In vivo antitumor effect of the mTOR inhibitor CCI-779 and gemcitabine in xenograft models of human pancreatic cancer

Authors


Abstract

Mammalian target of rapamycin (mTOR) is considered to be a major effector of cell growth and proliferation that controls protein synthesis through a large number of downstream targets. We investigated the expression of the phosphatidylinositol 3′-kinase (PI3K)/mTOR signaling pathway in human pancreatic cancer cells and tissues, and the in vivo antitumor effects of the mTOR inhibitor CCI-779 with/without gemcitabine in xenograft models of human pancreatic cancer. We found that the Akt, mTOR and p70 S6 kinase (S6K1) from the PI3K/mTOR signaling pathway were activated in all of the pancreatic cancer cell lines examined. When surgically resected tissue specimens of pancreatic ductal adenocarcinoma were examined, phosphorylation of Akt, mTOR and S6K1 was detected in 50, 55 and 65% of the specimens, respectively. Although CCI-779 had no additive or synergistic antiproliferative effect when combined with gemcitabine in vitro, it showed significant antitumor activity in the AsPC-1 subcutaneous xenograft model as both a single agent and in combination with gemictabine. Furthermore, in the Suit-2 peritoneal dissemination xenograft model, the combination of these 2 drugs achieved significantly better survival when compared with CCI-779 or gemcitabine alone. These results demonstrate promising activity of the mTOR inhibitor CCI-779 against human pancreatic cancer, and suggest that the inhibition of mTOR signaling can be exploited as a potentially tumor-selective therapeutic strategy. © 2005 Wiley-Liss, Inc.

Pancreatic cancer is the fifth leading cause of cancer-related death in most western countries.1 Since it is difficult to detect this disease at an early stage and because pancreatic cancer shows resistance to almost all available chemotherapy and radiation regimens, the prognosis remains dismal with a 5-year survival rate under 10%.2 Currently, the only active agent for advanced pancreatic ductal adenocarcinoma (PDA) appears to be gemcitabine, a DNA chain terminator. Even with this drug, however, the objective response rate is less than 20%.3 Until recently, the resistance of this cancer has generally been attributed to increased expression of detoxification mechanisms such as P-glycoprotein or antioxidants or to alterations of drug-metabolizing enzymes. Although these “classical” mechanisms are detectable in pancreatic cancer, there is no compelling evidence that their importance is greater than that in more responsive cancers,4 and so such mechanisms may not explain the high level of drug resistance seen in pancreatic cancer patients.

The mammalian target of rapamycin (mTOR) is one of the effectors regulated via the phosphatidylinositol 3′-kinase (PI3K)/Akt signaling pathway and it plays a central role in cell survival and proliferation.5 mTOR is a 289 kDa serine-threonine kinase that consists of 2,549 amino acids, and appears to modulate cellular signals in response to mitogenic stimuli and various nutrients, especially amino acids. In mammalian cells, growth factors (such as IGF, EGF and VEGF) regulate mTOR signaling through the PI3K/Akt pathway.5, 6 Intriguingly, ras shows mutation in more than 90% of PDAs, and it also positively regulates Akt.7 This pathway involves phosphatase and tensin homologue deleted from chromosome 10 (PTEN), 3-phosphoinositide-dependent kinase-1 (PDK1) and tuberous sclerosis complexes-1 and -2 (TSC-1 and -2). p70 S6 kinase (S6K1) and 4E-binding protein 1 (4E-BP1) are 2 major downstream targets of mTOR. These substrates bind to regulatory-associated protein of mTOR (Raptor) and are activated by phosphorylation of certain residues.8 In turn, S6K1 phosphorylates the 40s ribosomal S6 protein, leading to the translation of mRNAs bearing a 5′-TOP tract. Hypophospholyrated 4E-BP1 binds avidly to eIF4E, inhibiting the formation of the eIF4E complex and cap-dependent initiation of translation. After being hyperphosphorylated by mTOR, 4E-BP1 releases eIF4E and promotes cap-dependent protein synthesis.

