RT-PCR revealed that two pancreatic cancer cell lines, PANC-1 and SW1990, expressed c-kit mRNA (Fig. 1A). By Western blot analysis, c-kit protein was also present in those lines (Fig. 1B). Neither c-kit mRNA nor protein was detected in the other three pancreatic cancer cell lines, BxPC-3, Capan-2 and MIA PaCa-2. Thus, Western blot analysis confirms the PT-PCR results.

Figure 2 shows the effect of SCF on proliferation of KIT-positive and KIT-negative pancreatic cancer cell lines. In the two KIT-positive cell lines (PANC-1 and SW1990), proliferation was significantly enhanced by KIT ligand SCF when the concentration exceeded 1 ng/mL (Fig. 2A). In contrast, SCF did not enhance proliferation in the three KIT-negative lines (BxPC-3, Capan-2 and MIA PaCa-2) (Fig. 2B).

Figures 3A and 3B show the effect of SCF on cell invasiveness in the five pancreatic cancer cell lines. SCF enhanced the invasive ability in the two KIT-positive cell lines (Fig. 3A). On the other hand, the invasive potential was not enhanced by SCF in the three KIT-negative cancer cell lines (Fig. 3B).

We asked whether imatinib mesylate might inhibit KIT-stimulated proliferative and invasive activities of the two KIT-positive pancreatic cancer cell lines (PANC-1 and SW1990). As shown in Fig. 4, 5 μM imatinib mesylate significantly inhibited SCF-enhanced proliferation to the level shown by nonstimulated control. Similarly, SCF-enhanced invasive ability was significantly inhibited by 5 μM imatinib mesylate treatment of PANC-1 and SW1990 (Fig. 5).

KIT and SCF expression were evaluated in 42 patients with confirmed invasive ductal carcinoma of the pancreas. Fig. 6A shows one example of a KIT-positive pancreatic cancer cells and Fig. 6B shows one example of a SCF-positive pancreatic cancer cells. Table 1 shows the clinical and pathological characteristics of the patients. Pancreatic cancer cells expressed KIT in 16 of the 42 cases (38.1%). Characteristics of the patients, such as gender, age, location of the tumor, TNM stage and cancer cell differentiation, were not significantly different between the KIT-positive and KIT-negative groups. In lymph node metastases, lymphatic system invasion and intrapancreatic neural invasion, there were no significant differences between the two groups. But, the expression of KIT was correlated with venous system invasion (p = 0.046). Pancreatic cancer cells expressed SCF in fourteen KIT-positive specimens. In fact, fourteen specimens co-expressed KIT and SCF. Furthermore, KIT-positive specimens expressed SCF more frequently compared to KIT-negative specimens (p = 0.002). Concerning patient survival, KIT expression was not significantly correlated (Fig. 7A), but the cases co-expressing KIT and SCF had a tendency toward lower survival than others (Fig 7B). However, the difference was not statistically significant. The median follow-up time was 14.8 months after surgery.

Five human pancreatic cancer cell lines (SW1990, PANC-1, MIA PaCa-2, Capan2 and BxPC3) were obtained from the American Type Culture Collection (Rockville, MD). SW1990, PANC-1 and MIA PaCa-2 were maintained in Dulbecco's modified Eagle's medium (Sigma Chemical Co., St. Louis, MO) supplemented with 10% fetal bovine serum. Capan2 was maintained in McCoy's medium supplemented with 10% fetal bovine serum. BxPC3 was maintained in RPMI-1640 medium (Sigma Chemical Co., St. Louis, Mo) supplemented with 10% fetal bovine serum. All cells were incubated at 37°C in a humidified atmosphere of 5% CO₂ in air.

