| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
Departments of Genitourinary Medical Oncology [K. F., J. Y., C. R. S., E. L. K., D. D., M. O., C. J. L., N. M. N.], Endocrine Neoplasia and Hormonal Disorders [S. P.], and Pathology [K. A. R.], The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030; Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas 77030 [G. K.]; and Abbott Laboratories, Abbott Park, Illinois [T. J. J.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Design and Results: We show histological evidence of de novo bone formation as early as 2 weeks after injection of MDA PCa 2b cells in the bone of severe combined immunodeficient disease mice. In vitro, we show that PMOs induce MDA PCa 2b proliferation, suggesting a synergistic paracrine loop between these cells and PMOs. Endothelin (ET)-1, which is a mitogen for several cell types, is produced by all prostate cancer cell lines tested, and Atrasentan, an antagonist of ET-1 receptor A, partially reversed PMO proliferation induced by MDA PCa 2b cells. ET-1 is known to be comitogenic with a number of growth factors, including insulin-like growth factor (IGF)-I. In this study, we report that IGF-binding protein (IGFBP)-3 transcripts (that regulate levels of free IGF) are down-regulated in prostate cancer cells cocultured with PMO, whereas prostate-specific antigen (a protease known to cleave IGFBP-3) is detected in the 150–400 ng/ml range. Accordingly, IGFBP-3 has antiproliferative effects in PMOs, which were attenuated in our in vitro system. Taken together, our studies also implicate the IGF axis to play a role in this model of bone metastases. Secondly, the transcript level of mouse double minute 2 (a protein that regulate p53) was increased in prostate cancer cells grown with PMOs. The p53-dependent and -independent oncogenic activities of mouse double minute 2 suggest that osteoblasts induce a survival advantage in prostate cancer cells. Lastly, we show that expression of osteoprotegerin is decreased and of receptor activator of nuclear factor-
B ligand is increased in PMOs cultured in the presence of MDA PCa 2b cells, two events associated with osteoclast activation and bone resorption.
Conclusions: Our results provide evidence that multiple and distinct molecular events affecting both bone formation and bone resorption concur to the increase bone mass in prostate cancer bone metastases. These data also provide a rationale for developing therapeutic strategies designed to target these molecular changes.
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ET-1, a potent vasoconstrictor peptide has also been implicated in the pathophysiology of osteoblastic bone metastases from prostate cancer (8, 9, 10) : ET-1 is produced by prostate cancer cells; expression of the ET-1 clearance receptor B is lost during prostate cancer progression; and ET-1 levels are higher in plasma of prostate cancer patients with bone metastasis (9, 10, 11) . In vitro studies have demonstrated that ET-1 increases osteoblast proliferation and osteoblast-specific gene expression (10, 11, 12) . Furthermore, overexpression of ET-1 in murine bone was shown to induce an increase of bone formation that could be blocked by administration of an antagonist of ETA (11) . However, to date, no study has formally demonstrated a direct link between the ET-1 production or inhibition in prostate cancer cells and the occurrence of osteoblastic lesions in bone.
Although ET-1 is mitogenic for a variety of cell types, it has been shown that its effect is more potent as comitogenic of several growth factors, including IGFs (11) . IGFs are abundant in human bones, and they have mitogenic, chemotactic, and antiapoptotic effects on a wide variety of cells, including both prostate cancer cells (13 , 14) and osteoblasts (15) . IGF-I exerts its action through the IGF-IR, a tyrosine kinase receptor. Levels of IGFs available to interact with the receptor are modulated by six IGFBPs; these IGFBPs are, in turn, hydrolyzed by a number of proteases (16) . More than 90% of the circulating IGFs are bound to IGFBP-3. IGFBP-3 also have IGF-independent growth inhibitory effects on several cell types (16) . Prostate cancer cells have been shown to actively regulate bioavailable IGFs levels by urokinase- and PSA-mediated IGFBP-3 proteolysis (4 , 7 , 8 , 16) . IGFBP-3 inhibits the osteoblastic growth effect of IGFs in osteoblasts (15) , whereas its expression in prostate cancer cells inversely correlates with progression from benign to malignant disease (16) .