Rapamycin is a macrocyclic lactone that causes cell cycle arrest in G1 phase and apoptosis of many normal cells and tumor cells.9, 10, 11, 12, 13, 14 It is well known that rapamycin forms a complex with FK506-binding protein (FKBP), and specifically inhibits mTOR activity through the binding of this complex to the FKBP rapamycin-binding domain of mTOR, although the details of the mechanism by which this complex acts are still unclear. CCI-779 is a novel water-soluble synthetic rapamycin ester that is more stable in vivo than rapamycin, and is currently being developed for the treatment of patients with cancer.

Dysregulation of mTOR signaling occurs in a wide variety of human tumors, and can lead to increased susceptibility to mTOR inhibitors.15, 16, 17, 18 Since activation of Akt has been reported in pancreatic cancer19, 20, 21 and since the mTOR-S6K1 signaling pathway is constitutively activated and is essential for proliferation of pancreatic cancer cells in vitro,22, 23 mTOR (which lies downstream of the PI3K/Akt pathway) might be a promising objective of novel molecular targeting therapy for pancreatic cancer. In addition, recent studies have revealed that CCI-779 showed antitumor activity against advanced refractory renal cell carcinoma in a randomized phase II study24 and human multiple myeloma cells in a xenograft model.25 These considerations also provide a rationale for using the mTOR inhibitor CCI-779 to treat pancreatic cancer.

The aims of this study were to investigate whether or not mTOR signaling is activated in PDA tissues and pancreatic cancer cell lines, as well as whether or not CCI-779 (an mTOR inhibitor) has an antiproliferative effect in animal models of human pancreatic cancer as a single agent and in combination with gemcitabine.

Material and methods

Antibodies

Rabbit anti-Akt (no. 9272), anti-mTOR (no. 2972), anti-S6K1 (no. 9202), anti-4E-BP1 (no. 9271), anti-phospho-Akt (Ser473, no. 9277), anti-phospho-PDK1 (Ser241, no. 3061), anti-phospho-GSK3 (Ser9, no. 9336), anti-phospho-mTOR (Ser2448, no. 2971), anti-phospho-S6K1 (Thr421/Ser424, no. 9204) and anti-phospho-4E-BP1 (Thr70, no. 9455) were purchased from Cell Signaling Technology, Inc. (Beverly, MA). Mouse anti-PTEN (clone 6H2.1, no. ABM2052) was purchased from Cascade Bioscience (Winchester, MA). Mouse anti-β-actin (cat. no. A-5441) was obtained from Sigma (St. Louis, MO).

Tissue samples

Formalin-fixed, paraffin-embedded tissues were obtained from the Department of Surgery and Basic Surgical Science at Kyoto University (Kyoto, Japan) between January 1994 and December 2000. These specimens were harvested from 20 patients with invasive PDA and were collected after informed consent was given, in accordance with institutional guidelines.

Immunohistochemical analysis

Serial sections (4-μm thick) were deparaffinized in 3 changes of xylene, rehydrated in descending concentrations of ethanol, and washed 3 times for 5 min each with double distilled water. After rehydration, the sections were heated in a microwave oven in 10 mM sodium citrate buffer (pH 6.0) for 1 min at full power followed by 9 min at medium power, and then incubated for 10 min at room temperature (RT) in 1% hydrogen peroxide in methanol to block endogenous peroxidase activity. Next, the sections were incubated for 60 min at RT with 10 mM phosphate-buffered saline (PBS) (pH 7.4) containing 5% normal goat serum, followed by overnight incubation at 4°C with the primary antibody diluted 1:50 to 1:100 in 10 mM PBS containing 5% normal goat serum. Then the sections were washed 3 times for 5 min in PBS and incubated for 60 min with horseradish peroxidase-conjugated anti-rabbit immunoglobulin G (Envision Kit, DakoCytomation, Kyoto, Japan) as the secondary antibody. The serial sections were stained for assessment of p-AKT, p-mTOR and p-S6K1 expression. Negative controls were done using nonspecific rabbit IgG as the primary antibody. The expression of each phosphorylated molecule was classified into 3 grades: 0 = undetectable, 1 = weak staining and 2 = strong staining. The proportion of positive cancer cells was also classified into 3 grades: 0 = none, 1 = 1–49% and 2 = 50–100%. Then the staining intensity score was multiplied by the score for positive cells to obtain the overall score. Specimens with a score of ≥2 were regarded as positive, while specimens with a score ≤1 as negative. Evaluation of immunostaining was performed by 2 investigators (D. I. and K. F.).