Recombinant human SCF was purchased from R&D Systems (Abingdon, UK). For Western blot analysis, rabbit polyclonal anti-KIT antibody was obtained from Med. & Biological Laboratories (Nagoya, Japan) and rabbit monoclonal anti-β-actin antibody was obtained from Cell Signaling Technology (Beverly, MA). For immunohistochemistry, rabbit polyclonal anti-KIT antibody and rabbit polyclonal anti-SCF antibody were purchased from Immuno-Biological Laboratories Co., Ltd. (Gunma, Japan). KIT inhibitor imatinib mesylate was kindly provided by Novartis Pharma AG (Basel, Switzerland).

Total RNA was extracted from five pancreatic cancer cell lines using Isogen Kits (Nippon Gene Tokyo, Japan), and quantities determined spectrophotometrically. Total RNA aliquots (1 μg) were pretreated with Dnase I (Boehringer Mannheim, Germany) for 20 minutes at room temperature, denatured at 70°C for 10 minutes, chilled on ice, and then reverse-transcribed into cDNA in a reaction mixture containing 10 mmol/L dithiothreitol, 0.5 mmol/L dNTPs, first standard buffer, and 1U Superscript II (Invitrogen, San Diego, CA) at 42°C for 60 minutes. Reactions were terminated by heating at 72°C for 10 minutes. Reaction mixture aliquots (1 μL) were used as templates for PCR. The following primers for c-kit were used: forward 5'-gcccacaatagattggtattt-3' and reverse 5'-agcatctttacagcgacagtc-3'. The PCR product size was 570 bp. The PCR conditions were as follows: denaturation at 94°C for 4 minutes, annealing at 51°C for 1 minute, extension at 72°C for 1 minute, followed by 34 cycles of denaturation at 94°C for 1 minute, annealing at 51°C for 1 minute, and extension at 72°C for 1 minute. The final extension was at 72°C for 7 minutes. The PCR products were separated and detected using ethidium-bromide stained 1.5% agarose gels.

Total cell lysates from confluent cultures were prepared using ice-cold radioimmunoprecipitation assay (RIPA) buffer containing leupeptin, aprotinin, and pepstein A. Protein concentration of the cell lysates was measured using a BCA protein assay kit (Pierce, Rockford, IL). 30 μg of protein from cell lysates were separated on 7.5% SDS-PAGE and transferred to a PVDF membrane. The membrane was incubated in blocking buffer for 2 hours at room temperature. The blocking buffer consisted of 5% nonfat dry milk in Tris buffered saline containing 0.1% Tween 20 (TBS-T). After washing the membrane with TBS-T, the membrane was incubated with rabbit anti-KIT antibody diluted 1:1000 overnight at 4°C. Then, the membrane was washed with TBS-T, and subjected to horseradish peroxidase-conjugated anti-rabbit immunoglobulin (DAKO, Copenhagen, Denmark) diluted 1:2000 as a second antibody for 1 hour at room temperature. Protein-antibody complexes were visualized with an ECL Western blotting detection and analysis system (Amersham Biosciences, Buckinghamshire, UK). The membrane was washed and stripped using Restore Western Blot Stripping Buffer (Pierce, Rockford, IL) and β-actin was detected using rabbit anti-β-actin antibody by the same method.

To examine the influence of SCF on cell proliferation, we used the Premix WST-1 Cell Proliferation Assay System (TAKARA BIO INC., Shiga, Japan). Five human pancreatic cancer cell lines were seeded at a density of 3 × 10³/100 μL in 96-well plates and allowed to adhere for 24 hours. Then, cultures were refed with fresh media containing various concentration of SCF. After 72 hours incubation, 10 μL WST-1 reagent was added to each well and the trays were incubated for 2 hours at 37°C, at which point the absorbance was measured using a microplate reader with a test wave length of 450 nm and a reference wavelength of 690 nm. In studies of imatinib mesylate, cells were incubated with SCF (10 ng/mL) alone or with both SCF (10 ng/mL) and imatinib mesylate (5 μM). In controls, cells were incubated without either SCF or imatinib mesylate. After 72 hours incubation, the inhibitory effect was evaluated using the WST-1 Cell Proliferation Assay System

