Cell proliferation was assessed in MTT proliferation assays by incubation of cells with two-fold dilutions of Lys⁸-ψ-Lys⁹NT (8–13) at concentrations ranging from 0.07–16.67 nM under serum-free conditions for seven days. Dose-response curves for Lys⁸-ψ-Lys⁹NT (8-13) and BxPC-3, MIA PaCa-2 and HT-29 cells are shown in Figure 1. Proliferation of BxPC-3 cells was significantly stimulated by 0.07–1.04 nM Lys⁸-ψ-Lys⁹NT (8–13), whereas a concentration of 16.67 nM resulted in significant growth inhibition. Cell proliferation was not impaired by application of 20 μM SR 142948 in serum-free control medium alone, however, 20 μM SR 142948 in combination with Lys⁸-ψ-Lys⁹NT (8–13) revealed a dose-response relationship significantly different from NT-analog-stimulated growth over the whole concentration range, except at 8.3 nM Lys⁸-ψ-Lys⁹NT (8–13). Differences in the growth of MIA PaCa-2 cells treated with 0.07–16.67 nM Lys⁸-ψ-Lys⁹NT (8–13) and 20 μM SR 142948, either alone or in combination, were not significantly different from basal cell proliferation over the whole concentration range. NTR1-positive HT-29 colon cancer cells revealed increased growth at concentrations of 0.07–2.08 nM Lys⁸-ψ-Lys⁹NT (8–13); however, proliferation was significantly inhibited at 16.67 nM Lys⁸-ψ-Lys⁹NT (8–13), similar to BxPC-3 cells. The presence of 20 μM SR 142948 alone did not suppress proliferation below medium control levels in HT-29 cells; however, treatment with Lys⁸-ψ-Lys⁹NT (8–13) in combination with the antagonist impeded the cell growth at concentrations of 0.07–1.04 nM Lys⁸-ψ-Lys⁹NT (8–13) significantly.

Thus, application of Lys⁸-ψ-Lys⁹NT (8–13) exerted minor growth stimulatory effects at lower concentrations and inhibition of proliferation at highest concentrations in BxPC-3 and HT-29, but not in MIA PaCa-2 cells.

To investigate a putative correlation between cell cycle distribution and cell density, cells were seeded into six-well plates at increasing cell numbers and incubated in serum-supplemented culture medium until the cell layers in wells with highest cell numbers reached almost 100% of confluence. Figure 2 shows the dependence of the cell cycle distribution of BxPC-3, MIA PaCa-2 and HT-29 cells on the culture density. After an initial lag phase, BxPC-3 cells exhibited a considerable increase of cells in S phase with a maximum of DNA synthesis at 54% of confluence, followed by an increase of quiescent cells at high densities. MIA PaCa-2 cells, growing as scattered single cells, showed a large S phase fraction in highly dense cultures. Similar to BxPC-3 cells, proliferation of HT-29 cells was low during the lag phase and reached a maximum at approximately 28%, followed by a decrease of S phase cells in dense monolayers. According to these results, the different cell lines seem to become partially arrested at the S/G2M transition.

The effect of different cell densities on NTR1 cell surface expression was studied by flow cytometry using the monoclonal antibody B-N6, and results are shown in Figure 3. Cell surface expression of NTR1 in BxPC-3 cells decreased significantly with increasing cell density, but reached a maximum in cultures exceeding 77% of confluence (Figure 3, left, top). By contrast, MIA PaCa-2 cells exhibited low overall levels of NTR1 independent of cell density (Figure 3, left, middle). NTR1 levels of HT-29 cells were maximal at densities exceeding approximately 28 % (Figure 3, left, bottom).

For comparison, cell surface expression of the other growth factor receptor EGFR was quantified in dependence of cell density by use of the monoclonal antibody clone 225. In general, EGFR revealed higher expression levels than NTR1 in all cell lines (Figure 4, right). BxPC-3 cells exhibited augmented expression of EGFR at densities up to 77% and significantly reduced EGFR levels near confluence (Figure 3, right, top). EGFR expression in MIA PaCa-2 cells was significantly decreased above 37% of confluence (Figure 3, right, middle). HT-29 cells revealed considerable EGFR levels in scattered cells and reduced expression in cultures at and above 50% of confluence (Figure 3, right, bottom).