Cell lines and culture conditions

Six human pancreatic cancer cell lines were used. AsPC-1, BxPC-3 and Panc-1 cells were obtained from the ATCC (Rockville, MD). KMP-3 and KMP-4 were cell lines established in our department,26 while Suit-2 cells were kindly provided by Dr. Tomoda (National Kyushu Cancer Center, Fukuoka, Japan). Cells were maintained in the following media during incubation at 37°C in a humidified atmosphere of 5% CO2. AsPC-1 cells and BxPC-1 cells were cultured in RPMI 1640 medium (Gibco-BRL, Grand Island, NY) with 10% fetal bovine serum (FBS) (Gibco-BRL). KMP-3 cells and KMP-4 cells were cultured in an equal mixture of RPMI 1640 and F-12 nutrient medium (HAM) (Gibco-BRL) with 2% FBS. Panc-1 cells and Suit-2 cells were cultured in DMEM (Gibco-BRL) with 10% FBS. Each medium contained 100 units/ml of penicillin and 0.1 mg/ml of streptomycin.

Agents

CCI-779 was kindly provided by Wyeth-Ayerst Reseach Laboratories (Princeton, NJ) and gemcitabine was obtained from Eli Lilly (Bad Homburg, Germany). CCI-779 was stored as a dry powder at −20°C and dissolved in ethanol on the day of use. Gemcitabine was stored as a 50 mg/ml solution in sterile PBS at −20°C.

Cell proliferation assay

Cells (1.5 × 104/ml) were seeded into 24-well tissue culture dishes. After overnight incubation with complete medium, the cells were cultured under serum-starved conditions for 24 hr. Then the cells were treated with various concentrations of CCI-779 alone or combined with gemcitabine in the presence of 10% FBS. After 48 hr, the number of cells was counted using a cell counter (Coulter Z1; Beckman-Coulter, Fullerton, CA).

Western blot analysis

Cells were incubated with 20 nM CCI-779 or vehicle for 6 hr after culture with serum starvation for 24 hr. Then the cells were lysed in RIPA buffer containing 50 mM HEPES (pH 7.0), 250 mM NaCl, 0.1% Nonidet P-40, 1 mM phenylmethylsulfonylfluoride, and 20 μg/ml gabexate mesilate, followed by incubation on ice for 10 min. Subsequently, the lysate was sonicated for 10 sec. The extracts were cleaned by centrifugation at 15,000 rpm for 10 min at 4°C and the supernatants were harvested. The protein concentration was measured by a protein assay (cat. no. 23223, 23224; Pierce, Rockford, IL). Next, the lysates were resuspended in 1 vol of gel loading buffer containing 50 mM Tris–HCl (pH 6.7), 4% SDS, 0.02% bromophenol blue, 20% glycerol and 4% 2-mercaptoethanol. After heating at 95°C for 5 min, the extracted protein was subjected to western blotting as described previously.27 In brief, 30-μg aliquots of protein were size-fractionated to a single dimension by SDS-PAGE (8–15% gels) and transblotted onto a 0.45-μm polyvinylidene difluoride membrane (IPVH304F0; Millipore, Billerica, MA) using a semidry electroblot apparatus (Bio-Rad, Richmond, CA). The blots were then washed 3 times in Tris-buffered saline with 0.1% Tween-20 (TBST) and incubated for 1 hr at room temperature in blocking buffer (Block Ace; Dainipponseiyaku, Osaka, Japan). Subsequently, the blots were immunoblotted with an appropriate primary antibody for 1 hr at room temperature or overnight at 4°C. Unbound antibody was removed by washing the membrane 3 times for 10 min each with TBST, after which the membrane was incubated with a horseradish peroxidase-conjugated secondary antibody for 1 hr at room temperature. Reaction products were detected by the enhanced chemiluminescence system (Amersham, Buckinghamshire, UK) and membranes were treated with enhanced chemiluminescence reagents according to the manufacturer's protocol and were exposed to X-ray film for 5–120 sec.

