SEARCH

SEARCH BY CITATION

Keywords:

  • integrins;
  • cell adhesion;
  • cell proliferation;
  • type I collagen;
  • desmoplasia;
  • malignancy;
  • metastasis

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References

The authors have previously demonstrated that α2β1 integrin-mediated pancreatic cancer cell adhesion to Type I collagen is Mg2+-dependent, inhibited by Ca2+, and that this integrin, purified from cell lysates using Type I-collagen-sepharose in Mg2+, can be eluted with Ca2+. In the present study, the authors examined the divalent cation-dependency of α2β1 integrin-mediated pancreatic cancer cell adhesion, migration and proliferation on Type I collagen, an extracellular matrix protein shown to be highly up-regulated, and to promote the malignant phenotype in vitro and in vivo. The results indicate that cells attach to Type I collagen maximally when Mg2+ is greater than 1 mM, and that addition of increasing concentrations of Ca2+ reduces this adhesion. These effects are reversible, in that previous cell attachment in Mg2+ can be reversed by adding Ca2+, and vice versa. They also demonstrate that pancreatic cancer cells migrate and proliferate on Type I collagen in Mg2+ alone, but maximally when Mg2+ is present at concentrations that promote maximal cell adhesion and Ca2+ is present at concentrations less than Mg2+. Cell adhesion and proliferation assays, as well as affinity chromatography on Type I collagen using anti-integrin function-blocking monoclonal antibodies indicate that the effects of these divalent cation shifts are mediated specifically by the α2β1 integrin. As pancreatic juice contains over 1,200-fold more Mg2+ than Ca2+ and solid tumors are characterized by increased magnesium load, these data indicate that such pathophysiological divalent cation shifts could be involved in the activation of the α2β1 integrin-mediated malignant phenotype on Type I collagen in the pancreatic cancer. © 2008 Wiley-Liss, Inc.

Integrins are a family of heterodimeric transmembrane receptor proteins that mediate the binding of cells to the extracellular matrix (ECM) as well as, in some cases, other cells.1–6 A common characteristic of all integrins is their absolute requirement for divalent cations, such as Mg2+ and Ca2+, to function.3, 4, 6–11 These cations presumably exert their effects by binding to the 3–5 putative cation-binding domains located on all integrin α subunits,12 and possibly by interacting directly with β subunits.13

We have recently demonstrated that α2β1 integrin-mediated pancreatic cancer cell adhesion to Type I collagen, an ECM protein shown to be highly upregulated in pancreatic cancer14–17 and to promote the malignant phenotype in vitro and in vivo,18–25 is promoted in 1.5 mM Mg2+ and inhibited in 1.5 mM Ca2+,21 and that the α2β1 integrin from pancreatic cancer cell lysates bound specifically to Type I collagen by affinity chromatography in 3 mM Mg2+, can be eluted with 3 mM Ca.2+26 These data are consistent with our previous observations regarding the α2β1 integrin, Type I collagen, and the various cell types involved in cutaneous wound repair,27–29 of which strong parallels to the pancreatic cancer paradigm have been described.30, 31

During normal cutaneous wound healing in vivo, shifts in the concentrations of extracellular Mg2+ and Ca2+ have been shown to occur early in the process, activating the α2β1 integrin-mediated migration of various wound healing cell types on Type I collagen, including epithelial keratinocytes.29 This shift in extracellular divalent cation concentration is characterized by increased Mg2+ and decreased Ca2+. Interestingly, similar shifts probably occur in pancreatic cancer as well as pancreatic juice, produced in the range of 1,500–2,000 mL/day, contains Mg2+ in the range of 137–244 μM and Ca2+ in the range of 0.111–0.197 μM.32 This represents a more than 1,200-fold increase of extracellular Mg2+ relative to Ca2+. As pancreatic ductal epithelial basement membranes have been shown to be discontinuous or absent in pancreatic cancer,15, 16, 33 pancreatic juice could be expected to leak into the pancreatic cancer tumor microenvironment. Additionally, and similar to cutaneous wound healing,34 solid tumors are distinguished by increased Mg2+ load, even at the expense of healthy adjacent tissue.35, 36 Collectively, these data suggest the hypothesis that pathophysiological divalent cation shifts similar to that found in cutaneous wound repair could also occur in the local pancreatic cancer tumor microenvironment, activating the α2β1 integrin-mediated malignant phenotype on Type I collagen.

In the present study, we examined the potential regulatory role of extracellular Mg2+ and Ca2+ on adhesion, migration and proliferation of pancreatic cancer cells on Type I collagen. Our results indicate that pancreatic cancer cells attach to Type I collagen maximally when Mg2+ is greater than about 1 mM, and that addition of increasing concentrations of Ca2+ reduces this Mg2+-dependent adhesion. These divalent cation-mediated effects are reversible and not detrimental to the cell, in that the previous cell attachment in Mg2+ can be subsequently reversed by the addition of Ca2+, and vice versa. We also demonstrate that pancreatic cancer cells migrate and proliferate on Type I collagen in Mg2+ alone, but maximally when Mg2+ is present at concentrations that promote maximal cell adhesion and Ca2+ is present at concentrations less than Mg2+. Cell adhesion and proliferation assays, as well as affinity chromatography on Type I collagen using anti-integrin function-blocking monoclonal antibodies indicate that these divalent cation effects are mediated specifically by the α2β1 integrin. These data indicate a potential role for extracellular Mg2+ and Ca2+ shifts in the activation of the α2β1 integrin-mediated malignant phenotype on Type I collagen in pancreatic cancer.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References

Cells

Capan-1, CFPAC, Colo-357, AsPC-1, BxPC-3, MiaPaCa-2 and Panc-1 cells were from ATCC (Rockville, MD). FG cells are a fast-growing, metastatic variant of the Colo-357 pancreatic cancer cell line.37 Cells were cultured in Dulbecco's modified Eagles' medium (DMEM) (high-glucose) supplemented with 10% fetal bovine serum in a humidified atmosphere containing 5% CO2 at 37°C.

Adhesion assays

Adhesion assays were conducted as previously described11, 21, 38 using 5 × 104 pancreatic cancer cells in tris-buffered saline (TBS) containing 1 mg/mL bovine serum albumin (BSA) supplemented with either MgCl2 or CaCl2 at the indicated concentrations and times, and 96-well plates that were previously coated with 5 μg/mL of Type I collagen. This coating concentration has been previously shown to promote maximal adhesion in the pancreatic cancer cell lines used in these studies.20 In reversibility studies, MgCl2 or CaCl2 were added to each Type I collagen-coated well at constant concentration (3.5 mM) and mixed with the cells. After 45 min of incubation at 37°C the alternate cation was titrated into the wells at the indicated concentrations and incubated at 37°C. The remainder of the assay was conducted as previously described.11, 21, 38

Cell viability studies

Viability studies were conducted using 5 × 105 pancreatic cancer cells/mL suspended in TBS supplemented with 1 mg/mL BSA and 3.5 mM CaCl2. Cells were incubated at 37°C in polypropylene tubes under nonadherent conditions. At the indicated time points, cell viability was determined using trypan-blue exclusion according to the manufacturer's instructions (Invitrogen, Carlsbad, CA).