Since acidic regions are frequently found within solid tumors, it was investigated whether distinct extracellular pH conditions altered membrane expression of NTR1 and EGFR. For this purpose, cells were incubated in serum-free media of pH 7.1, 7.4 or 7.8 for four days and receptor levels analyzed by flow cytometry. As demonstrated in Figure 4 (left), BxPC-3 and PANC-1 cells revealed significant upregulation of NTR1 expression at pH_e 7.1 and downregulation at pH_e 7.8, in comparison to pH_e 7.4. Treatment with 10 nM Lys⁸-ψ-Lys⁹NT (8-13) resulted in significant internalization of NTR1 in BxPC-3 and PANC-1 cells independent of pH_e (data not shown). On the contrary, the low NTR1 expression in MIA PaCa-2 cells declined with decreasing pH_e and the presence of Lys⁸-ψ-Lys⁹NT (8–13) did not alter NTR1 expression significantly.

Parallel experiments to evaluate a putative dependence of EGFR expression on pH_e were performed for comparison (Figure 4, right). BxPC-3 cells exhibited high levels of EGFR that were significantly decreased at pH_e 7.8; however, receptor expression was not altered in PANC-1 cells.

Again, MIA PaCa-2 cells revealed effects opposite to those of BxPC-3 and PANC-1 cells with downregulation of EGFR at acidic and alkaline pH_e, respectively.

Similar to BxPC-3 and PANC-1 cells, NTR1 levels in HT-29 cells decreased with increasing pH_e and expression of the receptor was considerably reduced in the presence of the NT analog, independent of pH_e. EGFR expression was highest under physiological conditions and decreased at acidic and alkaline pH_e in Lys⁸-ψ-Lys⁹NT (8-13)-treated cells (data not shown).

In addition, effects of acidic or alkaline extracellular conditions on constitutive and Lys⁸-ψ-Lys⁹NT (8-13)-induced IL-8 production were investigated by stimulation of cells with 100 nM NT-analog in serum-free media of pH 7.1, 7.4 or 7.8 for seven hours. As shown in Figure 5 for BxPC-3 and PANC-1 cells, constitutive as well as NT analog-induced secretion of IL-8 was significantly elevated under acidic and alkaline conditions. The proportion of IL-8 that was enhanced by Lys⁸-ψ-Lys⁹NT (8-13) was higher under acidic and alkaline conditions compared to pH_e 7.4. BxPC-3 cells exhibited an increase of NT-analog-induced IL-8 of +20.4 ± 8.5% under acidic and 19.7 ± 4.0% under alkaline conditions, while the fraction of NT-analog-induced IL-8 production amounted to +9.7 ± 3.5% under physiological conditions. PANC-1 cells revealed increases of +22.2 ± 2.5% and +10.8 ± 3.7% at pH_e 7.1 and 7.8, respectively, in response to Lys⁸-ψ-Lys⁹NT (8–13) compared to ΔIL-8 of +13.1 ± 4.8% at physiological pH_e.

Pancreatic adenocarcinoma is characterized by highly aggressive tumor growth and early metastasis, resulting in a dismal prognosis for affected patients. EGF and neuropeptides like bombesin, cholecystokinin, neuromedin N and NT may act as growth factors and stimulate proliferation on the one hand and/or affect other cellular functions in different tumor cell types, including pancreatic cancer, on the other hand [1,15]. Expression of EGFR was reported in approximately 70% and production of NT and expression of NTRs in up to 90% of pancreatic tumors and derived cell lines [2,3,6]. Despite their abundant presence in cancer cells, suggesting them as useful drug targets, approaches to suppress EGFR- or NTR1-mediated signal transduction pathways by use of anti-EGFR monoclonal antibodies, such as cetuximab, small molecule tyrosine kinase inhibitors like erlotinib or the nonpeptide NTR1 antagonist SR 48692, have not met expectations in clinical studies so far [7]. This has been attributed to limited accessibility of the tumor cells for the large monoclonal antibodies. Furthermore, KRAS mutations leading to persistent activation of cellular signaling events downstream of EGFR are found in 50–95% of pancreatic cancer samples and render proliferation of tumor cells independent of EGFR [7,16]. However, application of gemcitabine combined with erlotinib for first-line treatment of advanced pancreatic cancer was approved by the U.S. Food and Drug Administration in 2005 [17].