As it was discussed, bone metastases of human prostate cancer possess osteoblastic features, however, an underlying osteolytic component is present in most cases (8 , 17 , 18) . In support of this concept, preliminary evidence of a Phase III trial suggested that oral clodronate (an inhibitor of bone resorption) delays symptomatic progression of bone metastases and may improve survival (19) . Major molecules implicated in cancer-induced osteolysis include PTHrP, RANKL, OPG, and MMPs. The PTHrP protein stimulates osteoclastic activity (20) . RANKL, a tumor necrosis factor, induces the differentiation and activation of osteoclasts. OPG, a soluble tumor necrosis factor receptor, inhibits the formation and activity of osteoclasts by sequestering RANKL (21, 22, 23) . Treatment with OPG abolishes cancer-induced bone destruction in vivo (24) . MMPs are a group of neutral proteinases thought to be involved in bone formation and remodeling. Osteoblasts secrete several proteases including MMPs such as collagenases, thus remodeling their own ECM (25) .
In an effort to understand the molecular basis of prostate cancer bone metastases, we established two human prostate cancer cell lines, MDA PCa 2a and MDA PCa 2b, that express PSA and the androgen receptor, and their growth is androgen regulated (26 , 27) . We then established an in vitro model of bone metastases, consisting of cocultured MDA PCa 2a and MDA PCa 2b cells with PMOs and showed that these prostate cancer cells induced a specific increase in osteoblast growth and differentiation (28) . In the current study, we used this model to define the contribution of various genes and cellular pathways in the pathogenesis of prostate cancer bone metastases.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Intrabone Injections.
Male SCID mice were obtained from Charles River Laboratories, Inc. (Wilmington, MA), and housed in a facility with constant humidity and temperature and a 12-h light-dark cycle. They had ad libitum access to standard mouse feed and water. All animal experimentation was conducted in accordance with accepted standards of humane animal care and was approved by the Animal Care and Use Committee of M. D. Anderson Cancer Center. Animals were anesthetized with intramuscular injections of ketamine (100 mg/kg) plus acepromazine (2.5 mg/kg). Aliquots of 0.5 x 106 of MDA PCa 2b cells were diluted in 5 µl of growth medium and then injected into the distal epiphysis of the right femur of each mouse using a 28-gauge Hamilton needle. The contralateral femur was used as an internal control, and four control mice were injected with saline only. Six additional mice served as a second control group; they were injected with 0.5 x 106 of PC3 cells. Mice were monitored twice weekly for tumor bulk. Serum PSA levels were measured, and X-rays of the legs were obtained at 2, 3, 4, 6, 10, and 16 weeks after injection. Blood was obtained from a small incision in the main tail vein, and the serum was separated for PSA testing. A set of five mice was killed at each point (or earlier in the case of a bulky tumor), and pathologic examination of the subject bones was performed.
Tissue Samples and PSA Assay.
Formalin-fixed, paraffin-embedded tissue samples from the tumors were prepared by the Department of Veterinary Medicine at M. D. Anderson Cancer Center. The subject bones were dissected free of muscle, fixed in 10% buffered formalin, decalcified in 5% formic acid, and then embedded in paraffin. Longitudinal 3-µm thick sections were obtained from each sample and stained with H&E. PSA concentration in the culture medium and PSA serum of mice were measured with a microparticle enzyme immunoassay (IMx PSA assay; Abbott Laboratories, Abbott Park, IL).
Primary Cultures of Mouse Calvaria Osteoblasts and Coculture with Prostate Cancer Cells.
