in human pancreatic cancer cell lines induced by CDC25B inhibitors

To determine further the mechanism of G2/M arrest induced by CDC25B inhibitors, we investigated the phosphorylation of the CDC25B substrate cdc2. In Mia-PaCa-2 cells, the level of phosphorylated cdc2 increased time dependently following addition of the CDC25B inhibitors, with maximal effects occurring after 12 h of incubation (Figure 4c). In contrast, in BxPc-3 cells there was no difference in CDC2 phosphorylation following addition of the CDC25B inhibitors (Figure 4d).

Discussion

The eucaryotic cell cycle is controlled by a family of protein kinases named CDKs, whose activation and inactivation is regulated by CDK inhibitors (e.g., p16) and activators (e.g., CDC25s). The G1–S phase and the G2–M phase checkpoints of the cell cycle are precisely controlled by numerous protein–protein interactions such as between the 14-3-3 protein, Cdk2 and chk-1 and others (van Hemert et al., 2001). Cell-cycle deregulation can lead to uncontrolled growth and contributes to tumor formation. Defects in a variety of molecules that regulate the cell cycle such as p53, cdk inhibitors (e.g., p15, p16, p18, p21 and p27) and Rb have been implicated in tumorigenesis.

CDC25s have been characterized as dual-specificity phosphatases that activate cyclin/CDKs complexes at different points of the cell cycle (Galaktionov and Beach, 1991; Sebastian et al., 1993). Overexpression of CDC25A and CDC25B has been documented in many human malignancies, suggesting their involvement in the pathogenesis and progression of malignant transformation (Gasparotto et al., 1997; Kudo et al., 1997; Hernandez et al., 2000, 2001; Ito et al., 2002; Miyata et al., 2001; Sasaki et al., 2001; Sato et al., 2001; Ngan et al., 2003). PDAC is characterized by uncontrolled cell-cycle progression, resistance to cell-cycle arrest and apoptosis, in part due to alterations involving mutations of the proto-oncogene K-ras and p53, Smad4/DPC4 and p16/CDKN2A tumor-suppressor gene mutations (Ozawa et al., 2001; Bardeesy and DePinho, 2002).

In the present study, we have investigated the expression and functional role of CDC25B in PDAC. At the mRNA level, cDNA array revealed that CDC25B expression levels were increased in pancreatic cancer compared with chronic pancreatitis and the normal pancreas, especially in metastatic lesions, in which CDC25B mRNA levels were 2.2- and 4.1-fold increased in comparison to primary cancer and normal pancreatic samples, respectively. In contrast, CDC25A and CDC25C showed no significant difference between normal and cancerous samples. These results are in agreement with previous observations in other tumors. For example, in gastric cancer, CDC25B is expressed at high levels compared to both CDC25A and CDC25C (Kudo et al., 1997). Quantitative PCR confirmed these findings, demonstrating that CDC25B mRNA levels in pancreatic cancer samples were increased compared to the normal pancreas and chronic pancreatitis.

Interestingly, TGF-β led to transcriptional repression of CDC25B mRNA expression in two TGF-β-responsive pancreatic cancer cell lines, but was without effect in three other pancreatic cancer cell lines with Smad4 mutations or other alterations in the TGF-β signaling pathway. Since pancreatic cancers in vivo frequently exhibit alterations of the TGF-β signaling pathway such as Smad4 mutations (Hahn et al., 1996), low expression of the type I TGF-β receptor (Wagner et al., 1998) or Smad6/Smad7 overexpression (Kleeff et al., 1999a, 1999b), these results provide another evidence of how these tumors might benefit from disturbances in the TGF-β signaling pathway, that is, by becoming resistant to the repressive effects of TGF-β on CDC25B expression.