Inhibition of adhesion assays

Inhibition of cell adhesion assays were performed as previously described.20, 27 Briefly, 96-well microtiter plates were coated with Type I collagen (5 μg/mL) and blocked with 1 mg/mL BSA. Then 5 × 104 pancreatic cancer cells were added to each well in serum-free DMEM supplemented with 1 mg/mL BSA. Function-blocking anti-integrin monoclonal antibodies (P1E6, anti-α2; P1B5, anti-α3; P5D2, anti-β1) (Millipore, Temecula, CA) of the same IgG1 isotype,39, 40 were added at a final concentration of 25 μg/mL just prior to cell addition in the serum-free medium as described above. After 45 min at 37°C, media were removed; attached cells were fixed, stained, destained, solubilized and quantified as described.20, 27

Migration assays

Migration assays were conducted using the modified Boyden chamber as previously described.20, 27, 41 Briefly, the chamber consists of 2 compartments separated by a filter, and migration was measured by counting the number of cells crossing the membrane through pores of defined size. Lower chambers were filled with serum- and Ca2+-free DMEM (Invitrogen, Carlsbad, CA) supplemented with 1 mg/mL BSA. On the basis of the adhesion data shown in Figure 1a, the normal Mg2+ concentration of the media (1 mM) was supplemented to a final concentration of 3.5 mM and Ca2+ was titrated at the indicated concentrations. 8 μm pore polycarbonate membrane filters (Neuro Probe, Gaithersburg, MD) were coated with 5 μg/mL of Type I collagen. This coating concentration has been previously shown to promote maximal α2β1 integrin-mediated cell migration with the pancreatic cancer cells used in these studies.20 Upper chambers were filled with 2.5 × 104 pancreatic cancer cells that were serum-starved 24 hr prior to assay, in media consistent with that of the lower chamber. Lower chamber final volumes were 30 μL and the upper chambers were 50 μL. The entire apparatus was then incubated 18–22 hr at 37°C. After the incubation period, the filters were fixed in methanol and stained with 0.5% toluidine blue in 3.7% formaldehyde. Excess stain was washed away with water, the attached cells on the upper side of the filters were mechanically removed using wet, cotton-tipped applicators, and the migratory cells on the underside of the filters were quantitated by counting 5 high-powered fields (250 × magnification) per well using an inverted light microscope (Olympus BH 2).

thumbnail image

Figure 1. The effect of divalent cations on α2β1 integrin-mediated pancreatic cancer cell adhesion on Type I collagen. (a) 5.0 × 104 of the indicated pancreatic cancer cells were added to each well of 96-well microtiter plates previously coated with 5 μg/mL Type I collagen in the presence of increasing concentrations of Mg2+ in TBS supplemented with 1 mg/mL BSA for 1 hr at 37°C as described in “Material and Methods.” Attached cells were fixed, stained, destained and quantitated by measuring the absorbance at 595 nm. The results are expressed as % Max, comparing all 8 cell lines, and represent the mean ± SEM of 3 experiments done in duplicate. One hundred percent is the maximum mean absorbance obtained at 595 nm including all cell lines tested. (b) Adhesion assays on Type I collagen were conducted as described in (a) above in the presence of 3.5 mM Mg2+ with a titration of Ca2+ at the indicated concentrations. The results are expressed as % Max comparing all 8 cell lines, and represent the mean ± SEM of 3 experiments done in duplicate. One hundred percent is the maximum mean absorbance obtained at 595 nm including all cell lines tested.

Download figure to PowerPoint

Proliferation assays

Proliferation assays were conducted as previously described.20, 37 Briefly, 96-well polystyrene culture dishes, not treated for tissue culture, were coated with Type I collagen at 5 μg/mL. This coating concentration has been previously shown to promote maximal cell proliferation on Type I collagen for the pancreatic cancer cell lines used in these studies.20 Twenty-four hour, serum-starved pancreatic cancer cells (5 × 103/well) were cultured under serum-free conditions on Type I collagen over a 3-day time course. The normal Mg2+ concentration of the media (1 mM) was supplemented to a final concentration of 3.5 mM, and Ca2+ was titrated at the indicated concentrations. Triplicate proliferation determinations were quantified by measuring the absorbance at 450 nm and subtracting the value obtained for each cell line on Type I collagen at initial seeding using CellTiter 96 Aqueous One Solution Cell Proliferation Assay™ reagent according to the manufacturer's instructions (Promega, Madison, WI). This reagent is composed of a novel tetrazolium compound [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, innersalt; MTS], and an electron coupling reagent (phenazine ethosulfate; PES). For inhibition of proliferation, pancreatic cancer cell proliferation was determined after 72 hr as described earlier in the presence or absence of the indicated function-blocking anti-integrin monoclonal antibodies (P1E6, anti-α2; P1B5, anti-α3; P5D2, anti-β1) at 25 μg/mL. All antibodies are the same IgG1 isotype,39, 40 and were added at the time of initial cell seeding or 24 hr later as indicated.

Affinity chromatography

Type I collagen (Chemicon International, Temecula, CA) was coupled to Cyanogen bromide-activated Sepharose 4B according to the manufacturer's instructions (Sigma Aldrich, St. Louis, MO). Coupling concentration: Type I collagen-sepharose—5.2 mg/mL of packed sepharose beads. Pancreatic cancer cell lines were cell-surface biotinylated using EZ-Link™ Sulfo-NHS-LC-Biotin [sulfosuccinimidyl-6-(biotinamido)] hexanoate according to the manufacturer's instructions (Pierce Biotechnology, Rockford, IL) as previously described.26, 42 Cell lysates were prepared in TBS supplemented with 100 mM octyl-β-D-glucopyranoside (OG) (CalBiochem, La Jolla, CA) and 3 mM MgCl2. After 30 min incubation on ice and subsequent centrifugation, equivalent volumes of cleared lysates were incubated overnight at 4°C with 25 μg of function-blocking monoclonal antibodies directed against the indicated integrin subunits. All antibodies used (P1E6, anti-α2; P1B5, anti-α3; P5D2, anti-β1) are the same IgG1 isotype.39, 40 Lysates were subsequently incubated with Type I collagen-sepharose (0.7 mL) that was previously equilibrated with wash buffer containing TBS supplemented with 50 mM OG and 3 mM MgCl2. Incubations periods were 2 hr at room temperature or overnight at 4°C. Flow-through extract was collected and the columns were washed with 10 column-volumes of wash buffer. Elution was conducted with TBS supplemented with 50 mM OG and 3 mM CaCl2. Fractions were subsequently analyzed by electrophoresis and immunoblotting as described below.

Immunoblotting

Immunoblotting was conducted as previously described.26, 28 Briefly, eluted fractions were separated on 12% NU-PAGE gels under nonreducing conditions, and transferred to nitrocellulose. After blocking with 3% BSA in PBS/0.1% tween-20, membranes were incubated with horseradish peroxidase-conjugated streptavidin (1:25,000 dilution) (Pierce Biotechnology), and peroxidase activity of the biotin–streptavidin complexes detected using chemiluminescence according to the manufacturer's instructions (Amersham Biosciences, Little Chalfont, UK). Immunoblotting results were quantitatively analyzed with a digital imaging system (Alpha Innotech, San Leandro, CA). The intensities of the bands were assigned integrated density values, which represent the sum of all pixel values in the box. Relative expression values were generated by comparing the values obtained using lysates in the presence of antibody with control lysates in the absence of antibody, which was arbitrarily assigned the value 1.