Primary cultures of osteoblasts were obtained from CD1 mice as reported previously (28)
. PMOs were then plated in 
-MEM plus 10% FBS for 48 h. Trypsin was subsequently added, and the cells were replated in culture dishes for experimentation. An in vitro bicompartmental culture system was used (28)
. Briefly, PMOs were seeded in tissue culture plates, whereas prostate cancer cells were seeded in cell culture inserts (0.4-µm pore size; Falcon/Becton Dickinson Labware, Franklin Lakes, NJ) so that the two different cell types shared the culture medium but were not in physical contact.
Mitogenic Assays.
[3H]Thymidine (NEN Life Science Products, Boston, MA) incorporation into DNA was evaluated as described previously (29)
.
Reagents.
ET-1 was purchased from Sigma Chemical (St. Louis, MO) and IGFBP-3 from R&D Systems, Inc. Atrasentan, was provided by Abbott Laboratories.
Protein Analysis.
Radioimmunoanalysis was used to assess ET-1 production by prostate cancer cells (Biotrak; Amersham) and IGF-I concentration in the culture medium (Nichols Institute Diagnostics, San Juan Capistrano, CA). Western blot analysis was performed as described previously (26)
.
Bone Differentiation Studies.
Alkaline phosphatase activity in PMOs was determined as described previously (28)
. Mineralization was monitored using a Von Kossa’s staining assay as described previously (28)
.
RNA Extraction and Northern Blot Analysis.
Northern blot analysis was performed as reported previously (28)
. Human IGF-I probe containing the entire coding region was a generous gift from Dr. Rodrigo Bravo (Bristol Myers Squib, Princeton, NJ). Human PTHrP probe was prepared from a 306-bp coding region cDNA fragment of a human thyroid carcinoma. The forward primer was 5'-CTGGTTCAGCAGTGGAGCGTC-3' starting at exon 3 of the human gene, and the reverse primer was 5'-CCTGAGTTAGGTATCTGCC-3' from exon 4 of the human gene. IGFBP-3 probe (clone ID no. 898218) was a generous gift from Millennium Pharmaceuticals, Inc. (Cambridge, MA), and the mdm2 probe was kindly provided by Dr. B. Vogelstein (Johns Hopkins University, Baltimore, MD). OPG probe (5130-bp coding region cDNA) was kindly provided by Dr. Christopher Wood (M. D. Anderson Cancer Center). Human and mouse RANKL probes were kindly provided by Amgen, Inc. (Thousand Oaks, CA). cDNA probe for 36b4 was used as a loading control (30)
.
Gene Array Analysis.
Differential expression of multiple genes by MDA PCa 2b cells growing alone or in coculture with PMOs was analyzed using the Human PathwayFinder-1 GEArray Kit (SuperArray, Inc., Bethesda, MD). The studied genes include egr-1, c-fos, c-myc, activating transcription factor-2, p53, hsf1, heat shock protein 27, heat shock protein 90, iNos, nuclear factor-B, inhibitor of nuclear factor-B , interleukin 2, Fas (Apo-1), CD5, p16, p21, p57Kip2, gadd45, pig7, pig8, mdm2, bax, and CYP19 (aromatase P450).
Statistical Analysis.
Numerical data were expressed as means ± SE. Statistical differences between means for the different groups were evaluated with Sigma Plot 4 using one-way ANOVA and the t test. Ps of <0.05 were considered statistically significant.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
summarizes tumor take of SCID mice after intrafemoral injection of MDA PCa 2b cells as assessed by serum PSA levels and X-ray imaging of the injected limb. Typical results of X-ray analysis are presented in Fig. 1
. Pathologic examinations were performed after the mice had been killed; results are summarized in Fig. 2
.
|
|
|
, Table 1
). A slight increase in serum PSA was already found at this early stage (Table 1)
. Histopathological examination of the injected limbs 2 weeks postinjection revealed colonies of tumor cells in the medullar cavity always surrounded by bone matrix, which suggested that these cells induce bone remodeling (Fig. 2, A and B)
. Four weeks after cell injection, tumor cells replaced the bone marrow in the metaphysis or the epiphysis; changes at this time included destruction of trabecular bone in some cases without destruction of cortical bone (Fig. 2C)
.