Initial evidence of the role of the NT-NTR system in various tumor entities concentrated attention to development of NTR inhibitors that were demonstrated to be partially effective in inhibiting basal and, preferentially, NT-stimulated growth of tumors in experimental animal models [3]. In particular, the selective nonpeptide NTR1 antagonist SR 48692 was tested in a double-blind, randomized, phase II-III maintenance study versus placebo in patients with extensive stage small cell lung cancer following a first-line chemotherapy with cisplatin and etoposide (NCI identifier NCT00290953) and revealed only sporadic responses at best, which resulted in discontinuation of the development as an anticancer drug (Evaluation of the Overall Survival of Meclinertant Versus Placebo After a First Line Chemotherapy With Cisplatin + Etoposide; Clinical Trials, 2009) [18]. This failure to suppress tumor cell proliferation may be explained by the capability of cancer cells to employ paracrine/autocrine growth factors other than NT for proliferation and/or by the significance of NT-NTR signaling for tumor-associated cellular functions besides growth control, for example the acidification of extracellular compartment [13].

Figure 6A demonstrates an overview of the mechanisms leading to extracellular acidosis commonly found in solid tumors of larger size. Low interstitial pH may be either due to insufficient supply of the tumor with nutrients and oxygen as a result of irregular vascularization. Decrease of the partial pressure of oxygen (pO2) with increasing distance from blood vessels leads to a metabolic switch of hypoxic tumor cells from respiration to anaerobic glycolysis with extracellular accumulation of lactic acid (Pasteur effect) [19]. Moreover, cancer cells draw energy from anaerobic glycolysis even under normoxic conditions (Warburg effect) [20,21]. According to our own results, we propose a model for extracellular acidosis, which is known to increase the metastatic potential of tumor cells and promotes dissemination responsible for poor clinical outcome involving NT (Figure 6B). In contrast to NTR1, EGFR seems to become downregulated in the tumor cell aggregates during their switch to metastasis.

Own experiments revealed expression of NTR1 and NTR3 at the mRNA level in all three pancreatic cancer cell lines and cell surface expression of NTR1 in BxPC-3, PANC-1 and MIA PaCa-2 pancreatic cancer cells was demonstrated using the monoclonal antibody B-N6 directed to NTR1. In the present study, proliferative responses of the cell lines to the NT analog were tested in MTT assays. Growth of BxPC-3 and HT-29 cells was stimulated to a minor degree at lower concentrations of Lys⁸-ψ-Lys⁹NT (8–13) in an SR 142948-sensitive manner and was inhibited at higher concentrations, contrary to MIA PaCa-2 cells lacking a specific growth response.

In order to assess the relationship between NTR1 and cell density, membrane expression of the receptor was determined in cultures of varying cell numbers. According to these results, the cell lines studied here seem to become arrested in the S/G2M phases upon reaching confluence. NTR1 expression varied with the cell density and was found to be maximal at medium density in HT-29 cells and at confluence in BxPC-3, respectively, whereas MIA PaCa-2 cells exhibited lower NTR1 expression in general, independent of cell density. Since the relative significance of EGFR and NTR1 in regard to tumor cell proliferation and dissemination has not been elucidated for the pancreatic cancer cell lines so far, we compared the dependence of the expression of both receptors on cell density and pH_e. The ubiquitous EGFR exhibited significantly decreased cell surface expression at cell densities near confluence in BxPC-3 and MIA PaCa-2 pancreatic cancer as well as HT-29 colon cancer cells, in contrast to NTR1. These findings do not correlate with the KRAS status of the cell lines, since BxPC-3 and HT-29 are wildtype, whereas MIA PaCa-2 has mutated KRAS (G12C) [22]. Anti-EGFR treatment is expected to have no effect in MIA PaCa-2 and PANC-1 cells characterized by overactivated mutant KRAS. Treatment of cells with 10 nM Lys⁸-ψ-Lys⁹NT (8–13) did not modulate EGFR in BxPC-3 and PANC-1 cells (data not shown). The results are in good agreement with experiments which described cell density-dependent inhibition of EGFR signaling by p38α MAPK [23]. Confluent activation of p38α MAPK inhibits EGFR-induced proliferation through receptor degradation and may be enhanced further by NT-induced stimulation of the same kinase in NT-producing cell lines like BxPC-3 and HT-29 [11,24]. Data point to a significant function of NTR1 in biological processes of non-dividing, resting pancreatic cancer cells, corroborating the conflicting experimental and clinical reports on the role of NT as a growth factor and the inability of the NTR1 inhibitor SR 48692 to impede tumor progression in patients.

