Background
Prostate cancer is the second leading cause of cancer death among men in the United States, and there is clinical evidence of bone metastases in approximately 80 % of those who have died [
1,
2]. A comprehensive understanding of signaling interactions between invading epithelial-derived prostate cancer cells and the host bone stromal environment that promote bone metastasis is crucial to the development of effective therapeutic strategies.
Although markers of bone production and resorption may be present in patients, prostate carcinoma bone metastases are generally characterized by new bone formation initiated by the differentiation of mesenchymal progenitor cells into osteoblasts [
3‐
5]. We have previously demonstrated that human prostate cancer cells, which express high levels of Sonic hedgehog (Shh), activate the signaling pathway in MC3T3 pre-osteoblasts and induce osteoblast differentiation [
6].
Shh is a secreted glycopeptide that plays critical functions in the normal development of many organs including the prostate; and, deregulation of the Shh pathway has been linked to human cancer [
7‐
9]. Expression of Shh and other members of the signaling pathway have been reported in human primary prostate carcinomas and metastases, including bone [
10‐
13]. Paracrine induction of osteoblast differentiation via the Shh pathway may be a mechanism through which Shh-expressing prostate cancer cells initiate changes in the bone microenvironment that favor the development of metastases.
A hallmark of osteoblast differentiation both in vivo and in vitro is the formation of an extracellular matrix [
14,
15]. Type 1 collagen accounts for about 95 % of the organic matrix proteins in bone [
4]. The role of matrix collagen in the formation of bone metastasis is not well understood.
In the present study, we investigated the formation of a collagenous ECM by osteoblasts induced to differentiate by Shh-expressing prostate cancer cells, and the effects of a collagenous matrix on paracrine signaling between prostate cancer cells and osteoblasts.
Discussion
We provide novel evidence that AA potentiates Shh signaling between prostate cancer cells and osteoblasts, and synergistically enhances Shh-induced osteoblast differentiation.
The ability of AA to upregulate Shh signaling in osteoblasts requires Gli transcriptional activity. The expression of target genes
Gli1 and
Ptc1 in MC3T3 pre-osteoblasts cultured with LNShh cells was increased by exogenous AA to at least twice their levels in the absence of AA. However, the effect of AA was blocked in MC3T3 pre-osteoblasts that expressed a dominant negative form of GLI1, the M-TAD cells. The GLI1 translated product in M-TAD cells is expected to bind to the consensus DNA GLI binding site but not activate the pathway [
6]. AA does not directly impact Gli transcriptional activity since treatment with AA did not increase the expression of
Ptc1 ( and
Gli1; not shown) in MC3T3 pre-osteoblasts cultured alone; whereas, exogenous Shh peptide did. And, the combined treatment of AA and Shh peptide did not significantly upregulate gene expression above that attained with exposure to the Shh peptide alone. In agreement with our earlier observations, we indicate an indirect mechanism of AA action on the Shh pathway [
6].
The effect of AA on Shh signaling requires collagen production. The specific collagen synthesis inhibitor DHP blocked the AA-promoted upregulation of
Gli1 in MC3T3 pre-osteoblasts cultured with LNShh cells. DHP is a proline analog which incorporates into nascent pro-α collagen chains and prevents prolyl hydroxylation, a critical step in procollagen triple helix formation and secretion [
24]. The AA-stimulated incorporation of
3 H]proline into α
1(I) and α
2(I)-procollagen molecules in MC3T3 cells was blocked by DHP at 0.5 mM [
16], the same concentration that inhibited AA-mediated upregulation of the Shh pathway in MC3T3 cells in the present studies.
The mechanism for collagen-mediated augmentation of Shh signaling is not clear. Collagen actions are mediated via the integrin signaling pathway. Integrins are cell-surface heterodimers consisting of α and β subunits. Type 1 collagen binds most commonly to cell surface integrins: α
1β
1 and α
2β
1[
25]. The interaction between type 1 collagen and osteoblast integrins activate downstream cascades including the mitogen-activated protein kinase and phosphatidylinositol 3-kinase pathways, which lead to increased expression of target genes [
26,
27]. Recently, these pathways have been implicated in non-canonical activation of Shh signaling [
28,
29]. It is intriguing to suppose that intracellular cross-talk between the downstream mediator(s) of collagen-integrin signaling and the Shh pathway may have contributed to increased transcriptional activation of
Ptc1 and
Gli1 message in differentiating osteoblasts.