Two months later, pathologic examination revealed the presence of tumor cells located in the epiphysis, metaphysis, and diaphysis. Tumor cells in the diaphysis were often necrotic, which was likely related to insufficient vascularization in the diaphysis. In some cases, the tumor was growing through holes from the metaphysis into the surrounding soft tissues. At this point, there was evidence of new bone formation in the epiphysis (Fig. 2D)
and bone thickening in the diaphysis. Proliferation of active osteoblasts surrounding the tumor cells and in areas once occupied by tumor cells were also detectable (Fig. 2
, inset). Osteoclasts were not usually seen in these lesions.
Ten to 16 weeks after injection, serum PSA had reached levels of up to 232 ng/ml (Table 1)
. Histopathologically, bulky tumors invading the entire bone marrow cavity of the metaphysis (and usually of the diaphysis) were seen. Tumor cells were mostly necrotic in the diaphysis, although they were viable in the distal metaphysis and often in the proximal metaphysis as well (data not shown). In most cases, tumor cells were found in the soft tissue surrounding the distal metaphysis. New bone formation was obvious in the distal epiphysis, the metaphysis and the diaphysis, and active osteoblasts were frequently seen in the proximal metaphysis (Fig. 2E)
. Necropsy showed no macroscopic evidence of distant metastases, although in some cases, tumor emboli were found in blood vessels of the lung.
To determine whether this sequence of events was specific to this cell line or could be observed with any prostate cancer cell line, we used PC3 cells. PC3 cells are also derived from a bone metastasis of prostate cancer but do not produce PSA and are not regulated by androgens. The injection of the prostate cancer cell line PC3 resulted in intrabone tumors in five of six mice 4 weeks after intrafemoral injections. In contrast to the bone lesions induced by MDA PCa 2b cells, the lesions induced by PC3 cells were all osteolytic (Fig. 2F)
, and the femur of one mouse was broken (Fig. 2F
, inset). Viable PC3 cells were found in the bone medulla of the distal metaphysis and in the surrounding soft tissue. The cortical bone of the metaphysis was completely destroyed in most cases, and many osteoclasts were found surrounding the PC3 tumor cells (Fig. 2F)
. No evidence of new bone formation was found.
Increased Proliferation Rate of MDA PCa 2a and MDA PCa 2b Prostate Cancer Cells in an in Vitro Model of Bone Metastasis.
We have previously reported that bone-derived MDA PCa 2a and MDA PCa 2b prostate cancer cells induce osteoblast growth and differentiation when the cells share medium during coculturing (28)
. In this study, we tested the effect of PMOs cocultured with prostate cancer cells in the proliferation rate of prostate cancer cells using [3H]thymidine incorporation into DNA as a direct measure of DNA synthesis. We found that DNA synthesis was increased in prostate cancer cells MDA PCa 2a and MDA Ca 2b but not PC3 at 72 h of coculture with PMOs (Fig. 3A)
. These results suggest that soluble factors secreted by PMO and by bone-forming prostate cancer cells are involved in a paracrine loop that results in new bone formation and prostate cancer growth.