Immunohistochemistry revealed that in PDAC, 48.6±16.3% of the cancer cells displayed positive nuclear CDC25B immunoreactivity, whereas in the normal pancreas only 8.3±1.8% cells exhibited positive nuclear staining. In addition to the nuclear localization of CDC25B, some cancer cells also exhibited cytoplasmic staining. There are conflicting reports regarding the localization of CDC25B in other tissues and cells. Ito and coworkers as well as other research groups have reported cytoplasmic accumulation of CDC25B (Gabrielli et al., 1996; Gasparotto et al., 1997; Broggini et al., 2000; Takemasa et al., 2000; Hernandez et al., 2001; Sato et al., 2001). Karlsson et al. (1999) reported that CDC25B is located in the nucleus in the G1 phase and exported to the cytoplasm during the S phase. Davezac et al. (2000) demonstrated that the localization of CDC25B in the cytoplasm or in the nucleus depends on the combined effects of a nuclear localization signal, a nuclear export signal and on the interaction with 14-3-3 proteins. This difference in the observed localization in pancreatic tissues might be due to different, tissue-specific transportation and trafficking of the protein. The functional consequences of this observation, however, are currently not known.

In the present study, the percentage of CDC25B-positive cells was significantly higher in PDAC and metastatic lesions in comparison to the normal pancreas (P<0.01). However, there was no correlation between the percentage of CDC25B-positive cells and stage or grade of the tumor. In addition, PanIN lesions exhibited a comparable percentage of CDC25B-positive cells, suggesting that overexpression of CDC25B is an early event in the pathogenesis of PDAC. In line with this hypothesis, in areas surrounding the tumor, the percentage of CDC25B-positive cells was significantly increased in comparison to normal pancreatic tissues. Although a correlation between CDC25B expression and stage or grade of the tumor has been observed in several human malignances such as non-small-cell lung carcinoma, ovarian cancer, breast cancer and colorectal cancer (Galaktionov et al., 1995; Broggini et al., 2000; Takemasa et al., 2000; Sasaki et al., 2001), we could not detect such a correlation. Our results instead suggest that in PDAC, increased CDC25B expression occurs early in the process of malignant transformation.

Recent studies have shown that the role of CDC25B in tumor pathogenesis and progression is dependent on post-transcriptional regulation. One important post-transcriptional alteration is splicing, and five CDC25B splicing variants have been identified so far. The CDC25B3 splice variant is the main isoform in normal and cancerous colorectal tissues. In contrast, CDC25B2 seems to be important in the pathogenesis of non-Hodgkin's lymphomas (Kudo et al., 1997; Hernandez et al., 2000), and CDC25B1 in regulating the initiation of the mammalian mitosis (Baldin et al., 2002). In the present study, the CDC25B3 isoform was predominant in normal as well as cancerous tissues, suggesting that this isoform, and not other splice variants, plays an important role under physiological and pathophysiological conditions in the pancreas.

Since CDC25s are important components of the cell-cycle machinery and since their expression is frequently altered in human malignancies, efforts have been made to inhibit their activity specifically. Recently, novel CDC25B inhibitors were utilized to inhibit tumor cell growth (Lazo et al., 2001). In our study, two CDC25B inhibitors – NSC663284 and NSC668394 – dose dependently inhibited pancreatic cancer cell growth. Three lines of evidence suggest that their effects could specifically be attributed to CDC25B inhibition. First, FACS analysis revealed that the two compounds resulted in G2/M arrest and G2 accumulation. Second, the levels of phosphorylated cdc2 increased time dependently after addition of CDC25B inhibitors, suggesting that these inhibitors indeed blocked CDC25B phosphatase activity. Third, Mia-PaCa-2 cells that displayed the highest expression levels of CDC25B were most sensitive to the CDC25B inhibitors, whereas BxPc-3 cells that expressed only low levels of CDC25B were resistant to the inhibitors.

In conclusion, our findings demonstrate that CDC25B is overexpressed in PDAC and metastatic lesions, providing a novel aspect in the pathogenesis of this disease. CDC25B inhibitors act to inhibit pancreatic cancer cell growth by blocking G2/M phase transition via inhibition of dephosphorylation of cdc2. Together, our findings raise the possibility that inhibition of CDC25B phosphatase may ultimately have a therapeutic role in this disorder.