To determine the effect of trypsinization on α2β1 integrin expression, 24 hr, serum-starved pancreatic cancer cells grown on 6-well tissue culture plates to ∼70% confluency were either lysed directly (1 mL/well), or detached by trypsinization (Invitrogen), followed by washing with soybean trypsin inhibitor (STI) (Sigma-Aldrich, St. Louis, MO). This procedure is consistent with the previously described protocol used in the present studies for preparation of cells for adhesion, migration and proliferation assays.11, 21, 38 Cells collected using trypsin and STI were subsequently lysed (1 mL/well). The lysis buffer was prepared as follows: 250 mM Tris, pH 7.4, 0.25% Igepal CA-630 (NP-40) (Sigma-Aldrich), 1 mM MgCl2 and 1 mM CaCl2. Cell lysates were frozen, thawed and cleared by centrifugation for 15 min at 16,000g. Supernatants were transferred to fresh tubes and analyzed for integrin expression levels by immunoblotting with a polyclonal antibody directed against the α2, Type I collagen-binding integrin subunit (1:1,000 dilution) (Millipore, Temecula, CA), with a monoclonal antibody (Clone P4C10) directed against the β1 integrin subunit (1:500 dilution) (Millipore), and with a monoclonal antibody directed against β-actin (Clone AC-15) (1:2,000 dilution) (Sigma-Aldrich) to control for equivalent loading of the paired lysates for each cell line. After washing, membranes were exposed to peroxidase-conjugated goat anti-mouse or anti-rabbit IgG (1:2,000 dilution) (Amersham Biosciences, Buckinghamshire, England), and the peroxidase detected using chemiluminescence and autoradiography as previously described.21 Densitometry analyses were conducted as described for the affinity chromatography immunoblotting experiments by comparing lysates prepared directly from tissue culture plastic to those prepared after the trypsinization procedure.

Statistical analysis

Statistical significance was determined using 2-tailed Student ttests.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References

Ca2+ reverses the Mg2+-dependent, α2β1 integrin-mediated attachment of pancreatic cancer cells to type I collagen substrates

Of the 4 known collagen-binding integrins, including α1β1, α2β1, α10β1 and α11β1,43, 44 only the expression of the α1β1 and α2β1 integrins has been examined in pancreatic adenocarcinoma so far.25 Results from these previous studies indicate that both the α1β1 and α2β1integrins are expressed in vivo at both the mRNA and protein levels in pancreatic cancer. In vitro, α2β1 integrin expression has been shown in 18/19 cell lines (95%), while α1β1 integrin expression has been demonstrated in 7/13 cell lines (54%). We have previously shown using inhibition of adhesion assays with function-blocking monoclonal antibodies directed against specific integrin subunits in both two- and three-dimensions, in inhibition of migration assays, and by affinity chromatography and subsequent immunoprecipitation analyses, that the α2β1 integrin specifically and exclusively mediates the interaction of Panc-1, BxPC-3, CFPAC, Capan-1, Colo-357, FG and AsPC-1 pancreatic cancer cell lines with Type I collagen.20, 21, 26, 37 These studies indicate further that the α1β1 integrin is not involved in pancreatic cancer cell interactions with Type I collagen.

We have also previously shown in cell adhesion assays with Colo-357, Panc-1 and Capan-1 cells on Type I collagen in the presence of either 1.5 mM Mg2+ or 1.5 mM Ca2+, that α2β1 integrin-mediated pancreatic cancer cell attachment is Mg2+-dependent and inhibited by Ca2+.21 In the present study, we examined the Mg2+-dependency of pancreatic cancer cell adhesion to Type I collagen in titration experiments. As shown in Figure 1a, our results indicate that maximal adhesion occurs at concentrations of Mg2+ greater than about 1 mM for all cell lines tested except MiaPaCa-2 cells. MiaPaCa-2 cells are the only known α2β1 integrin-negative pancreatic cancer cell line, and these cells do not attach to Type I collagen.20, 26, 45, 46 Furthermore, the attachment of pancreatic cancer cells to Type I collagen substrates in the presence of optimal Mg2+ concentrations (3.5 mM) decreased with increasing Ca2+ concentration (Fig. 1b).

Because Mg2+-dependent, α2β1 integrin-mediated pancreatic cancer cell adhesion to Type I collagen was inhibited by increasing concentrations of Ca2+, we next asked whether the effects of these divalent cations on α2β1 integrin-mediated pancreatic cancer cell adhesion were reversible, as previously shown by our laboratory for WI38 human lung fibroblasts.27 The panels on the right side of Figure 2 demonstrate with Panc-1, Colo-357 and FG cells, that previous cell attachment using optimal concentrations of extracellular Mg2+ (3.5 mM) could be reversed by the addition of Ca2+, and the extent of this detachment increased with increasing extracellular Ca2+ concentration. Maximal cell detachment varied for each cell line, and was between 50 and 70% of that shown with 3.5 mM Mg2+ only. Similar results were also obtained with BxPC-3 cells (data not shown). To test whether the inhibitory effect of Ca2+ was detrimental to the cells, we preincubated cells in the presence of Ca2+ (3.5 mM) and subsequently exposed them to a Mg2+ titration. The panels on the left side of Figure 2 indicate that, while no cell attachment was observed on Type I collagen in the presence of Ca2+ alone, subsequent addition of increasing concentrations of extracellular Mg2+ yielded increased attachment. Similar results were also obtained with BxPC-3 cells (not shown). The maximum mean absorbance achieved for each cell line for each experiment (attachment/detachment) is also shown parenthetically above each curve in Figure 2. These absorbance data indicate that the extent of Mg2+-dependent cell attachment after a 45 min preincubation in 3.5 mM Ca2+ varied for each cell line, ranging between 20 and 100% of that observed in Mg2+ only.

thumbnail image

Figure 2. The effects of divalent cations on pancreatic cancer cell adhesion on Type I collagen are reversible. 5.0 × 104 Panc-1, FG, and Colo-357 cells were added to each well of microtiter plates coated with 5 μg/mL Type I collagen and incubated for 45 min at 37°C in the presence of either 3.5 mM Mg2+ or 3.5 mM Ca2+. After the initial incubation a titration of the alternate cation was added to the wells at the indicated concentrations and reincubated for another 45 min. Attached cells were fixed, stained, destained and quantitated by measuring the absorbance at 595 nm. The results are expressed as % Max for each cell line for each divalent cation titrated, and represent the mean ± SEM of 3 experiments done in quadruplicate. One hundred percent is the maximum mean absorbance (indicated parenthetically above each curve) obtained at 595 nm for each cell line tested for each divalent cation titrated.

Download figure to PowerPoint

A number of control experiments were also conducted to examine the divalent cation-dependent reversibility of pancreatic cancer cell adhesion on Type I collagen in further detail. First, it is possible that addition of Ca2+ after 45 min in 3.5 mM Mg2+ is preventing additional adhesion during the second 45 min incubation, but not reversing previous Mg2+-dependent adhesion that occurred during the first 45 min incubation. In cell adhesion assays with Panc-1, BxPC-3 and FG cells, we compared the absorbance values at 595 nm obtained after 45 min in 3.5 mM Mg2+ to those obtained after 90 min. Figure 3a demonstrates that there is no significant difference in Mg2+-dependent BxPC-3 or FG cell adhesion between 45 and 90 min. For Panc-1 cells, there is a statistically significant increase in cell adhesion between 45 and 90 min (p = 0.035), but this difference is less than 7%. These data collectively indicate that Ca2+ is detaching cells and not preventing additional Mg2+-dependent adhesion during the second 45 min incubation.

thumbnail image

Figure 3. Examination of the reversibility of α2β1 integrin-mediated interactions with Type I collagen by divalent cations. (a) Cell adhesion assays were conducted in the presence of 3.5 mM Mg2+ on Type I collagen as described for Figures 1 and 2 and in “Material and Methods.” The absorbance after 45 min was compared to that obtained after 90 min. Results, expressed as the percentage of the 90 min absorbance value for each cell line, represent the mean ± SEM of 3 experiments with 12 replicates for each cell line per experiment. *p = 0.035 (b) Cell viability studies using trypan-blue exclusion were conducted under nonadherent conditions in the presence of 3.5 mM Ca2+ over a 90 min time course as described in “Material and Methods.” Results, expressed as the percentage of T = 0, represent the mean ± SEM of 5 experiments. (c) Immunoblotting studies were conducted to determine the effect of the trypsinization procedure on the stability of the α2β1 collagen-binding integrin as described in “Material and Methods.” Representative immunoblots are shown for the α2 and β1 integrin subunits from lysates of Panc-1, BxPC-3 and FG cells prepared with (+) and without (−) trypsinization and soybean trypsin inhibitor (STI) treatment. Immunoblotting results for β-actin are also shown to demonstrate equivalent protein loading for each pair of samples from each cell line. This experiment was conducted twice with similar results.