However, a role for collagen in the upregulation of the Shh pathway in osteoblasts must involve, at least in part, a Shh ligand-activated mechanism i.e., canonical hedgehog signaling. The upregulatory effect of AA on the Shh pathway was observed in MC3T3 pre-osteoblasts cultured with LNShh cells but not with LNCaP cells although matrix collagen was present in both mixed cultures. The collagen matrix may sequester Shh ligands and present them more effectively to Ptc receptors on target cells. Investigators have shown that the activity of ligands stored in the ECM such as transforming growth factor-β1 (TGF-β1) and bone morphogenetic proteins (BMPs) is influenced by matrix proteins including collagen [
30‐
32]. Additionally, there is evidence that Shh molecules bind to ECM components vitronectin and laminin [
33,
34]. Although binding of Shh to collagen has yet to be demonstrated, the orientation and/or proximity of Shh molecules to Ptc receptors on osteoblasts might be optimized when the ligands are complexed with collagen. Collagen might also enhance the expression and/or function of membrane-bound components of the pathway including the Ptc receptor and Smoothened (Smo) leading to increased activation of Shh signaling. The cell surface expression of ligand-binding TGF-β1 receptors in MC3T3 cells was downregulated by AA treatment; and, this effect was linked to AA’s action on the synthesis of collagen and its interaction with α
2β
1 integrins [
30]. Further investigations are needed to unravel the mechanisms through which collagen activates the Shh pathway.
We have previously shown that paracrine activation of the Shh pathway induces osteoblast differentiation in the absence of exogenous AA [
6]. We have confirmed these findings and further showed that AA-enhanced osteoblast differentiation requires Gli transcriptional activation since M-TAD cells failed to respond to the osteogenic effect of LNShh cells. More importantly, we have demonstrated a synergistic effect of AA on Shh-induced osteoblast differentiation. Levels of
Akp2 expression in MC3T3 pre-osteoblasts exposed to both Shh signaling and AA (i.e., cultured with LNShh cells and treated with AA) were markedly greater than those in MC3T3 cells exposed to Shh alone or AA alone. The effect of AA is linked to the presence of matrix collagen since the collagen synthesis inhibitor DHP blocked the AA-promoted upregulation of ALP activity in MC3T3 osteoblasts cultured with LNShh cells. These findings are consistent with previous reports of collagen-dependent synergistic interactions between AA and osteogenic stimuli including BMP2 and interleukin-11 [
31,
35,
36].
The biological significance of collagen fibril structural properties in cancer development and progression is just beginning to be recognized. Distinct signatures of collagen fibril organization were observed in the stromal environment of breast tumor tissues, and collagen fiber alignment appeared to correlate with tumor cell invasiveness [
37]. Cancer cells, particularly metastatic cells, move linearly along collagen fibers [
38]. Conceivably, the parallel alignment of collagen fibrils in mixed cultures of MC3T3 pre-osteoblasts and LNShh cells could facilitate prostate cancer cell motility and migration. The higher percentage of collagen fibrils with smaller diameter sizes in mixed cultures of MC3T3 and LNShh cells might indicate a more dynamic process of fibril assembly and matrix organization where early stage small-diameter fibril intermediates are continuously formed [
39,
40]. Further studies to determine the role of Shh-expressing prostate cancer cells in regulating the structural properties of bone matrix collagen, particularly under in vivo conditions, will increase our understanding of the significance of the Shh pathway in shaping the bone stromal microenvironment to support metastasis.
The role of matrix collagen on cell functions has been investigated mostly through use of scaffolds formed by either commercially-available, laboratory-prepared native type 1 collagen extracted from rat tail tendons, or collagen formed by osteoblasts. In these studies, cells are either added onto the gellified matrix or mixed with the matrix as it solidifies. Thus, matrix density and collagen fibril structural characteristics are largely pre-determined for the inoculated cells. Our mixed cell culture system enables both cancer cells and osteoblasts to interact during the process of AA-dependent collagen matrix formation. This allows investigations into reciprocal in situ interactions among cancer cells, osteoblasts, and osteoblast-synthesized matrix proteins that should provide significant insights into signaling processes relevant to bone metastasis.
Methods
Cells and plasmid transfections
Parental LNCaP human prostate cancer cells and mouse calvaria-derived non-transformed pre-osteoblast cells MC3T3-E1 (subclone 4; designated as MC3T3 cells) were commercially obtained (ATCC, Rockville, MD).