Animal study

The animal study was performed in accordance with the guidelines for animal experiments of Kyoto University. To create the subcutaneous xenograft model, athymic female nude mice (4–6 weeks old) were subcutaneously inoculated with 1 × 106 AsPC-1 cells in 100-μl of Hank's Balanced Salt Solution (HBSS) (Gibco-BRL) containing 20% matrigel (BD Biosciences, Bedford, MA). Tumor-bearing animals were divided into the following 5 treatment groups: (i) CCI-779 (1 mg/kg, 5 times/week for 2 weeks), (ii) CCI-779 (10 mg/kg, 5 times/week for 2 weeks), (iii) gemcitabine (125 mg/kg, 2 times/week for 2 weeks), (iv) CCI-779 + gemcitabine (10 and 125 mg/kg for 2 weeks) and (v) vehicle alone (control). Each group consisted of 10 animals. CCI-779 and gemcitabine were administered intraperitoneally and treatment was started when subcutaneous tumors reached 20 mm3 in size. Tumor size was measured weekly with calipers and the volume was calculated by the following formula: tumor volume = (length × width2)/2.

For the peritoneal dissemination xenograft model, 1 × 106 Suit-2 cells in 250-μl sterile PBS were injected into the intraperitoneal cavity of nude mice. After injection, the mice were divided into the following 5 treatment groups:(i) CCI-779 (10 mg/kg, 5 times/week for 1 week), (ii) CCI-779 (20 mg/kg, 5 times/week for 1 week), (iii) gemcitabine (125 mg/kg, 2 times/week for 2 weeks), (iv) CCI-779 + gemcitabine (20 mg/kg, 5 times/week for 1 week and 125 mg/kg, 2 times/week for 2 weeks) and (v) vehicle alone (control). Each group consisted of 5 animals. Treatment was initiated 3 days after the injection of Suit-2 cells, and survival was measured from the date of injection.

Statistical analysis

Results are presented as the mean ± SEM for quantitative experiments. For statistical analysis of differences between the groups, one-way ANOVA was performed and statistical significance was considered to be present at p < 0.05. Each experiment was performed at least 3 times independently. Survival was calculated by the Kaplan–Meier method, and differences between each group were evaluated with the log-rank test.

Results

Expression of mTOR pathway-related molecules in PDA tissue specimens

To examine the role of the mTOR signaling pathway in human pancreatic cancer, we performed immunohistochemical analysis of tissue specimens from PDAs. Expression of p-Akt (Ser473), p-mTOR (Ser2448) and p-S6K1 (Thr421/Ser424) was observed in 50, 55 and 65% of the specimens, respectively (Fig. 1d1f). Staining was mainly seen in the cytoplasm. As shown in Figure 1g1i, positive expression of p-Akt was associated with positive expression of p-mTOR and p-S6K1 in serial sections of the same specimen. There was a significant correlation among the expression of these 3 molecules (p-Akt vs p-mTOR: p = 0.0013, p-mTOR vs p-S6K1: p = 0.045 and p-mTOR vs p-S6K1: p = 0.0004, respectively). When the expression of these molecules was compared with various clinicopathological features (pT, pN, pM, UICC Stage, tumor size and histological grade), however, there were no significant associations. There was also no statistical correlation between the expression of these molecules and survival.

Figure 1.

Immunohistochemical staining of phospho-Akt (Ser473, no. 9277) (a, d and g), phospho-mTOR (Ser2448, no. 2971) (b, e and h) and phospho-p70S6K (Thr421/Ser424, no. 9204) (c, f and i) in specimens of pancreatic ductal adenocarcinoma (PDA). The upper panel (a, b and c, serial sections) represents the sections of PDA specimens regarded as negative. The middle panel (d, e and f) shows typical photomicrographs of positive sections of PDA specimens. The lower panel (g, h and i) shows expression of these 3 molecules in serial sections of PDA. Strong cytosolic staining for p-Akt, p-mTOR and p-p70S6K is seen. The nuclei were counterstained with Mayer's hematoxylin (original magnification: ×400).