|
. In initial experiments, we determined that ET-1 stimulated the growth of PMOs with an equivalent effect in the range of concentration tested (10-11-10-8 M; data not shown). ET-1 (10-9 M) induced PMO growth, whereas Atrasentan (10-5 M) reversed the ET-1 induction of PMO cell growth (Fig. 3B)
. When Atrasentan was used as a single agent, there was no significant inhibitory effect on the growth of PMOs (Fig. 3B)
. Furthermore, the stimulating effect on the growth of PMOs induced by MDA PCa 2b cells in the coculture system was partially inhibited by Atrasentan (Fig. 3B)
. Similar results were obtained when using the osteoblastic 2T3 cell line instead of PMOs (data not shown). Results were consistent with the previous experiment and supported the inhibitory effect of Atrasentan on MDA PCa 2b-induced growth of osteoblasts. That the inhibitory effect was not complete suggests that other mitogenic factors are produced by prostate cancer cells. The coculture with MDA PCa 2b cells increased alkaline phosphatase activity and calcified matrix formation by PMOs (Table 3
and Fig. 4
, respectively). Although Atrasentan strongly inhibited mineralization by PMOs grown alone (Fig. 4)
, it did not inhibit mineralization or alkaline phosphatase activity induced by MDA PCa 2b cells. This suggests that MDA PCa 2b cells produce other soluble factors besides ET-1, which promote osteoblast differentiation. ET-1 can act alone as a mitogen, but its effect is more potent as a comitogen with a variety of growth factors (11)
. Among them is IGF-I, which plays a key role in bone formation, is an important regulator of the activity of mature osteoblasts (15)
, has strong mitogenic and antiapoptotic effects on various cancer cells, including prostate cancer, and has been implicated in the pathogenesis of osteoblastic bone metastases of prostate cancer (4
, 8
, 31, 32, 33, 34)
.
|
|
|
1 to 7.5 kb. Yet, these multiple transcripts encode one IGF-I protein. Northern blot analysis demonstrated that both PMOs and prostate cancer cells express IGF-I transcripts. The major transcript expressed by PMOs is 2.7 kb long, whereas the major transcript expressed by prostate cancer cells is 1.1 kb long (data not shown). We then studied the concentration of IGF-I in the cultured medium conditioned by prostate cancer cells and PMOs growing alone and in coculture. IGF-I is secreted by both prostate cancer cells and PMOs (Table 4)
. Moreover, culture medium conditioned by MDA PCa 2b and PMOs in coculture has IGF-I levels between 32 and 42 ng/ml (4.2 and 5.6 nM), which are well within the range reported to induce DNA synthesis and collagen production in rodent calvaria cells, as well as proliferation of prostate cancer cells.
|
.
|
As an additional attempt to understand the implications of the decreased IGFBP-3 expression in this model system, we added IGFBP-3 to PMOs grown alone and in coculture with MDA PCa 2b cells. We found that IGFBP-3 inhibited PMOs cell growth at all concentrations used (Fig. 5B)
. When PMOs are grown in coculture with MDA PCa 2b cells, the inhibitory effect of IGFBP-3 is abolished at doses of 10 and 50 ng/ml. An inhibitory effect on PMOs cocultured with MDA PCa 2b cells was still detected (though of less intensity) at 100 and 1000 ng/ml IGFBP-3; these findings were confirmed in an independent experiment.
Taken together, these results implicate the IGF axis in the osteoblastic reaction of prostate cancer bone metastases and suggest that paracrine regulatory loops between prostate cancer cells and osteoblasts exist and may result in enhanced IGF bioavailability locally in bone by decreasing IGFBP-3 expression. This may in turn potentiate the mitogenic effect of ET-1.
Molecular Analysis of Prostate Cancer Cells Cultured in this in Vitro Model of Bone Metastasis.
We have shown that PMO provides a growth advantage to MDA PCa 2b cells in the in vitro model of bone metastases (Fig. 3A)
. We therefore studied gene expression of MDA PCa 2b cells grown alone and after 4 days of coculture with PMOs by a human gene expression profiling system (Human PathwayFinder-1 GEArray Kit; SuperArray, Inc.) and found a 2-fold increase in MDM2 expression in cocultured cells, compared with controls. The increased expression of MDM2 in MDA PCa 2b cells cocultured with PMOs was confirmed by Northern blot analysis (Fig. 5C)
.
p53 and bcl-2 alterations are involved in the regulation of apoptosis and have been implicated in the progression of prostate cancer to the androgen-independent metastatic phenotype (36
, 37) . MDA PCa 2a and MDA PCa 2b express wild-type p53 and do not overexpress bcl-2 as with 50 and 70% of bone metastasis from prostate cancers, respectively (27)
. We thus examined whether MDA PCa 2a and MDA PCa 2b cells grown with PMO would change the pattern of p53 or bcl-2 expression. No difference was found in p53 and bcl-2 expression as assessed by Western blot analysis (Fig. 5D)
.