LNCaP cells have been previously stably transfected with a 1.44 kb human
Shh cDNA cloned into a pIRES2-EGFP mammalian cell expression vector (designated LNShh cells) or with pIRES2-EGFP vector alone as controls (designated LNCaP cells) [
10]. We have previously confirmed the increased expression of Shh at the gene and protein levels in LNShh cells compared to LNCaP cells [
10]. The morphology of LNCaP and LNShh cells appears identical and these cells exhibit similar growth properties in culture [
10,
41]. LNCaP and LNShh cells were maintained at 37 C, 5 % CO
2 in complete culture medium consisting of RPMI-1640 supplemented with 10 % fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin (Gibco Invitrogen). Shh gene and protein expression were routinely determined by quantitative real time RT-PCR and western blot analysis, respectively, and GFP expression was monitored by fluorescence microscopy.
MC3T3 cells have been previously stably transfected with pCMV-
GLI(−)TAD: a human
GLI1 cDNA lacking a
t rans
a ctivation
d omain and cloned into pcDNA3 plasmid (designated M-TAD cells) [
6]. We have previously shown that both M-TAD cells and parental MC3T3 cells (used as controls) express endogenous mouse
Gli1 message; but, only the M-TAD cells express the message for the human
GLI1(−TAD) transgene whose GLI1 translated product is expected to bind to the consensus DNA GLI binding site but not activate the pathway; thus, acting as a dominant negative transcription factor [
6]. MC3T3 and M-TAD cells were maintained at 37 C, 5 % CO
2 in non-differentiation complete culture medium consisting of ascorbic acid (AA)-free α-MEM supplemented with 10 % FBS, 100 U/ml penicillin and 100 μg/ml streptomycin (Gibco Invitrogen).
Mixed culture of cells
LNCaP or LNShh cells (5 × 104) and MC3T3 or M-TAD cells (0.5 × 104) were mixed in AA-free α-MEM complete culture medium and seeded per well of 6-well tissue culture plates. When grown in chamber slides, pre-osteoblasts and prostate cancer cells were mixed at equal concentrations of 1 × 104 cells per cell line. Cultures were maintained for the length of time specified in the experiments with media changes every 2–3 days.
Effect of ascorbic acid
L-ascorbic acid (AA; Aldrich) was dissolved in AA-free α-MEM complete culture medium at 50 μg/ml final concentration. Mixed cultures were maintained in complete culture medium with AA or in complete culture medium only as controls. The AA concentration used in these experiments is below pharmacologic concentrations that may be cytotoxic to prostate cancer cells including LNCaP [
42,
43]. To determine the direct effect of AA, MC3T3 cells were seeded alone onto 6-well culture plates at 1 × 10
5 cells per well and treated with AA as above.
Effect of Shh peptide
Shh-N, a modified active N-terminal peptide of human Shh (kindly provided by Curis Inc., Cambridge, MA), was prepared in serum-free AA-free α-MEM culture medium at 1 μg/ml final concentration. MC3T3 cells were seeded onto 6-well tissue culture plates at 1 × 105 cells per well in AA-free α-MEM complete culture medium. Following overnight incubation, cells were maintained for 24 h in serum-free culture medium with Shh-N or in serum-free culture medium only as controls.
Effect of collagen synthesis inhibitor
The collagen synthesis inhibitor 3,4-dehydro-L-proline (DHP; Sigma) was dissolved in AA-free α-MEM complete culture medium. Mixed cultures were maintained in complete culture medium with 50 μg/ml AA and varying concentrations of DHP or in complete culture medium only (i.e., without both AA and DHP) as controls. In some experiments, single cultures of MC3T3 cells were used.
RNA isolation and real time quantitative RT-PCR
Total RNA was extracted using Trizol (Invitrogen), purified using the RNeasy Mini Kit (Qiagen) and subjected to DNase treatment with RQ1 RNase-free DNase (Promega) to remove contaminating genomic DNA. The TaqMan® Gold PCR Core Reagent Kit along with MuLV Reverse Transcriptase and RNase Inhibitor (Applied Biosystems) were used for cDNA synthesis. PCR primers (Invitrogen) and FAM-QSY7 probes (MegaBases, Inc.) for genes of interest and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase were designed using the Primer Express 3.0 software program. Mouse species-specific primer sequences, which amplified genes of interest in mouse MC3T3 but not in human prostate cancer cells, have been published [
6]. mRNA expression was measured in duplicate or triplicate per sample using 40 cycles of amplification in the 7500 Fast Real-Time PCR System (Applied Biosystems). Reactions were routinely performed without Reverse Transcriptase to demonstrate RNA dependence of the reaction products. Results were analyzed using the comparative C
t method as described previously [
6]. Data are expressed as relative fold change in gene expression.
Cell proliferation assay
MC3T3 cells were seeded onto 24-well tissue culture plates at 0.2 × 104 cells per well and maintained in AA-free α-MEM complete culture medium. Proliferation was determined using the Cell Counting Kit-8 (Dojindo Laboratories, Japan) which is based on the formation of a water-soluble formazan dye through the activity of dehydrogenases in living cells. Absorbance measurements at 450 nm are in direct proportion to the number of living cells.