Expression of mTOR pathway-related molecules by pancreatic cancer cell lines

Next, we performed western blot analysis to examine the expression of mTOR pathway-related molecules by pancreatic cancer cell lines. As previously reported,28 PTEN (a negative regulator of PI3K) was not expressed by KMP-3 and KMP-4 cells. However, the other molecules, including Akt, mTOR, S6K1 and 4E-BP1, were expressed by all of the cell lines examined (Fig. 2a).

Figure 2.

(a) Western blot analysis of mTOR signaling pathway-related proteins in pancreatic cancer cell lines. Twenty micrograms of total protein extracted from cells starved of serum for 24 hr was applied to each lane. Akt, mTOR, p70S6K and 4E-BP1 were expressed at almost equivalent levels by all of the cell lines. PTEN expression was not detected in KMP3 and KMP-4 cells. Representative results from 3 independent experiments are shown. (b) Western blot analysis of the phosphorylation of mTOR signaling pathway-related proteins in pancreatic cancer cell lines. The phosphorylation site and catalog number of each protein are as follows: phospho-Akt (Ser473, no. 9277), phospho-PDK1 (Ser241, no. 3061), phospho-GSK3 (Ser9, no. 9336), phospho-mTOR (Ser2448, no. 2971), phospho-S6K1 (Thr421/Ser424, no. 9204), and phospho-4E-BP1 (Thr70, no. 9455). Twenty micrograms of total proteins was applied to each lane. All of the proteins examined were phosphorylated in all of the cell lines even after 24 hr of serum starvation. Reflecting the loss of PTEN expression, Akt was strongly phosphorylated in KMP-3 and KMP-4 cells. Representative results from three independent experiments are shown.

We also examined the phosphorylation of these molecules after 24 hr of serum starvation. Phospho-PDK1, which phosphorylates Akt, showed almost equal expression in all of the cell lines. With respect to phospho-Akt, which reflects loss of PTEN expression, its expression was much stronger in KMP-3 and KMP-4 cells than in the other cell lines. Interestingly, mTOR was almost equally phosphorylated in all of the cell lines, a result that was inconsistent with p-Akt expression. Furthermore, phosphorylation of S6K1 and 4E-BP1 was also observed in all of the cell lines. On the other hand, the pattern of phospho-GSK expression, which is another target of Akt, was the same as that of phospho-Akt expression (Fig. 2b).

Effect of CCI-779 alone and combined with gemcitabine on cultured pancreatic cancer cells

These results suggested that the mTOR signaling pathway might be constitutively activated in PDAs. Therefore, we next examined the growth-inhibitory effect of a specific mTOR inhibitor, CCI-779, on cultured pancreatic cancer cells. After serum starvation for 24 hr, 6 pancreatic cancer cell lines were cultured for 48 hr with complete growth medium containing increasing concentrations of CCI-779. KMP-3 (a PTEN-deficient cell line) and AsPC-1 showed a high sensitivity to CCI-779 (IC50 = 2 and 20 nM, respectively), while Suit-2 cells had the highest resistance (IC50 > 500 nM) (Fig. 3, Table I).

Figure 3.

Growth-inhibitory effect of CCI-779 on pancreatic cancer cell lines. Cells were cultured in the serum-free medium for 24 hr, followed by incubation with various concentrations of CCI-779 for 48 hr and the number of viable cells was counted. Data represent the mean ± SEM of triplicate determinations from 1 of 3 representative experiments.

Table I. Summary of Expression Patterns of mTOR Related Molecules and The Effects of CCI-779 in Pancreatic Cancer Cell Lines
 AsPC-1BxPC-3KMP-3KMP-4Panc-1Suit-2
PTEN++++
p-Akt+++++++++++++++
p-mTOR+++++++++++++
p-p70S6K+++++++
p-4EBP1+++++++++++
IC50 (nM)20>2002100100>500

We also examined suppression of the phophorylation of S6K1 by CCI-779 treatment. As shown in Figure 4, phosphorylation of S6K1 was inhibited in all of the cell lines by treatment with 20 nM CCI-779 for 6 hr. In AsPC-1 cells, the phospho-specific band of S6K1 was decreased by 88% after CCI-779 treatment when compared with that after vehicle treatment. In Suit-2 cells that were resistant to this drug, however, p70S6K phosphorylation was only decreased by 39% (Fig. 4). From these data, AsPC-1 and Suit-2 cells were chosen as representative sensitive and resistant cells to CCI-779 for further experiments.