Expression of PTHrP, OPG, and RANKL in this in Vitro Model of Bone Metastasis.
To elucidate whether a decrease in osteoclastic bone resorption may contribute to the increase in bone mass observed in prostate cancer bone metastases, we analyzed the expression of three genes involved in the regulation of bone resorption: PTHrP, OPG, and RANKL. As Fig. 6A
shows, the expression of PTHrP was higher in PC3 cells than it was in all other prostate cancer cell lines tested, which is consistent with the fact that PC3 cells induce osteolytic lesions. The expression of OPG was also much higher in PC3 cells than it was in MDA PCa 2a, MDA PCa 2b, and LNCaP cells, which all had undetectable levels of OPG (Fig. 6B)
. RANKL was undetectable in all cell lines tested (data not shown). In vivo data showed that bone lesions induced by MDA PCa 2b cells are associated with increased proliferation and activation of osteoblasts (Fig. 2)
, whereas in vitro, this cell line does not express any significant amount of OPG, an inhibitor of osteoclastic bone resorption. Taken together, these findings suggest that osteoblastic lesions induced by MDA PCa 2b are related to stimulation of osteoblastic activity rather than to an inhibition of osteoclast activation. We then studied whether the expression of these genes is modulated in PMO or MDA PCa 2b cells during coculture. MDA PCa 2b cells and PMOs were grown in this in vitro system of bone metastases, and RNA was harvested after 4 days of coculturing. Results showed that prostate cancer cells induce a decreased expression of OPG and an increased expression of RANKL in PMOs, which may lead to osteoclast activation (Fig. 6C)
. In contrast, OPG and PTHrP expression by prostate cancer cells was not changed when these cells were cocultured with PMOs (data not shown).
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
|
Previous studies using an osteoblastic breast cancer cell line suggested that the new bone formation induced by these cells was ET-1 mediated. Moreover, the new bone formation was inhibited by an ETA receptor antagonist (8) . In preclinical studies, Atrasentan, an antagonist selective for the ETA receptor, blocked ET-1-induced biological responses that were mediated through the ETA receptor. In in vitro studies, Atrasentan is 1800-fold more selective for ETA than the ET receptor B with a Ki of 0.034 nM for the ETA receptor (40) . The results reported here showed that ET-1 partially mediates osteoblast proliferation in this in vitro model of bone metastasis; moreover, Atrasentan decreased mineralization by PMOs. It is therefore expected that systemic administration of Atrasentan would result in a reduction of bone formation. Thus, our results are in agreement with preliminary findings of Phase I/II studies (41, 42, 43) . This underlines the value of this in vitro model of bone metastasis to study the interaction between prostate cancer cells and osteoblasts at the molecular level.
We have shown in this study that the transcript levels of IGFBP-3 produced by MDA PCa 2a and MDA PCa 2b decrease when the cells are cocultured with PMOs. Furthermore, MDA PCa 2a and MDA PCa 2b, as with most human prostate cancers, express PSA, which is a known protease of IGFBP-3. It is therefore likely that alternative and/or additive mechanisms decrease IGFBP-3 expression in prostate cancer, thus increasing the bioavailable levels of IGFs. Numerous in vitro studies have documented the ability of IGF-I to influence osteoblast growth and differentiation (15) . Moreover, targeted overexpression of IGF-I in the osteoblasts of transgenic mice demonstrated an increase in bone formation rate and mineral density (44) .