Materials and methods

Patients and tissue samples

Samples from 48 human primary PDACs (30 male, 18 female; median age 66 years; range 38–84 years), 10 cases of liver and/or lymph node metastasis (seven males, three females; median age 64 years; range 48–83 years), and samples of 34 chronic pancreatitis (28 males, six females; median age 50 years; range 16–64 years) were collected from patients who underwent pancreatic resection in the Department of Visceral and Transplantation Surgery (University of Bern, Switzerland) and the Department of General Surgery (University of Heidelberg, Germany). In all, 24 normal human pancreatic tissue samples were obtained from previously healthy organ donors (14 males, 10 females; median age 42 years; range 14–73 years) through a donor program. According to the TNM classification of the Union Internationale Contre le Cancer (UICC), there were 10 stage I ductal adenocarcinomas, 10 stage II, 24 stage III and four stage IV ductal adenocarcinomas. There were five grade 1, 23 grade 2 and 20 grade 3 tumors. Freshly removed samples were snap frozen in liquid nitrogen and stored at −80°C until analysis. In addition, a portion of the tissue samples were fixed in 5% paraformaldehyde solution for 12–24 h and paraffin embedded for histological analysis. The study was approved by the Ethics Committees of the University of Bern (Switzerland) and the University of Heidelberg (Germany) and by the Human Subjects Committee of the University of California (Irvine, CA, USA).

RNA extraction

Total RNA from human pancreatic tissues and pancreatic cancer cell lines was isolated either by the single-step guanidinium thiocyanate method as described previously or MagNAPure LC mRNA isolation kit II (Roche Diagnostic GmbH, Mannheim, Germany) (Friess et al., 1993; Kleeff et al., 1998; Koliopanos et al., 2001).

cDNA synthesis

cDNA was synthesized from total RNA by reverse transcription using a cDNA kit according to the manufacturer's instructions (First Strand cDNA Synthesis Kit, Roche Diagnostic GmbH, Mannheim, Germany).

cDNA array

The GeneChip® HG-U95Av2 array used in this study was fabricated by Affymetrix Inc. (Santa Clara, CA, USA). It contains about 10 000 full-length human genes from the UniGene database (Build 95). Poly (A)+RNA isolation, cDNA synthesis and cRNA in vitro transcription was carried out as reported previously (Friess et al., 2001; Tureci et al., 2003). The in vitro transcription product was purified and fragmented as described (Friess et al., 2001; Tureci et al., 2003). Hybridization of the fragmented in vitro transcription product to oligonucleotide arrays was performed as suggested by the manufacturer (Affymetrix Inc.).

Quantitative RT–PCR

Quantitative PCR was performed with the LightCycler FastStart DNA SYBR Green kit (Roche Applied Science, Mannheim, Germany). The primer sets were designed for CDC25B (forward 5-IndexTermGGG CAA GTT CAG CAA CAT CGT GGA-3, reverse 5-IndexTermGTA GCC GCC TTT CAG GAT ATA CAT C-3) and the CDC25B3 splice variant (forward 5-IndexTermGAA ACC CCA AAG AGT CAG GTA G-3, reverse 5-IndexTermGAG TTG GTG ATG TTC CGA AGC AC-3). To control for specificity of the amplification products, a melting curve analysis was performed. The number of transcripts was calculated from a standard curve obtained by plotting known input concentrations of four different representative plasmids (log dilutions) to the PCR cycle number at which the detected fluorescence intensity reaches a fixed value (z). Data of two independent analyses for each sample and parameter were averaged. The copy number of the different CDC25B and variant B3 was normalized by the housekeeping genes HPRT and cyclophillin B (Search-LC GmbH, Heidelberg, Germany).