Figure 4.

Western blot analysis of the phosphorylation of S6K in pancreatic cancer cells after treatment with CCI-779. Twenty micrograms of total protein extracted from untreated cells or cells treated with 20 nM CCI-779 was applied to each lane. The inhibitory rate was calculated by densitometry (ATTO, Osaka, Japan). A representative result from 3 independent experiments is shown.

Before we performed in vivo experiments, the growth inhibitory effect of CCI-779 combined with gemcitabine was examined in vitro using AsPC-1 and Suit-2 cells. Our preliminary study demonstrated that AsPC-1 cells were resistant to gemcitabine, while Suit-2 cells were sensitive (data not shown). As shown in Figure 5, combination therapy had neither a synergistic nor an additive effect on both cell lines.

Figure 5.

Growth-inhibitory effect of combined treatment with CCI-779 and gemcitabine in (a) AsPC-1 and (b) Suit-2 cells. Since AsPC-1 cells were sensitive to CCI-779 and resistant to gemcitabine, they were treated with increasing concentrations of CCI-779 and a fixed concentration of gemcitabine (10 nM) for 48 hr. In contrast to AsPC-1 cells, Suit-2 cells were incubated with a fixed concentration of CCI-779 (20 nM) and increasing concentrations of gemcitabine for 48 hr, since these cells were resistant to CCI-779 and sensitive to gemcitabine. As shown in both graphs for AsPC-1 and Suit-2 cells, no additive or synergistic effects were found for both cell lines. Data represent the mean ± SEM of 9 determinations from 3 representative experiments.

Antitumor activity of CCI-779 in xenograft models

The antitumor effect of CCI-779 was also examined using AsPC-1 and Suit-2 xenograft models. After inoculation of 1 × 106 cells into the right flank (AsPC-1) or the peritoneal cavity (Suit-2) of nude mice, CCI-779 and/or gemcitabine was administered by intraperitoneal injection. In the AsPC-1 subcutaneous xenograft model, CCI-779 was able to delay tumor growth as single agent (Fig. 6). Treatment with a low dose of CCI-779 significantly decreased tumor volume by 74% after 35 days (p = 0.037) and treatment with a high dose of CCI-779 also significantly decreased tumor volume by 68% when compared with the control (p = 0.009). Consistent with the in vitro results, no growth inhibition was observed in the mice treated with gemcitabine alone. Interestingly, treatment with a combination of CCI-779 and gemcitabine caused a synergistic decrease of tumor volume by 41% when compared with the control (p = 0.0002). In addition, the combination therapy resulted in a significant decrease of tumor volume even when compared with the CCI-779 alone as well as the gemcitabine alone (p = 0.0042).

Figure 6.

Suppression of the growth of subcutaneous xenografts of AsPC-1 cells by CCI-779, gemcitabine, or a combination of both agents. Athymic mice bearing AsPC-1 xenografts were divided into 5 treatment groups. Open triangles indicate the day of CCI-779 administration and open circles indicate gemcitabine administration.

In the Suit-2 intraperitoneal xenograft model, our preliminary study revealed that untreated mice died of peritoneal dissemination after approximately 7 weeks with bloody ascites. All of the mice in the control and CCI-779 monotherapy group died within 45 days, and there were no significant differences among the groups (Fig. 7). Consistent with the high in vitro sensitivity to gemcitabine, the mice treated with gemcitabine alone showed significantly longer survival time than untreated mice or CCI-779-treated mice (p = 0.0198 and 0.0026, respectively). Interestingly, just as in the AsPC-1 model, mice treated with a combination of CCI-779 and gemcitabine showed the best survival. In fact, 2 of the 5 mice given combination therapy survived for more than 100 days, and no tumor was seen in the peritoneal cavity when the mice were killed for autopsy on day 107.

Figure 7.