Immunocytochemistry
Cells were maintained as mixed cultures in Lab-TekII CC2-treated chamber slides (Nunc) for the length of time specified in the experiments with media changes every 2–3 days. Cells were fixed in 10 % neutral buffered formalin for 10 minutes and processed for von Gieson staining for collagen. For positive control, a section of aorta was similarly processed and stained. Slides were viewed in a Leica DMR-HC Upright Microscope and images were captured with imaging software (Improvision Openlab).
Alkaline phosphatase activity
Quantitative determination of ALP activity was done using the p-Nitrophenyl Phosphate (pNPP) Liquid Substrate System (Sigma Aldrich) as previously described [
6]. Absorbance at 405 nm was measured using a microplate reader, and ALP activity was calculated according to manufacturer’s instructions. Protein determination was done using the Bio-Rad
DC Protein Microplate Assay according to manufacturer’s protocol.
Staining for ALP activity was performed on mixed cultures which were fixed with 10 % neutral buffered formalin for 10 minutes and incubated with alkaline phosphatase substrate solution (Sigma-Aldrich) for at least 30 minutes at room temperature in the dark as previously described [
6].
Transmission electron microscopy
Mixed cultures of MC3T3 pre-osteoblasts and LNCaP or LNShh cells were maintained in the presence of AA (50 μg/ml) on Thermanox cover slips (13 mm diameter; Electron Microscopy Services) kept in 24-well tissue culture plates for 7, 14, and 21 days. At end of culture, samples were fixed overnight at 4 C in 0.1 M sodium cacodylate buffer pH7.3 containing 2 % paraformaldehyde and 2.5 % glutaraldhyde, post-fixed with 2 % osmium tetroxide in 0.1 M sodium cacodylate buffer, and rinsed with distilled water. Following, samples were stained en bloc with 3 % uranyl acetate, rinsed in distilled water, dehydrated in ascending grades of ethanol, embedded in resin mixture of Embed 812 and Araldite, and cured in oven at 60 C. Samples were sectioned on a Leica Ultracut UC6 ultramicrotome, and 70 nm thin sections were collected on 200 mesh copper grids and post-stained with 3 % uranyl acetate and Reynolds lead citrate. Samples were sectioned either parallel to cell layers to reveal fibril orientation or perpendicular to cell layers to allow morphometric measurement of fibril diameters [
23,
39]. Samples were examined on FEI Tecnai Spirit G2 TEM, and digital images were captured on an FEI Eagle camera at magnifications ranging from 1900x to 49000x. Two to three grids per sample were examined and images from several sections per grid were taken.
Analysis of collagen fibril diameter
Diameters of collagen fibrils were measured, using the analysis measurement tool of the Adobe Photoshop CS3 Extended software, on TEM images of samples sectioned perpendicular to cell layers and examined at magnification of 49000×. Fibril diameters were measured from 500 randomly selected collagen fibrils from 8–14 TEM images from different grid sections from each of 2 samples per group.
Data analysis
Data were analyzed by ANOVA and pair-wise multiple comparisons were done using the Bonferroni
t-test at
P < 0.05. Comparison between two groups was done by Student’s
t-test. Fibril diameter size distributions were compared using the asymptotic Kolmogorov-Smirnov two-sample test [
21].
Acknowledgements
We thank the following: Ying Zhou, Ph.D., Biostatistics Research Core, Children’s Memorial Research Center, for help with statistical analysis; the Histology Facility at Children’s Memorial Hospital, for help with collagen staining; and, Curis Inc., Cambridge, MA, for generously providing the Shh-N peptide. This work was supported by grants from the National Cancer Institute, Career Development Award (MLGL); American Cancer Society Illinois Division (MLGL); Center to Reduce Cancer Health Disparities, Continuing Umbrella of Research Experience Program Award (MV); Illinois Regenerative Medicine Institute (MLGL and PMI); George M. Eisenberg Foundation for Charities (PMI); and, Eisenberg Research Scholarship (MLGL). TEM work was performed at the Northwestern University Cell Imaging Facility generously supported by NCI CCSG P30 CA060553 awarded to the Robert H Lurie Comprehensive Cancer Center.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SMZ performed experiments and helped with data analysis. MV and TD performed experiments and contributed to gene expression data analysis. DW and PI contributed to data analysis and manuscript editing. MLGL designed the studies, performed experiments, performed data analysis, and prepared the manuscript. All authors read and approved the final manuscript.