Download figure to PowerPoint

Second, the maximum absorbance obtained in the attachment assays for each cell line when Mg2+ is added after a 45 min preincubation in 3.5 mM Ca2+ compared to Mg2+-only adhesion varies between 20 and 100%, depending on the cell line. These data suggest that Ca2+ may, indeed, have a detrimental effect on some pancreatic cancer cell lines. We, therefore, conducted cell viability studies over a 90 min time course using trypan-blue exclusion with pancreatic cancer cells suspended in the presence of 3.5 mM Ca2+. Figure 3b indicates that there is no significant difference in Panc-1, BxPC-3 or FG cell viability throughout the 90 min time course. These data suggest that the inability of some of the cell lines to achieve comparable cell adhesion after a preincubation in Ca2+ compared to that observed Mg2+ alone is not due to detrimental effects of Ca2+ on cell viability.

Finally, to determine the effect of the trypsinization procedure on the stability of the α2β1 collagen-binding integrin, we compared control cell lysates prepared directly from tissue culture dishes to those prepared after trypsin and STI treatment as previously described for the preparation of cells for cell adhesion, migration and proliferation assays.11, 21, 38 Immunoblotting results shown in Figure 3c demonstrate for Panc-1, BxPC-3 and FG cells, that there is essentially no difference in the expression of the α2, Type I collagen-binding integrin subunit or the β1 integrin subunit when cells are lysed directly on the tissue culture dish compared to cells lysed after the trypsinization procedure. Subsequent densitometric analyses confirm that there are no statistically significant differences in α2 (p ≥ 0.146) or β1 (p ≥ 0.725) integrin subunit expression in any of the cell lines tested. These data suggest further that the subsequent use of these trypsinized cells would not be expected to have any adverse effects on α2β1 integrin-mediated interactions with Type I collagen in either short- or long-term culture.

Taken together, these data indicate that α2β1 integrin-mediated pancreatic cancer cell attachment to Type I collagen is Mg2+-dependent and inhibited by Ca2+. These data indicate further that the effect of divalent cations on integrin-mediated pancreatic cancer cell adhesion are reversible to varying degrees depending on the particular cell line, and have no apparent detrimental effect on cell viability through the experimental time course.

Divalent cation shifts promote maximal α2β1 integrin-mediated pancreatic cancer cell migration on type I collagen

The cell adhesion and reversibility data together with previous affinity chromatography and immunoprecipitation results demonstrating that Ca2+ can elute the α2β1 integrin from pancreatic cancer cell lysates bound specifically to Type I collagen-sepharose in Mg2+26 suggested that, as we have previously shown for fibroblasts, keratinocytes, endothelial cells and macrophages,27–29 Ca2+ and Mg2+ could also be involved in the modulation of α2β1 integrin-mediated pancreatic cancer cell migration on Type I collagen. We have previously demonstrated using anti-integrin function-blocking monoclonal antibodies in inhibition of cell migration assays, that the α2β1 integrin specifically and exclusively mediates pancreatic cancer cell migration on Type I collagen, with no involvement from the α1β1 integrin.20 In the same modified Boyden chamber migration assays, we found that while BxPC-3 cells were migratory on Type I collagen using concentrations of extracellular Mg2+ that supported maximal cell adhesion (3.5 mM) and not in Ca2+ alone (not shown), maximal migration was observed when these cations were utilized in combination as long as Ca2+ was present at concentrations less than Mg2+ (3.5 mM) (Fig. 4). As the extracellular Ca2+ concentration increased over that of Mg2+, cell migration rapidly declined. Figure 4 also shows the dramatic morphologic differences in BxPC-3 pancreatic cancer cells observed under these different divalent cation conditions. For example, in the presence of 3.5 mM Mg2+ only, cells were singular and well spread. These data are consistent with the known Ca2+-dependency of E-cadherin-mediated cell–cell adhesion.47 Addition of 0.47 mM Ca2+ to 3.5 mM Mg2+-containing medium resulted in nearly a 2-fold increase in cell migration and loose cell–cell contacts. Addition of extracellular Ca2+ concentrations greater than Mg2+ resulted in increased cell–cell adhesion and decreased cell migration on Type I collagen. Together, these data indicate that even subtle changes in the concentrations of Mg2+ and Ca2+ can effect pancreatic cancer cell migration.

thumbnail image

Figure 4. The effect of divalent cations on pancreatic cancer cell migration on Type I collagen. BxPC-3 cell migration was determined and quantitated as described in “Material and Methods” using 5.0 × 104 cells/well in 3.5 mM Mg2+ plus a titration of the Ca2+ at the indicated concentrations on Type I collagen (5 μg/mL)-coated filters. The results, expressed as % Max for each cell line, represent the mean ± SEM of 3 experiments done in triplicate. One hundred percent for BxPC-3 cells = 416 cells ± SEM/high-powered field). Light photomicrographs of BxPC-3 cells shown under 2 different magnifications are representative examples of migration in modified Boyden Chamber assays conducted on Type I collagen (5 μg/mL)-coated filters under the indicated divalent cation conditions. Photomicrographs were obtained using an Olympus BX-60 microscope equipped with a Spot Digital Imaging package (Diagnostic Instruments, Burlingame, CA). Bar, 100 μm and 20 μm, as indicated.

Download figure to PowerPoint

Divalent cation shifts promote maximal α2β1 integrin-mediated pancreatic cancer cell proliferation on type I collagen

Because Mg2+ is essential for virtually all major biological processes, including DNA, RNA, and protein synthesis, as well as cell division, membrane transport and oxidative phosphorylation,35, 36 we examined the effects of divalent cation shifts on the proliferation of pancreatic cancer cells on Type I collagen. As shown in Figure 5, Mg2+ alone supports cell proliferation over a 72-hr time course in BxPC-3, CFPAC and FG cells. Remarkably, the addition of Ca2+ increased cell proliferation by as much as 2-fold, as long as Ca2+ was present at concentrations less than Mg2+ (3.5 mM). As the Ca2+ concentration exceeded that of Mg2+, cell proliferation declined markedly. Figure 5 also shows the morphologic differences in BxPC-3 pancreatic cancer cell proliferation after 72 hr in the presence of these different divalent cation conditions. And like the migration data shown in Figure 4, in the presence of 3.5 mM Mg2+ only, where Ca2+-dependent E-cadherin-mediated cell–cell adhesion is inhibited,47 cells were singular and well spread. Addition of 0.47 mM Ca2+ to 3.5 mM Mg2+-containing medium resulted in nearly a 2-fold increase in cell proliferation and loose cell–cell contacts. Addition of extracellular Ca2+ concentrations greater than Mg2+ resulted in increased cell–cell adhesion and decreased cell proliferation on Type I collagen.

thumbnail image

Figure 5. The effect of divalent cations on pancreatic cancer cell proliferation on Type I collagen. BxPC-3, CFPAC and FG cell proliferation was determined and quantitated as described in “Material and Methods” using 5.0 × 103 cells/well in 3.5 mM Mg2+ plus a titration of the Ca2+ at the indicted concentrations on Type I collagen (5 μg/mL)-coated plates over a 72 hr time course. The results, expressed as % Max for each cell line, represent the mean ± SEM of 3 experiments done in duplicate. ▪, 24 hr; ▴, 48 hr; •, 72 hr. One hundred percent is the maximum mean absorbance obtained at 450 nm for each cell line throughout the time course as described in “Material and Methods.” For each cell line, maximum A450 was obtained at 72 hr with Ca2+ concentrations less than 3.5 mM. Light photomicrographs of BxPC-3 cells shown are representative examples of proliferation assays conducted on Type I collagen (5 μg/mL)-coated filters under the indicated divalent cation conditions after 72 hr. Photomicrographs were obtained using an Olympus BX-60 microscope equipped with a Spot Digital Imaging package (Diagnostic Instruments, Burlingame, CA). Bar, 50 μm.