Kaplan–Meier plot showing the survival of Suit-2-bearing mice after treatment with CCI-779, gemcitabine, or a combination of both agents. Suit-2 cells were injected intraperitoneally into athymic mice. Three days after injection, the mice were divided into 5 treatment groups and drug administration was initiated.

Discussion

The present study demonstrated that the signaling pathways regulating mTOR are frequently activated in pancreatic cancer, and that a rapamycin derivative (CCI-779) could inhibit tumor growth in xenograft models as a single agent as well as combined with gemcitabine.

It is well known that ras mutation can be detected in more than 90% of pancreatic cancers and that ras protein binds and activates the catalytic subunit of PI3K.29, 30 Recent studies have revealed that the genes encoding the catalytic subunit of PI3K and Akt are amplified in various human cancers, including pancreatic cancer.19, 20, 21, 31 Interestingly, Hu et al. have demonstrated that transfection of mutant K- or N-ras genes into multiple myeloma cells confers a high sensitivity to CCI-779.15 Furthermore, it has been reported that Akt regulates the sensitivity of tumors to mTOR inhibitors.32 These studies lead to the hypothesis that mTOR and its downstream targets, which are considered to be essential for cell survival and proliferation through regulation of protein synthesis, might be activated in pancreatic cancer, and suggest that the inhibition of mTOR signaling could block tumor growth.

First, we examined the expression and constitutive phosphorylation of molecules from the mTOR signaling pathway in pancreatic cancer tissues and cell lines. Using specific antibodies, immunohistochemical analysis revealed that pospho-Akt, phospho-mTOR, and phospho-S6K1 were expressed in at least half of the PDA specimens examined (50, 55 and 65%, respectively). Western blot analysis also demonstrated that the Akt/mTOR signaling pathway, including the downstream effectors S6K1 and 4E-BP1, was constitutively activated in all of the pancreatic cancer cell lines examined. These data suggested that the mTOR signaling pathway might play a role in the development of pancreatic cancer and that mTOR inhibitors might be active against pancreatic cancer.

Next, we used a cell proliferation assay to examine the in vitro growth inhibitory effect of CCI 779 against 6 pancreatic cancer cell lines, including 2 PTEN-deficient cell lines (KMP-3 and KMP-4 cells26, 28). Although PTEN protein has been suggested to have multiple important functions pertaining to cell cycle regulation and survival, its best characterized role is related to regulation of the PI3K/Akt signaling pathway.31 Recently, Neshat et al. have shown that CCI-779 inhibits the hyperproliferation of PTEN-deficient cells.33 Also, the administration of CCI-779 to PTEN heterozygous mice, which develop multiple neoplasms, attenuates tumor development.34 PTEN is a well-known negative regulator of PI3K, and acquired mutations of the PTEN gene have been reported in several tumors35; however, PTEN mutation occurs rarely in patients with pancreatic cancer.36 In this study, we demonstrated that KMP-3 and 4 cells were PTEN-deficient cells and were sensitive to CCI-779 (IC50: 2 and 100 nM, respectively) in spite of the relatively minimal suppression of S6K1 phosphorylation in both cells. We consider that PTEN status is also associated with the sensitivity of CCI-779 in KMP-3 and 4 cells besides S6K1 dephosphorylation.

Recent work has suggested that S6K1 activity may be an appropriate marker for mTOR-interacting agents.37, 38 It has been reported that rapamycin dephosphorylates S6K1 and inhibits the proliferation of pancreatic cancer cell lines such as BxPC-3, Panc-1 and MiaPaCa-2.22, 23 Consistent with these data, the suppression of S6K1 phosphorylation was maximum (88%) in AsPC-1 cells and was lowest (39%) in Suit-2 cells, which were highly sensitive and resistant to CCI-779, respectively, suggesting that suppression of S6K1 phosphorylation was correlated with the inhibitory effect of CCI-779 on pancreatic cancer cells (Fig. 4). With respect to 4E-BP1, another major downstream target of mTOR, CCI-779 had no effect on its phosphorylation in this study (data not shown). Interestingly, it has been shown that the treatment of nude mice bearing a CCI-779-sensitive breast cancer cell line (MDA-468) with CCI-779 results in marked inhibition of tumor tissue S6K1 activity, which is identical to the inhibition seen in peripheral blood mononuclear cells (PBMCs).37 This suggests that PBMCs may be an adequate surrogate tissue for measuring S6K1 activity in vivo. In the clinical setting, it may be very useful to assess the pharmacodynamic effects of CCI-779 by using an S6K1 assay of PBMCs. From these data, we considered that the S6K1 activity should be regarded as a suitable surrogate marker of biological effect for an mTOR inhibitor, CCI-779.