Conventional RT–PCR

RT–PCR was carried out under standard conditions, using the following primer pairs: forward 5-IndexTermCTC ATT AGT GCC CCA CTG GT-3 and reverse 5-IndexTermATC AGC TCT CGG TGG TCA CT-3 for CDC25B for subcloning; forward 5-IndexTermGCT TCC TCG CCG GTC ACC AC-3 and reverse 5-IndexTermCCT GCG GCT GGC CCA CTC-3 for the CDC25B for splicing variants; forward 5-IndexTermCTG GCC GGG ACC TGA CTG ACT-3 and reverse 5-IndexTermCTA GAA GCA TTT GCG GTG GAC GAT-3 for beta-actin as an internal control. Products were separated by electrophoresis in 1% agarose-TBE gels, and bands were visualized by ethidium bromide staining (24).

Subcloning

A 305 RT–PCR fragment of the human CDC25B gene was subcloned into the pGEMT-easy vector (Promega Biotechnology, Madison, WI, USA) using standard conditions. Authenticity was confirmed by sequencing (QIAGEN Genomic services, Hilden, Germany).

Northern blot analysis

Total RNA (15 μg/lane) was size fractionated on 1.2% agarose/1.8 M formaldehyde gels. Gels were stained with ethidium bromide for verification of RNA integrity and loading equivalency. Fractionated RNA was transferred onto Genescreen membranes (NEN™ Life Science Products, Boston, MA, USA) and crosslinked by UV irradiation. Blots were then prehybridized for 12 h at 42°C in 50% formamide, 1% SDS, 5 × Denhardt's, 100 μg/ml salmon sperm DNA, 50 mM Na2PO4, pH 7.4, 10% dextran, 75 mM NaCl and 5 mM EDTA. Blots were then hybridized for 24 h at 42°C in the presence of an α32P dCTP (Amersham Pharmacia Biotech, Freiburg, Germany) labeled cDNA probe for CDC25B (106 cpm/ml) or 7S (105 cpm/ml), rinsed twice with 2 × SSC and washed twice with 0.2 × SSC/2%SDS at 50 and 55°C for 20 min, respectively. All blots were exposed at −80°C to Kodak BiomaxMS films with Kodak-intensifying screens (Friess et al., 1993; Kleeff et al., 1998, 1999a).

Immunohistochemistry

Paraffin-embedded tissue sections (3 μm thick) were deparaffinized and rehydrated. Slides were covered with 3% H2O2 for 10 min and 10% normal goat serum for 30 min at room temperature before incubation overnight with a rabbit anti-CDC25B antibody (Cell Signaling Technology, Beverly, MA, USA) at 4°C. After washing with TBS+0.1%BSA, slides were incubated with an HRP-linked anti-rabbit secondary antibody (DAKO Envision Kit, Carpenteria, CA, USA) for 45 min, then washed three times with TBS+0.1%BSA, followed by incubation with a buffered substrate for liquid DAB/liquid DAB chromogen mixture for 20 s. Finally, sections were counterstained with Mayer's hematoxylin and mounted in permanent mounting medium (Friess et al., 1993; Kleeff et al., 1998). To calculate the percentage of CDC25B positive cells, 10 high-power fields were selected at random for each section, and at least 500 cells were evaluated.

Protein extraction

Cells and tissues were lysed in 0.5 ml lysis buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 2 mM EDTA (pH 8.0) and 1% SDS with proteinase inhibitors (one tablet/10 ml, Roche Molecular Biochemicals, Basel, Switzerland). Protein concentration was determined by the BCA protein assay (Pierce Chemical Co., Rockford, IL, USA).

Immunoblot

Total protein (20 μg/lane) was size fractionated by SDS–polyacrylamide gel electrophoresis (SDS/PAGE) and transferred onto nitrocellulose membranes (BioRad Laboratories, Hercules, CA, USA). Blots were blocked in TBS with Tween-20 (0.1%) at pH 7.5 containing 5% nonfat milk for 1 h and then incubated overnight with the indicated antibody (CDC25B, CDC2 and phospho-CDC2). Following three washes with TBS-T, membranes were incubated with the secondary antibody for 1 h at 23°C. Antibody detection was performed with an enhanced chemiluminescence (ECL) reaction system (Amersham Life Sciences, Amersham, UK).