Download figure to PowerPoint

The α2β1 integrin specifically mediates the divalent cation-dependent effects of pancreatic cancer cells on type I collagen

We have previously demonstrated with function-blocking monoclonal antibodies directed against particular integrins and integrin subunits, that the α2β1 integrin specifically and exclusively mediates pancreatic cancer cell adhesion on Type I collagen in both two- and three-dimensions, as well as cell migration.20, 26 Additionally, affinity chromatography and immunoprecipitation experiments using pancreatic cancer cell lines that express both the α2β1 and α1β1 integrins indicate further that only the α2β1 integrin binds Type I collagen.26 As a further demonstration of the specificity of divalent cation-dependent, α2β1 integrin-mediated binding of pancreatic cancer cells to Type I collagen, we conducted inhibition of affinity chromatography, adhesion and proliferation assays using the anti-integrin function-blocking monoclonal antibodies referred earlier. Figures 6a and 6b demonstrate by immunoblotting and densitometric analyses, that the α2β1 integrin, bound to Type I collagen-sepharose from cell-surface biotinylated extracts of BxPC-3 cells in 3 mM Mg2+ and eluted with 3 mM Ca2+, could be significantly inhibited with monoclonal antibodies directed against the α2 and β1 integrin subunits (p < 0.05), but not with an antibody directed against the α3 integrin subunit. All 3 antibodies are the same IgG1 isotype.39, 40 Figure 6c demonstrates further that the anti-α2 and anti-β1 integrin subunit antibodies inhibited BxPC-3 cell adhesion on Type I collagen in the presence of optimal divalent cation concentrations (3.5 mM Mg2+/0.5 mM Ca2+). The extent of this inhibition for BxPC-3 cells was similar to that previously shown using media with normal divalent cation concentrations (1 mM Mg2+/1.8 mM Ca2+).20 By contrast, the anti-α3 antibody had no effect. Figure 6d demonstrates that addition of the anti-α2 and anti-β1 integrin subunit antibodies, either at the time of initial cell seeding or 24 hr later, inhibited BxPC-3 cell proliferation on Type I collagen. By contrast, the anti-α3 antibody had no effect. The α2 integrin subunit antibody inhibited cell adhesion, proliferation and α2β1 integrin binding to Type I collagen in affinity chromatography to a lesser extent than the β1 integrin subunit antibody. These results are in agreement with those obtained with other pancreatic cancer cell lines as well as fibroblasts.20, 27 Collectively, these data indicate that the α2β1 integrin specifically mediates the divalent cation-dependent effects on adhesion, migration and proliferation of pancreatic cancer cells on Type I collagen.

thumbnail image

Figure 6. The α2β1 integrin specifically mediates divalent cation-dependent interactions of pancreatic cancer cells with type I collagen. Conduction of affinity chromatography using cell surface biotinylated extracts of BxPC-3 cells and Type I collagen-sepharose (a) and subsequent densitometric analyses (b) as described in “Material and Methods” indicate that the α2β1 integrin, bound in Mg2+ and eluted with Ca2+, is significantly inhibited after pre-incubation with 25 μg of the indicated monoclonal antibodies directed against the α2 and β1 integrin subunits (*p < 0.05), but not with an antibody directed against the α3 integrin subunit. Representative immunoblots are shown in (a). Densitometry results shown in (b) are expressed as relative increase/decrease compared to no antibody controls, and represent the mean ± SEM of 3 experiments. Two-tailed Student's t-tests were used to identify significant differences (*p < 0.05) compared to no antibody controls. (c) Inhibition of cell adhesion assays were conducted as described in “Material and Methods.” Addition of the anti-α2 and anti-β1 integrin subunit antibodies (25 μg/mL) at the time of initial cell seeding inhibited BxPC-3 cell adhesion on Type I collagen in the presence of optimal divalent cation concentrations (3.5 mM Mg2+/0.5 mM Ca2+). By contrast, the anti-α3 antibody had no effect. The results, expressed as percent Control (in the absence of antibody), represent the mean absorbance at 595 nm ± SEM of 4 experiments done in triplicate. Two-tailed Student's t-tests were used to identify significant differences (*p < 0.05) compared to control. (d) Inhibition of cell proliferation assays conducted as described in “Material and Methods” indicate that addition of the anti-α2 and anti-β1 integrin subunit antibodies (25 μg/mL), either at the time of initial cell seeding (black bars) or 24 hr later (gray bars), inhibited BxPC-3 cell proliferation on Type I collagen in the presence of optimal divalent cation concentrations (3.5 mM Mg2+/0.5 mM Ca2+). By contrast, the anti-α3 antibody had no effect. The results, expressed as % Control (in the absence of antibody), represent the mean absorbance at 450 nm with the value obtained for each cell line at initial seeding subtracted ± SEM of 3 experiments done in triplicate. Two-tailed Student's t-tests were used to identify significant differences (*p < 0.05) compared to control.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References

In the present study, we demonstrate that Ca2+ reverses Mg2+-dependent, α2β1-mediated binding in cell adhesion and affinity chromatography experiments with cell-surface biotinylated pancreatic cancer cell lysates, and markedly affects the migration and proliferation of pancreatic cancer cells on Type I collagen substrates. The observation that pancreatic cancer cells were still capable of Mg2+-dependent adhesion after preincubation with Ca2+ alone suggests that exposure to alterations in the concentrations of these extracellular divalent cations is not detrimental to the cell or to the α2β1 integrin. And while pancreatic cancer cells migrate and proliferate on Type I collagen in the presence of Mg2+ alone and not in Ca2+ alone, the inclusion of both divalent cations with optimal extracellular Mg2+ (3.5 mM) and Ca2+ present at concentrations less than Mg2+ resulted in maximal cell migration and proliferation. Together, these results indicate that the concentrations of extracellular Mg2+ and Ca2+ regulate α2β1 integrin function on Type I collagen in vitro and suggest that divalent cation shifts could potentially influence the malignant phenotype of pancreatic cancer cells in the Type I collagen-rich local tumor microenvironment in vivo.14–17, 48, 49

Consistent with our previous inhibition of adhesion studies using media containing normal divalent cation concentrations (1 mM Mg2+/1.8 mM Ca2+),20, 26, 41 the present studies demonstrate that antibodies directed against the α2β1 integrin inhibit pancreatic cancer cell adhesion on Type I collagen in the presence of optimal divalent cation concentrations (3.5 mM Mg2+/0.5 mM Ca2+) as well. For the first time, we also demonstrate that antibodies directed against the α2β1 integrin also inhibit pancreatic cancer cell proliferation on Type I collagen. It is especially noteworthy that cell proliferation is inhibited even when antibodies are added 24 hr after initial cell seeding on Type I collagen. Additionally, we demonstrate for the first time, that antibodies directed against α2β1 inhibit Mg2+-dependent integrin binding to Type I collagen in affinity chromatography.