Solid tumors of larger size with irregular vascularization are known to contain highly acidic regions with pH < 6.8 [20,25,26]. Previously, we had demonstrated for the first time that NT induced NHE1-dependent intracellular alkalinization and extracellular acidification mediated by mitogen- and stress-activated kinase 1/2 (MSK1/2) in pancreatic cancer cells [11]. Therefore, the pancreatic cell lines were investigated using media adjusted to different pH values. Normally, cancer cells exhibit reduced growth under acidic conditions; however, all three pancreatic cancer cell lines proliferated more rapidly when incubated in medium of pH 7.1, while cell numbers were markedly decreased at an alkaline pH of 7.8. Both BxPC-3 and PANC-1 cells revealed enhanced NTR1 cell surface expression under acidic extracellular conditions, unlike MIA PaCa-2 cells exhibiting NTR1 levels increasing with pH_e. EGFR expression was not significantly altered in dependence of pH_e, with the exception of slightly decreased receptor levels in BxPC-3 cells at alkaline and in MIA PaCa-2 cells at acidic and alkaline pH_e values.

The main source of intratumoral acidity has been considered to be lactate produced during anaerobic or aerobic glycolysis, respectively [27,28]. To define further targets of NT in pancreatic cancer cells, whole-genome gene expression analysis was performed using Lys⁸-ψ-Lys⁹NT (8–13)-treated BxPC-3 cells and, most interestingly, induction of hypoxia-inducible factor 1α (HIF-1α) and upregulation of genes implicated in glycolytic metabolism was observed although cells were incubated under normoxic conditions.

In conclusion, NT/NTR1 signaling in pancreatic cancer cells seems to promote the induction of a metastatic phenotype, prior to acidosis resulting from a metabolic switch to glycolysis common for larger cancers, in contrast to its varying effects on tumor cell proliferation. Localized acidic extracellular conditions favor tumor cell invasion at a very early stage of tumor development [29]. This is consistent with the literature of the contribution of acidic pH_e to tumor invasion through selection of aggressive tumor cells, suppression of apoptosis, induction of proinvasive mediators and stimulation of protease secretion [30,31]. A central player of metastasis is IL-8, an autocrine/paracrine factor that affects several cellular processes and, importantly, triggers release of matrix metalloproteinases or urinary plasminogen activator from cells [32]. Upregulation of IL-8 in solid tumors like colon and pancreatic cancer is known to be stimulated by the prevalent hypoxic and acidic environment [32,33]. Own experiments clearly demonstrated that Lys⁸-ψ-Lys⁹NT (8–13) induced secretion of IL-8 in BxPC-3 and PANC-1 cells. These findings, taken together, strongly indicate that NT signaling in pancreatic cancer cells can initiate and/or stimulate cellular processes which contribute to the transformation of cells holding less aggressive potential to more metastatic cells. Downregulation of EGFR at higher cell densities concomitant with upregulation of NTR1 seems to indicate mutual exclusive roles for these receptors. EGFR may be important in the initial growth of pancreatic tumor cells and replaced by increased expression of other growth factor receptors, like NTR1, during metastatic dissemination.

Unless otherwise noted, all chemicals and solutions were obtained from Sigma-Aldrich (St. Louis, MO, U.S.). The stable NT analog Lys⁸-ψ-Lys⁹NT (8–13) (H-Lys⁸-ψ-Lys⁹-Pro¹⁰-Tyr¹¹-Ile¹²-Leu¹³-OH) was purchased from Bachem (Weil am Rhein, Germany), dissolved in distilled water at a concentration of 5 mg/mL (6.7 mM Lys⁸-ψ-Lys⁹NT(8–13)) and aliquots stored at −20 °C. SR 142948 from Tocris Cookson (Bristol, U.K.) was dissolved in dimethyl sulfoxide (DMSO) at 5 mM. Aliquots were stored at −20 °C. Ca²⁺- and Mg²⁺-free Dulbecco's phosphate buffered saline (PBS) for trypsinization of cell cultures was purchased from Gibco/Invitrogen (Carlsbad, CA, U.S.).