Cell culture

Pancreatic cancer cell lines were cultured in RPMI1640 (AsPc-1, BxPc-3, Capan-1 and T3M4) or DMEM (Colo-357, Mia-PaCa-2 Panc-1) with 10% fetal bovine serum, 100 U/ml penicillin and 100 μg/ml streptomycin and maintained at 37°C in 5% CO2. For TGF-β induction experiments, cells were serum starved overnight. Subsequently, cells were incubated in a serum-free medium in the absence or presence of 200 pM TGF-β1 (R&D Systems Inc., Minneapolis, MN, USA) for the indicated time. After 48 h, RNA was extracted and QRT–PCR was carried out as described.

MTT assay

Two CDC25B inhibitors (NSC663284 and NSC668394) were the kind gift of Robert J Schultz (Drug Synthesis & Chemistry Branch, Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute). Both compounds were identified from 140 000 compounds in the NCI Compound Repository, as specific inhibitors of CDC25B (Lazo et al., 2001). Determination of the GI50 was carried out as described (http://www.dtp.nci.nih.gov/). Briefly, 5 × 103 cells/well were seeded in 96-well plates. After incubation for 48 h with different concentrations of NSC663284 or NSC668394, cell growth was determined by the MTT assay. GI50 was calculated from [(Ti−Tz)/(C−Tz)] × 100=50 (Tz: time zero, C: control growth, and Ti: test growth in the presence of drug at different concentration levels).

Cell-cycle analysis by flow cytometry

Cells seeded in 12-well plates (1 × 105 cells/well) were incubated for 48 h at the GI50 concentration. Cells were next harvested and washed with PBS, then resuspended in 1 ml of hypotonic PI buffer: 50 μg/ml PI, 0.1% Triton X-100 and 0.1% NaCitrate) and stored overnight at 4°C (Nicoletti et al., 1991). Measurement of DNA content for 20 000 cells in each sample was performed using FACScan (Becton Dickinson and Company, New Jersey, USA).

Statistics

Patient data are expressed as median and range. Experimental results are expressed as mean±s.e.m. of at least three independent experiments, unless indicated otherwise. For statistical analysis, the Mann–Whitney U-test was used. P<0.05 was taken as the level of significance.

References

  1. Baldin V, Pelpel K, Cazales M, Cans C and Ducommun B . (2002). J. Biol. Chem., 277, 35176–35182.

  2. Bardeesy N and DePinho RA . (2002). Nat. Rev. Cancer, 2, 897–909.

  3. Broggini M, Buraggi G, Brenna A, Riva L, Codegoni AM, Torri V, Lissoni AA, Mangioni C and D'Incalci M . (2000). Anticancer Res., 20, 4835–4840.

  4. Davezac N, Baldin V, Gabrielli B, Forrest A, Theis-Febvre N, Yashida M and Ducommun B . (2000). Oncogene, 19, 2179–2185.

  5. DiMagno EP, Reber HA and Tempero MA . (1999). Gastroenterology, 117, 1464–1484.

  6. Freeman JW, Mattingly CA and Strodel WE . (1995). J. Cell Physiol., 165, 155–163.

  7. Friess H, Ding J, Kleeff J, Liao Q, Berberat PO, Hammer J and Buchler MW . (2001). Ann. Surg., 234, 769–778 discussion 778–779..

  8. Friess H, Yamanaka Y, Buchler M, Ebert M, Beger HG, Gold LI and Korc M . (1993). Gastroenterology, 105, 1846–1856.

  9. Gabrielli BG, De Souza CP, Tonks ID, Clark JM, Hayward NK and Ellem KA . (1996). J. Cell Sci., 109 (Part 5), 1081–1093.