As stated earlier, there are 4 known collagen-binding integrins, which include α1β1, α2β1, α10β1 and α11β1.43, 44 Our present and previously published data indicate clearly that the α2β1 integrin specifically and exclusively mediates pancreatic cancer cell interactions with Type I collagen, with no direct involvement of the α1β1 integrin.20, 26, 41 However, α10β1 and α11β1 integrin expression has not yet been examined in pancreatic cancer. And while it cannot be completely ruled out, several lines of evidence suggest that α10β1 and α11β1 are not involved in pancreatic cancer cell interactions with Type I collagen.

First, our affinity chromatography and immunoprecipitation data using cell surface-biotinylated pancreatic cancer cell extracts indicate that only the α2β1 integrin binds to Type I collagen, even using cell lines that also express the α1β1 integrin.26 Neither the α10 nor α11 integrin subunits, which migrate at different relative masses compared to the α2 and α1 integrin subunits under nonreducing conditions,44 are present in the eluates after affinity chromatography of pancreatic cancer cell lysates on Type I collagen (at least in sufficient quantity to be detected by chemiluminescence). Second, α10 and α11 integrin expression are most often associated with bone and cartilage.43, 44 And third, recent quantitative RT-PCR analyses of cancer cell lines for collagen receptor expression indicate that α1β1, α2β1, α10β1 and α11β1 are abundantly expressed in sarcoma-derived cell lines, whereas most carcinoma-derived cell lines express only the α1β1 and α2β1 integrins.50 However, pancreatic adenocarcinoma cell lines were not investigated in that study. Together with the known affinity differences of the monoclonal antibodies used in these and our previous studies,20, 26, 27, 39–41 where the β1 antibody has a higher affinity than the α2 antibody, these data collectively suggest that only the α2β1 integrin is involved in pancreatic cancer cell interactions with Type I collagen.

In agreement with our previous studies of αvβ1-mediated attachment of IMR 32 human neuroblastoma cells to Arg-Gly-Asp (RGD)-containing peptides,11 the attachment of pancreatic cancer cells to Type I collagen substrates in the presence of 3.5 mM Mg2+ decreased as the Ca2+ concentration increased. Also consistent with our previous studies of α2β1 integrin-mediated WI38 fibroblast adhesion on Type I collagen substrates,27 previous pancreatic cancer cell attachment in Mg2+ could be reversed by the addition of Ca2+, and the extent of this detachment increased with increasing Ca2+ concentration. And like our previously published cell migration results using human wound healing cell types, including keratinocytes, endothelial cells, fibroblasts and macrophages,27–29 we found that while BxPC-3 cells were migratory on Type I collagen in Mg2+ alone and not in Ca2+ alone, maximal migration was observed when these cations were utilized in combination with Mg2+ present at concentrations that promote maximal cell adhesion (3.5 mM) and Ca2+ present at concentrations less than Mg2+. For example, maximal WI38 fibroblast adhesion on Type I collagen occurred in the presence of 1.5 mM Mg2+, and maximal migration occurred at 1.5 mM Mg2+ with Ca2+ present at less than 1.5 mM.27 Maximal HaCaT keratinocyte adhesion on Type I collagen occurred in the presence of 3.3 mM Mg2+, and maximal migration occurred at 3.3 mM Mg2+ with Ca2+ present at less than 3.3 mM.28

These previous studies of the effects of divalent cations on integrin-mediated interactions with the ECM examined only cell adhesion, and migration. And while it is well-established that the processes of DNA, RNA, protein synthesis and cell division are Mg2+-dependent,35, 36 to the best of our knowledge, this is the first direct demonstration of the differential effects of altered levels of extracellular Mg2+ and Ca2+ on the α2β1 integrin-mediated proliferation of pancreatic cancer cells on Type I collagen.

Under normal physiological conditions, the extracellular levels of total magnesium and calcium are about 1.0 and 2.5 mM, respectively.27–29 We have previously shown in porcine and rat partial- and full-thickness models of cutaneous wound repair, however, that the local concentrations of extracellular magnesium and calcium shift early in the wound healing response, with decreased calcium and increased magnesium.27–29 This pathophysiological shift activates the α2β1 integrin-mediated migratory phenotype of keratinocytes, fibroblasts, endothelial cells and macrophages on Type I collagen. Our results indicate that, similar to these wound healing cell types, pancreatic cancer cells respond similarly to such divalent cation shifts, including increased α2β1 integrin-mediated adhesion, migration and proliferation on Type I collagen.

It is not obvious whether such pathophysiological divalent cation shifts occur in pancreatic cancer. As mentioned earlier, pancreas juice contains Mg2+ at a 1,200-fold excess relative to Ca2+.32 Additionally, solid tumors are characterized by increased Mg2+ load, even at the expense of healthy adjacent tissue.35, 36 While it is not yet clear to what extent the extracellular divalent cation concentrations are altered in the tumor microenvironment in pancreatic cancer in vivo, these data collectively suggest that changes in the extracellular concentrations of Mg2+ and Ca2+ similar to that seen during cutaneous wound repair probably occur, at least to some extent. Our data indicate that such shifts would activate the α2β1 integrin-mediated malignant phenotype on Type I collagen. Furthermore, the effect of divalent cation shifts may not be limited to cancer cells in the tumor microenvironment, as other cell types known to be localized in the tumor stroma, including inflammatory cells, endothelial cells and fibroblasts,30 have also been shown to react similarly to shifts in divalent cation concentrations.27–29

We have previously shown that the increase in α2β1 integrin-mediated keratinocyte migration on Type I collagen fostered in the presence of increased Mg2+ and decreased Ca2+ is associated mechanistically with downregulated expression and localization of Ca2+-dependent E-cadherin-mediated cell–cell adhesion.28 Similarly, we have also shown that increased extracellular Mg2+ and decreased Ca2+ down-regulate E-cadherin-mediated cell–cell adhesion and increases the migration of FG pancreatic cancer cells on fibronectin, an ECM substrate that promotes strong E-cadherin-mediated cell–cell contacts but not significant cell migration in media containing normal extracellular concentrations of Mg2+ and Ca2+.21 These data are in agreement with our present observations on Type I collagen, where increased Mg2+ and decreased Ca2+ results in increased pancreatic cancer cell migration associated with loose or absent cell–cell contacts (Fig. 4). These data also suggest that the effects of divalent cations may apply to integrin-mediated interactions with other ECM components in the pancreatic cancer tumor microenvironment, and are the subject of current studies.

Most of the previous in vivo studies of the effects of divalent cations are related to cutaneous wound repair. Recently, however, studies of the influence of extracellular divalent cations on colon cancer cell adhesion in a murine transplantable tumor model demonstrated that Mg2+ and Mn2+ stimulated tumor formation to 96 and 92%, respectively, but Ca2+ reduced tumor formation to 56% with no apparent affect on serum Ca2+.51, 52 Together with our results, these data collectively suggest that during pancreatic cancer progression, a resulting shift in divalent cation concentrations in the local tumor microenvironment, derived from pancreatic juice or other physiological factors, and sharing many similarities to the wound healing microenvironment,30, 31 could activate the α2β1 integrin-mediated malignant phenotype on Type I collagen.