The in vitro invasive potential of human pancreatic cancer cells was determined using BioCoat Matrigel Invasion Chambers (Becton Dickinson, Bedford, MA). This system is separated by a PET membrane coated with Matrigel Matrix such that only invasive cells can migrate through the membrane to the reverse side. After rehydration for 2 hours in a humidified incubator at 37°C with 5% CO₂, cells were suspended in medium containing 10% FBS, placed in the upper chamber at a concentration of 1 × 10⁵ cells/chamber and incubated for 24 hours at 37°C with 5% CO₂ with various concentrations of SCF. After 24 hours incubation, noninvasive cells were removed from the upper surface of the membrane by scrubbing. The cells on the reverse side of the membrane were fixed with 70% ethanol, stained with Giemsa solution, and counted under a microscope at 200 × magnification. We also investigated the inhibitory effect of imatinib mesylate on SCF-mediated invasion using the same method. Namely, cells were incubated with SCF (10 ng/mL) or with both SCF (10 ng/mL) and imatinib mesylate (5 μM) for 24 hours. For controls, cells were incubated without either SCF or imatinib mesylate.

Forty-two patients diagnosed with invasive ductal adenocarcinoma of the pancreas underwent pancreatic resections at Nagoya City University Hospital between April 1989 and March 2005 and were included in this study.

Specimens were fixed in 10% formalin and then embedded in paraffin. Specimens were sectioned into 3 μm-thick slices and the sections were deparaffinized. Next, the sections were subjected to autoclave treatment in Target Retrieval Solution High pH (DAKO, Copenhagen, Denmark) for 20 minutes at 95°C for KIT antigen retrieval. For SCF, the sections were subjected to autoclave treatment in 0.01 M sodium citrate buffer (pH 6.0) for 20 minutes at 95°C. Then, all sections were immersed in blocking reagent (Block Ace, Dainippon Pharma Co., Ltd., Osaka, Japan) for 10 minutes to reduce non-specific binding. After removal of excess serum, the sections were incubated with primary antibodies against KIT or SCF (diluted 1:50) for 60 minutes at room temperature. After rinsing in phosphate buffered saline (PBS), the sections were treated with horseradish peroxidase-labeled anti-rabbit immunoglobulin (DAKO, Copenhagen, Denmark) as a second antibody for 30 minutes at room temperature. The peroxidase reaction was visualized by incubating the sections with 0.02% 3.3'-diaminobenzidine tetrahydrochloride in 0.05 M Tris buffer (pH 7.6) with 0.01% hydrogen peroxide, followed by hematoxylin counter staining. As a negative control, staining was also performed using normal mouse immunoglobulin G in place of the primary antibody. Results of each immunohistochemical staining were evaluated as follows. The intensity of tissue staining was graded semiquantitatively on a four-point scale (-, +, ++, +++), and the proportion of cells stained was assessed on a four-tier scale (1. 0%-15%; 2. 15%-50%; 3. 50%-85%; and 4. 85%-100% cells stained). The cases were classified into positive group and negative group by the intensity and proportion of immunostained cancer cells for KIT and SCF. Cases where the immunostaining intensity was more than ++ and the proportion was more than 2 were defined as positive group. When venous system invasion was evaluated, we classified v0 and v1 as the slight group, and v2 and v3 as the severe group. We assessed the relationship between KIT expression and two groups of venous invasion. With regard to intrapancreatic neural invasion and lymphatic system invasion, evaluations were conducted in the same way as venous invasion.

All experiments were performed in triplicate. Results are expressed as means ± SD (standard deviation). Multiple group comparisons were performed by one-way ANOVA with a post hoc test for subsequent individual group comparison. Differences between two samples were analyzed by an unpaired t test. Survival analyses were calculated by using the Kaplan-Meier method and the differences in survival among subjects were compared with the log-rank test. The Mann-Whitney U-test was used to compare the immunohistochemical characteristics. Differences were considered statistically significant at p < 0.05.