  10. Galaktionov K and Beach D . (1991). Cell, 67, 1181–1194.

  11. Galaktionov K, Lee AK, Eckstein J, Draetta G, Meckler J, Loda M and Beach D . (1995). Science, 269, 1575–1577.

  12. Gasparotto D, Maestro R, Piccinin S, Vukosavljevic T, Barzan L, Sulfaro S and Boiocchi M . (1997). Cancer Res., 57, 2366–2368.

  13. Gudjonsson B . (1987). Cancer, 60, 2284–2303.

  14. Hahn SA, Schutte M, Hoque AT, Moskaluk CA, da Costa LT, Rozenblum E, Weinstein CL, Fischer A, Yeo CJ, Hruban RH and Kern SE . (1996). Science, 271, 350–353.

  15. Hernandez S, Bessa X, Bea S, Hernandez L, Nadal A, Mallofre C, Muntane J, Castells A, Fernandez PL, Cardesa A and Campo E . (2001). Lab. Invest., 81, 465–473.

  16. Hernandez S, Hernandez L, Bea S, Pinyol M, Nayach I, Bellosillo B, Nadal A, Ferrer A, Fernandez PL, Montserrat E, Cardesa A and Campo E . (2000). Int. J. Cancer, 89, 148–152.

  17. Hoffmann I, Clarke PR, Marcote MJ, Karsenti E and Draetta G . (1993). EMBO J., 12, 53–63.

  18. Hoffmann I, Draetta G and Karsenti E . (1994). EMBO J., 13, 4302–4310.

  19. Ito Y, Yoshida H, Nakano K, Kobayashi K, Yokozawa T, Hirai K, Matsuzuka F, Matsuura N, Kakudo K, Kuma K and Miyauchi A . (2002). Br. J. Cancer, 86, 1909–1913.

  20. Jemal A, Thomas A, Murray T and Thun M . (2002). CA Cancer J. Clin., 52, 23–47.

  21. Karlsson C, Katich S, Hagting A, Hoffmann I and Pines J . (1999). J. Cell Biol., 146, 573–584.

  22. Kleeff J, Ishiwata T, Kumbasar A, Friess H, Buchler MW, Lander AD and Korc M . (1998). J. Clin. Invest., 102, 1662–1673.

  23. Kleeff J, Ishiwata T, Maruyama H, Friess H, Truong P, Buchler MW, Falb D and Korc M . (1999a). Oncogene, 18, 5363–5372.

  24. Kleeff J and Korc M . (1998). J. Biol. Chem., 273, 7495–7500.

  25. Kleeff J, Maruyama H, Friess H, Buchler MW, Falb D and Korc M . (1999b). Biochem. Biophys. Res. Commun., 255, 268–273.

  26. Kleeff J, Wildi S, Friess H and Korc M . (1999c). Pancreas, 18, 364–370.

  27. Koliopanos A, Friess H, Kleeff J, Shi X, Liao Q, Pecker I, Vlodavsky I, Zimmermann A and Buchler MW . (2001). Cancer Res., 61, 4655–4659.

  28. Kudo Y, Yasui W, Ue T, Yamamoto S, Yokozaki H, Nikai H and Tahara E . (1997). Jpn. J. Cancer Res., 88, 947–952.

  29. Lal G, Liu L, Hogg D, Lassam NJ, Redston MS and Gallinger S . (2000). Genes Chromosomes Cancer, 27, 358–361.

  30. Lammer C, Wagerer S, Saffrich R, Mertens D, Ansorge W and Hoffmann I . (1998). J. Cell Sci., 111 (Part 16), 2445–2453.

  31. Lazo JS, Aslan DC, Southwick EC, Cooley KA, Ducruet AP, Joo B, Vogt A and Wipf P . (2001). J. Med. Chem., 44, 4042–4049.

  32. Liao Q, Ozawa F, Friess H, Zimmermann A, Takayama S, Reed JC, Kleeff J and Buchler MW . (2001). FEBS Lett., 503, 151–157.