The pancreatic cancer cell lines BxPC-3, MIA PaCa-2 and PANC-1 and the colon carcinoma cell line HT-29 were obtained from the American Type Culture Collection (ATCC; Rockville, MD, U.S.) and cultivated in RPMI-1640 bicarbonate medium (Seromed, Berlin, Germany) supplemented with 10% heat-inactivated fetal bovine serum (Seromed), 4.0 mM L-glutamine, 50 units/mL penicillin, 50 μg/mL streptomycin and 100 μg/mL neomycin in a humidified incubator (Heraeus Cytoperm, Hanau, Germany; 5% CO₂, 37 °C, 95% humidity). Viability of cells assessed by trypan blue exclusion was > 95% for all experiments. Cell culture media yielding pH values of 7.1, 7.4 and 7.8 under the tissue culture conditions (5% CO₂) were prepared by supplementation of bicarbonate-free RPMI-1640 medium with 10, 25 and 45 mM NaHCO₃, respectively. Cells were subcultivated by trypsinization with 0.05% trypsin/0.02% ethylene diamine tetraacetic acid (EDTA) three times a week and checked for mycoplasma contamination (Mycoplasma PCR ELISA, Roche Diagnostics, Vienna, Austria).

In order to determine the effects of test compounds on cell proliferation, 6–10 serial two-fold dilutions of the compounds were pipetted into 96-well microtiter plates (Greiner, Kremsmuenster, Austria) at a volume of 100 μL medium per well followed by addition of 0.5 × 10⁴ cells at another 100 μL medium/well. Tests were performed in triplicate. Microtiter plates were incubated under tissue culture conditions and cell proliferation measured using a modified MTT (3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide) assay (EZ4U, Biomedica, Vienna, Austria) according to the manufacturer's instructions. Optical densities were measured at 450 nm using the EL × 800 microplate reader (BioTek, Bad Friedrichshall, Germany). Proliferation of each cell line in the absence of test substances was determined using separate wells and set to 100% for calculation.

4 × 10⁵ native cells were incubated with primary mouse monoclonal antibodies clone B-N6 (Cell Sciences, Canton, MA, U.S.) or clone 225 (Abcam, Cambridge, U.K.) in medium in microtiter plates (Greiner, Kremsmuenster, Austria) at 4 °C overnight for detection of NTR1 and EGFR, respectively. Cells were then washed with PBS, stained with goat anti-mouse fluorescein isothiocyanate (FITC)-labeled IgG secondary antibody (dilution 1:50, 4 °C, 20 min), washed again and analyzed using the Cytomics FC500 flow cytometer (Beckman Coulter, Fullerton, CA, U.S.) acquiring 5,000 cells per run. Antigen expression was calculated as relative fluorescence intensity = mean fluorescence antibody/mean fluorescence control.

5 × 10⁵ cells per well cultured in 6-well plates (Greiner) were trypsinized, washed with PBS, fixed in 70% ethanol at −20 °C for 30 min, washed again, resuspended in staining solution (PBS containing 20 μg/mL propidium iodide, 5 μg/mL ribonuclease A and 0.05% Nonidet P-40) and incubated at room temperature overnight. Washed cells were analyzed by flow cytometry (Cytomics FC500, Beckmann Coulter) acquiring 1 × 10⁴ cells per run. MultiCycle AV software (Phoenix Flow Systems, San Diego, CA, U.S.) was employed for calculation of cell cycle distribution from linear histograms. Percentages of cells in cell cycle phases G1/0 (resting), S (DNA synthesis) and G2M (mitotic cells) were recorded.

Cellular IL-8 production was detected in cell culture supernatants using the Strakine Tool Box ELISA Kit (Strathmann Biotec, Hamburg, Germany) according to the manufacturer's instructions.

All experiments were performed at least in duplicate and calculations were carried out using Origin Scientific Graphing and Analysis Software (Northampton, MA, U.S.) and Microsoft Excel (Redmond, CA, U.S.). Values are shown as mean ± standard deviation (SD). Statistical analysis was done using Student's t-test assuming normal distribution for the experimental measurements. Differences with * p < 0.05 were regarded as statistically significant.