A common problem encountered during surgical resection of pancreatic cancer tumors is a high incidence of postoperative liver metastasis, local recurrence and peritoneal tumor dissemination (38–86%, depending on the diagnostic methodology).53, 54 Our present data suggest that, as with colon cancer, irrigation of the peritoneum with low concentrations of CaCl2 during pancreatic cancer tumor resection may potentially and significantly reduce α2β1 integrin-mediated tumor cell dissemination. Future studies will examine this hypothesis in our fluorescent orthotopic murine models of pancreatic cancer.55–60

In conclusion, we demonstrate that even subtle changes in the concentrations of extracellular Mg2+ and Ca2+, which probably occur locally in the pancreatic cancer tumor microenvironment in vivo, can markedly increase the α2β1 integrin-mediated interactions of pancreatic cancer cells with Type I collagen. It may be important to consider these pathophysiological divalent cation conditions when designing therapeutics targeting the α2β1 integrin-mediated malignant phenotype in pancreatic cancer.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References
  • 1
    Hemler ME. VLA proteins in the integrin family: structures, functions, and their role on leukocytes. Annu Rev Immunol 1990; 8: 365400.
  • 2
    Hynes RO. Integrins: bidirectional, allosteric signaling machines. Cell 2002; 110: 67387.
  • 3
    Ruoslahti E. Integrins. J Clin Invest 1991; 87: 15.
  • 4
    Ruoslahti E,Noble NA,Kagami S,Border WA. Integrins. Kidney Int Suppl 1994; 44: S17S22.
  • 5
    Springer TA. Adhesion receptors of the immune system. Nature 1990; 346: 42534.
  • 6
    Ruoslahti E,Pierschbacher MD. New perspectives in cell adhesion: RGD and integrins. Science 1987; 238: 4917.
  • 7
    Dransfield I,Hogg N. Regulated expression of Mg2+ binding epitope on leukocyte integrin α subunits. Embo J 1989; 8: 375965.
  • 8
    Fitzgerald LA,Phillips DR. Calcium regulation of the platelet membrane glycoprotein IIb-IIIa complex. J Biol Chem 1985; 260: 1136674.
  • 9
    Fujimura K,Phillips DR. Calcium cation regulation of glycoprotein IIb-IIIa complex formation in platelet plasma membranes. J Biol Chem 1983; 258: 1024752.
  • 10
    Hynes RO. Targeted mutations in cell adhesion genes: what have we learned from them? Dev Biol 1996; 180: 40212.
  • 11
    Kirchhofer D,Grzesiak J,Pierschbacher MD. Calcium as a potential physiological regulator of integrin-mediated cell adhesion. J Biol Chem 1991; 266: 44717.
  • 12
    Argraves WS,Suzuki S,Arai H,Thompson K,Pierschbacher MD,Ruoslahti E. Amino acid sequence of the human fibronectin receptor. J Cell Biol 1987; 105: 118390.
  • 13
    Loftus JC,O'Toole TE,Plow EF,Glass A,Frelinger ALIII,Ginsberg MH. A β 3 integrin mutation abolishes ligand binding and alters divalent cation-dependent conformation. Science 1990; 249: 91518.
  • 14
    Binkley CE,Zhang L,Greenson JK,Giordano TJ,Kuick R,Misek D,Hanash S,Logsdon CD,Simeone DM. The molecular basis of pancreatic fibrosis: common stromal gene expression in chronic pancreatitis and pancreatic adenocarcinoma. Pancreas 2004; 29: 25463.
  • 15
    Linder S,Castanos-Velez E,von Rosen A,Biberfeld P. Immunohistochemical expression of extracellular matrix proteins and adhesion molecules in pancreatic carcinoma. Hepatogastroenterology 2001; 48: 13217.
  • 16
    Mollenhauer J,Roether I,Kern HF. Distribution of extracellular matrix proteins in pancreatic ductal adenocarcinoma and its influence on tumor cell proliferation in vitro. Pancreas 1987; 2: 1424.
  • 17
    Stein WD,Litman T,Fojo T,Bates SE. A Serial Analysis of Gene Expression (SAGE) database analysis of chemosensitivity: comparing solid tumors with cell lines and comparing solid tumors from different tissue origins. Cancer Res 2004; 64: 280516.
  • 18
    Armstrong T,Packham G,Murphy LB,Bateman AC,Conti JA,Fine DR,Johnson CD,Benyon RC,Iredale JP. Type I collagen promotes the malignant phenotype of pancreatic ductal adenocarcinoma. Clin Cancer Res 2004; 10: 742737.
  • 19
    Bachem MG,Schunemann M,Ramadani M,Siech M,Beger H,Buck A,Zhou S,Schmid-Kotsas A,Adler G. Pancreatic carcinoma cells induce fibrosis by stimulating proliferation and matrix synthesis of stellate cells. Gastroenterology 2005; 128: 90721.
  • 20
    Grzesiak JJ,Bouvet M. The α2β1 integrin mediates the malignant phenotype on type I collagen in pancreatic cancer cell lines. Br J Cancer 2006; 94: 131119.
  • 21
    Grzesiak JJ,Smith KC,Chalberg C,Burton DW,Deftos LJ,Bouvet M. Type I collagen and divalent cation shifts disrupt cell-cell adhesion, increase, migration, and decrease PTHrP, IL-6, and. IL-8 expression in pancreatic cancer cells. Int J Gastrointest Cancer 2005; 36: 13146.
  • 22
    Imamichi Y,Konig A,Gress T,Menke A. Collagen type I-induced Smad-interacting protein 1 expression downregulates E-cadherin in pancreatic cancer. Oncogene 2007; 26: 23815.
  • 23
    Koenig A,Mueller C,Hasel C,Adler G,Menke A. Collagen type I induces disruption of E-cadherin-mediated cell-cell contacts and promotes proliferation of pancreatic carcinoma cells. Cancer Res 2006; 66: 466271.
  • 24
    Shintani Y,Hollingsworth MA,Wheelock MJ,Johnson KR. Collagen I promotes metastasis in pancreatic cancer by activating c-Jun NH(2)-terminal kinase 1 and up-regulating N-cadherin expression. Cancer Res 2006; 66: 1174553.
  • 25
    Grzesiak JJ,Ho JC,Moossa ARBouvet M. The integrin-extracellular matrix axis in pancreatic cancer. Pancreas 2007; 35: 293301.
  • 26
    Grzesiak JJBouvet M. Determination of the ligand binding specificities of the α2β1 and α1β1 integrins in a novel three-dimensional in vitro model of pancreatic cancer. Pancreas 2007; 34: 2208.
  • 27
    Grzesiak JJ,Davis GE,Kirchhofer D,Pierschbacher MD. Regulation of α 2 β 1-mediated fibroblast migration on type I collagen by shifts in the concentrations of extracellular Mg2+ and Ca2+. J Cell Biol 1992; 117: 110917.
  • 28
    Grzesiak JJ,Pierschbacher MD. Changes in the concentrations of extracellular Mg++ and Ca++ down-regulate E-cadherin and up-regulate α 2 β 1 integrin function, activating keratinocyte migration on type I collagen. J Invest Dermatol 1995; 104: 76874.
  • 29
    Grzesiak JJ,Pierschbacher MD. Shifts in the concentrations of magnesium and calcium in early porcine and rat wound fluids activate the cell migratory response. J Clin Invest 1995; 95: 22733.