We hypothesized that constitutively elevated levels of phosphorylated mTOR in pancreatic cancer cells could protect against apoptosis induced by chemotherapy agents and contribute to drug resistance. Therefore, the inhibition of mTOR phosphorylation by CCI-779 may enhance the sensitivity of pancreatic cancer cells to gemcitabine. To clarify this hypothesis, the effects of combined therapy with CCI-779 and gemcitabine were investigated in vitro and in vivo. From the results shown in Figure 3, we chose AsPC-1 and Suit-2 cells as CCI-779-sensitive and -resistant cells, respectively. In the in vitro proliferation assay, AsPC-1 was resistant and Suit-2 was sensitive to gemcitabine treatment as a single agent, unlike their susceptibility to CCI-779. As shown in Figure 5, combined therapy with CCI-779 and gemcitabine did not have a synergistic effect on both cell lines. Although numerous studies have demonstrated that rapamycin and its derivatives could induce apoptosis in various cancer cell lines,39 the mechanisms of antiproliferative effects of CCI-779 in vitro in this study remain unclear. To elucidate the mechanisms, further investigation is now on-going in our laboratory. Next, the in vivo antitumor activity of CCI-779 as a single agent or combined with gemcitabine was examined. Since it is well known that many patients with PDA die from peritoneal dissemination, we studied an in vivo Suit-2 peritoneal dissemination model as well as an AsPC-1 subcutaneous model. Consistent with the data from our in vitro proliferation assay, CCI-779 showed an antitumor effect in the AsPC-1 subcutaneous model but not the Suit-2 peritoneal dissemination model, whereas gemcitabine showed the reverse activity as a single agent. Combined therapy with CCI-779 and gemcitabine led to delayed tumor growth in the AsPC-1 subcutaneous model and longer survival in the Suit-2 peritoneal dissemination model when compared with single agent therapy, suggesting that a synergistic effect might be obtained in the clinical setting.

This discrepancy between the susceptibility to combination therapy in vitro and in vivo is interesting. It was recently reported that rapamycin has a potent antiangiogenic effect by decreasing the production of vascular endothelial growth factor (VEGF).40 El-Hashemite et al. have also reported that the loss of TSC-1 and TSC-2, which are upstream negative regulators of mTOR, induces VEGF production via activation of mTOR.41 With respect to the relationship between VEGF and pancreatic cancer, we have previously shown that VEGF may play an important role in tumor angiogenesis.42 Recently, it has been reported that treatment of rapamycin suppressed primary and metastatic liver tumor growth by the inhibition of tumor angiogenesis in an orthotopic model of AsPC-1, one of pancreatic cancer cell lines.39 Intriguingly, in this model, combination treatment of rapamycin and anti-VEGF antibody 2C3 further improved the antitumor effects compared with single treatment of rapamycin or 2C3. Moreover, other investigators have demonstrated synergistic growth-inhibitory effects by combined treatment of gemcitabine and rapamycin in another orthotopic model of L3.6pl pancreatic cancer cells.43 They have revealed that rapamycin induced apoptosis of endothelial cells and tumor vessel thrombosis in this model and suggested that combining rapamycin with other cytotoxic drugs could be a feasible therapy of pancreatic cancer. From these data, we speculate that an antiangiogenic effect of CCI-779 might be involved in the mechanism of its synergistic antitumor activity in this study.

In conclusion, mTOR and related molecules are frequently activated in pancreatic cancer, and an mTOR inhibitor (CCI-779) inhibits the proliferation of certain pancreatic cancer cells in vitro and shows a synergistic effect with gemcitabine in vivo. These data suggest that mTOR might be a promising objective of novel molecular targeting therapy for pancreatic cancer.

Ancillary