  33. Miyata H, Doki Y, Yamamoto H, Kishi K, Takemoto H, Fujiwara Y, Yasuda T, Yano M, Inoue M, Shiozaki H, Weinstein IB and Monden M . (2001). Cancer Res., 61, 3188–3193.

  34. Neoptolemos JP, Dunn JA, Stocken DD, Almond J, Link K, Beger H, Bassi C, Falconi M, Pederzoli P, Dervenis C, Fernandez-Cruz L, Lacaine F, Pap A, Spooner D, Kerr DJ, Friess H and Buchler MW . (2001a). Lancet, 358, 1576–1585.

  35. Neoptolemos JP, Stocken DD, Dunn JA, Almond J, Beger HG, Pederzoli P, Bassi C, Dervenis C, Fernandez-Cruz L, Lacaine F, Buckels J, Deakin M, Adab FA, Sutton R, Imrie C, Ihse I, Tihanyi T, Olah A, Pedrazzoli S, Spooner D, Kerr DJ, Friess H and Buchler MW . (2001b). Ann. Surg., 234, 758–768.

  36. Ngan ES, Hashimoto Y, Ma ZQ, Tsai MJ and Tsai SY . (2003). Oncogene, 22, 734–739.

  37. Nicoletti I, Migliorati G, Pagliacci MC, Grignani F and Riccardi C . (1991). J. Immunol. Methods, 139, 271–279.

  38. Ozawa F, Friess H, Tempia-Caliera A, Kleeff J and Buchler MW . (2001). Teratog. Carcinog. Mutagen., 21, 27–44.

  39. Sasaki H, Yukiue H, Kobayashi Y, Tanahashi M, Moriyama S, Nakashima Y, Fukai I, Kiriyama M, Yamakawa Y and Fujii Y . (2001). Cancer Lett., 173, 187–192.

  40. Sato Y, Sasaki H, Kondo S, Fukai I, Kiriyama M, Yamakawa Y and Fujii Y . (2001). Jpn. J. Clin. Oncol., 31, 428–431.

  41. Schutte M, Hruban RH, Hedrick L, Cho KR, Nadasdy GM, Weinstein CL, Bova GS, Isaacs WB, Cairns P, Nawroz H, Sidransky D, Casero Jr RA, Meltzer PS, Hahn SA and Kern SE . (1996). Cancer Res., 56, 2527–2530.

  42. Sebastian B, Kakizuka A and Hunter T . (1993). Proc. Natl. Acad. Sci. USA, 90, 3521–3524.

  43. Sherr CJ . (1996). Science, 274, 1672–1677.

  44. Takemasa I, Yamamoto H, Sekimoto M, Ohue M, Noura S, Miyake Y, Matsumoto T, Aihara T, Tomita N, Tamaki Y, Sakita I, Kikkawa N, Matsuura N, Shiozaki H and Monden M . (2000). Cancer Res., 60, 3043–3050.

  45. Tsai RY and McKay RD . (2002). Genes Dev., 16, 2991–3003.

  46. Tureci O, Ding J, Hilton H, Bian H, Ohkawa H, Braxenthaler M, Seitz G, Raddrizzani L, Friess H, Buchler M, Sahin U and Hammer J . (2003). FASEB J., 17, 376–385.

  47. Ulloa L and Tabibzadeh S . (2001). J. Biol. Chem., 276, 21397–21404.

  48. van Hemert MJ, Steensma HY and van Heusden GP . (2001). Bioessays, 23, 936–946.

  49. Wagner M, Kleeff J, Lopez ME, Bockman I, Massaque J and Korc M . (1998). Int. J. Cancer, 78, 255–260.

  50. Wu S and Wolgemuth DJ . (1995). Dev. Biol., 170, 195–206.

Download references

Acknowledgements

This work was supported, in part, by US Public Health Service Grant CA-40162 to M Korc.

Author information

Correspondence to Helmut Friess.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • cell cycle, cyclin-dependent kinase, CDC25B, pancreatic ductal adenocarcinoma

Further reading