  • 30
    Bardeesy N,DePinho RA. Pancreatic cancer biology and genetics. Nat Rev Cancer 2002; 2: 897909.
  • 31
    Menke A,Adler G. TGFβ-induced fibrogenesis of the pancreas. Int J Gastrointest Cancer 2002; 31: 416.
  • 32
    Ishihara N,Yoshida A,Koizumi M. Metal concentrations in human pancreatic juice. Arch Environ Health 1987; 42: 35660.
  • 33
    Shimoyama S,Gansauge F,Gansauge S,Oohara T,Beger HG. Altered expression of extracellular matrix molecules and their receptors in chronic pancreatitis and pancreatic adenocarcinoma in comparison with normal pancreas. Int J Pancreatol 1995; 18: 22734.
  • 34
    Cunningham JJ,Anbar RD,Crawford JD. Hypomagnesemia: a multifactorial complication of treatment of patients with severe burn trauma. JPEN J Parenter Enteral Nutr 1987; 11: 3647.
  • 35
    Durlach J. Magnesium and its relationship to oncology. In: SigelH,SigelA, eds.Metal ions in biological systems, vol. 26. New York: Marcel Dekker, 1990. 54978.
  • 36
    Durlach J,Bara M,Guiet-Bara A,Collery P. Relationship between magnesium, cancer and carcinogenic or anticancer metals. Anticancer Res 1986; 6: 135361.
  • 37
    Grzesiak JJ,Clopton P,Chalberg C,Smith K,Burton DW,Silletti S,Moossa AR,Deftos LJ,Bouvet M. The extracellular matrix differentially regulates the expression of PTHrP and the PTH/PTHrP receptor in FG pancreatic cancer cells. Pancreas 2004; 29: 8592.
  • 38
    Ruoslahti E,Hayman EG,Pierschbacher M,Engvall E. Fibronectin: purification, immunochemical properties, and biological activities. Methods Enzymol 1982; 82(Part A): 80331.
  • 39
    Hoffstrom BG,Wayner EA. Immunohistochemical techniques to study the extracellular matrix and its receptors. Methods Enzymol 1994; 245: 31646.
  • 40
    Wayner EA,Carter WG,Piotrowicz RS,Kunicki TJ. The function of multiple extracellular matrix receptors in mediating cell adhesion to extracellular matrix: preparation of monoclonal antibodies to the fibronectin receptor that specifically inhibit cell adhesion to fibronectin and react with platelet glycoproteins Ic-IIa. J Cell Biol 1988; 107: 188191.
  • 41
    Grzesiak JJ,Smith KC,Burton DW,Deftos LJ,Bouvet M. GSK3 and PKB/Akt are associated with integrin-mediated regulation of PTHrP, IL-6 and IL-8 expression in FG pancreatic cancer cells. Int J Cancer 2005; 114: 52230.
  • 42
    Grzesiak JJ,Smith KC,Chalberg C,Truong C,Burton DW,Deftos LJ,Bouvet M. Heat shock protein-70 expressed on the surface of cancer cells binds parathyroid hormone-related protein in vitro. Endocrinology 2005; 146: 356776.
  • 43
    White DJ,Puranen S,Johnson MS,Heino J. The collagen receptor subfamily of the integrins. Int J Biochem Cell Biol 2004; 36: 140510.
  • 44
    Popova SN,Lundgren-Akerlund E,Wiig H,Gullberg D. Physiology and pathology of collagen receptors. Acta Physiol (Oxf) 2007; 190: 17987.
  • 45
    Li X,Ma Z,Zhu J,Li K,Wang S,Bie P,Dong J. Laminin-5 promotes cell motility by regulating the function of the integrin α6β1 in pancreatic cancer. Carcinogenesis 2004; 25: 2283.
  • 46
    Sawai H,Yamamoto M,Okada Y,Sato M,Akamo Y,Takeyama H,Manabe T. Alteration of integrins by interleukin-1α in human pancreatic cancer cells. Pancreas 2001; 23: 399405.
  • 47
    Takeichi M. Cadherin cell adhesion receptors as a morphogenetic regulator. Science 1991; 251: 14515.
  • 48
    Iacobuzio-Donahue CA,Ashfaq R,Maitra A,Adsay NV,Shen-Ong GL,Berg K,Hollingsworth MA,Cameron JL,Yeo CJ,Kern SE,Goggins M,Hruban RH. Highly expressed genes in pancreatic ductal adenocarcinomas: a comprehensive characterization and comparison of the transcription profiles obtained from three major technologies. Cancer Res 2003; 63: 861422.
  • 49
    Iacobuzio-Donahue CA,Maitra A,Olsen M,Lowe AW,van Heek NT,Rosty C,Walter K,Sato N,Parker A,Ashfaq R,Jaffee E,Ryu B, et al. Exploration of global gene expression patterns in pancreatic adenocarcinoma using cDNA microarrays. Am J Pathol 2003; 162: 115162.
  • 50
    Mirtti T,Nylund C,Lehtonen J,Hiekkanen H,Nissinen L,Kallajoki M,Alanen K,Gullberg D,Heino J. Regulation of prostate cell collagen receptors by malignant transformation. Int J Cancer 2006; 118: 88998.
  • 51
    Thavmilselvan V,Fomby M,Walsh M,Basson MD. Divalent cations modulate human colon cancer cell adhesion. J Surg Res 2003; 110: 25565.
  • 52
    van der Voort van Zyp J,Conway WC,Thamilselvan V,Polin L,Basson MD. Divalent cations influence colon cancer cell adhesion in a murine transplantable tumor model. Am J Surg 2005; 190: 7017.
  • 53
    Nakao A,Fujii T,Sugimoto H,Kanazumi N,Nomoto S,Kodera Y,Inoue S,Takeda S. Oncological problems in pancreatic cancer surgery. World J Gastroenterol 2006; 12: 446672.
  • 54
    van Grevenstein WM,Hofland LJ,Jeekel J,van Eijck CH. The expression of adhesion molecules and the influence of inflammatory cytokines on the adhesion of human pancreatic carcinoma cells to mesothelial monolayers. Pancreas 2006; 32: 396402.
  • 55
    Bouvet M,Spernyak J,Katz MH,Mazurchuk RV,Takimoto S,Bernacki R,Rustum YM,Moossa AR,Hoffman RM. High correlation of whole-body red fluorescent protein imaging and magnetic resonance imaging on an orthotopic model of pancreatic cancer. Cancer Res 2005; 65: 982933.
  • 56
    Bouvet M,Tsuji K,Yang M,Jiang P,Moossa AR,Hoffman RM. In vivo color-coded imaging of the interaction of colon cancer cells and splenocytes in the formation of liver metastases. Cancer Res 2006; 66: 112937.
  • 57
    Bouvet M,Wang J,Nardin SR,Nassirpour R,Yang M,Baranov E,Jiang P,Moossa AR,Hoffman RM. Real-time optical imaging of primary tumor growth and multiple metastatic events in a pancreatic cancer orthotopic model. Cancer Res 2002; 62: 153440.
  • 58
    Katz MH,Bouvet M,Takimoto S,Spivack D,Moossa AR,Hoffman RM. Survival efficacy of adjuvant cytosine-analogue CS-682 in a fluorescent orthotopic model of human pancreatic cancer. Cancer Res 2004; 64: 182833.
  • 59
    Katz MH,Takimoto S,Spivack D,Moossa AR,Hoffman RM,Bouvet M. A novel red fluorescent protein orthotopic pancreatic cancer model for the preclinical evaluation of chemotherapeutics. J Surg Res 2003; 113: 15160.
  • 60
    Katz MH,Takimoto S,Spivack D,Moossa AR,Hoffman RM,Bouvet M. An imageable highly metastatic orthotopic red fluorescent protein model of pancreatic cancer. Clin Exp Metastasis 2004; 21: 712.