We also found that IGFBP-3 has antiproliferative effects in PMOs. This effect was attenuated in this in vitro model of bone metastasis, probably attributable to PSA-mediated proteolysis of IGFBP-3. Previous studies have implicated PSA-dependent proteolysis of IGFBP-3 in the osteoblastic phenotype of prostate cancer bone metastases (31 , 32 , 35) . A decrease in serum IGFBP-3 in patients with metastatic prostate cancer and a significant negative correlation between serum PSA and IGFBP-3 were reported (32) . In support of those findings, tissue and serum concentrations of PSA vary inversely with IGFBP-3 levels in patients with metastatic prostate cancer (31 , 35) , and it was recently reported that plasma IGF-I and IGFBP-3 levels are predictors of biochemical progression and advanced stages of prostate cancer (33 , 34) . IGFBP-3 modulation may thus be an essential event in prostate cancer bone metastases. Despite the important role of IGFs in bone, there is a paucity of studies on the effect of IGFBPs on osteoblast growth and differentiation. Our laboratory is currently studying the specific mechanism of IGFBP-3 growth inhibitory effect on osteoblasts.
We have shown that transcript levels of MDM2 are increased in prostate cancer cells grown with PMO (Fig. 5C)
. MDM2 is one of the main negative regulators of the p53 tumor suppressor gene and 50% of bone metastases from prostate cancer are thought to have wild-type p53 (36)
. MDM2 and p53 form an autoregulation loop. The binding between MDM2 and p53 can negatively regulate the transrepression function of p53, inhibit p53-induced apoptosis, and target p53 for degradation (45)
. MDM2 has also non-p53 partners (e.g., Rb, ADP ribosylation factor, E2F1/DP1) that might balance-out its effect on p53 function (45)
. This indicates that the outcome of MDM2-p53 interaction in cancer cells will likely depend on the molecular alterations present in a given tumor cell. The cells that were used in this study (MDA PCa 2a and MDA PCa 2b) have a wild-type p53. Rb and p16INK4A were considered to be normal based on the pattern of protein expression (27)
. On the basis of our current understanding of MDM2-p53-Rb interaction, it is difficult to predict the specific effect of MDM2 overexpression on p53 expression/function. Results presented in Fig. 5D
show that p53 expression levels remain similar to the control cells. However, the oncogenic activities of MDM2 go beyond p53 to increase E2F1-mediated transactivation, thus leading to uncontrolled growth (45)
.
Bone metastases of prostate cancer are recognized for their osteoblastic features, yet osteolytic features are also present in most cases. The mechanisms of prostate cancer-induced osteolysis at the bone site have not been completely defined, however (8) . MDA PCa 2b cells growing in the bone produced osteoblastic reaction, and although osteolysis was also observed, few osteoclasts or scalloped bone surfaces were detected. We therefore analyzed the expression of three major genes involved in bone-resorption, namely PTHrP, OPG, and RANKL. Guise et al. (20) provided strong evidence that PTHrP mediates the osteolytic metastases of breast cancer cells in vivo. We found that PTHrP expression was higher in PC-3 than in MDA PCa 2a and MDA PCa 2b cells. These findings and those of previous studies (8) suggest that the level of expression of PTHrP in prostate cancer cells inversely correlates with the propensity of those cells to form osteoblastic lesions when injected into mouse bone or into human bone implants. Moreover, evidence has been reported that PSA can cleave PTHrP and completely abolish its ability to stimulate cyclic AMP production in vitro (8) . MDA PCa 2a and MDA PCa 2b cells, unlike PC-3 cells, produce and secrete PSA (26) . Taken together, our results suggest that PTHrP does not play a major role in the resorptive component of prostate cancer.
Expression of OPG, a factor that inhibits the formation and activity of osteoclasts by sequestering RANKL, was not detected in MDA PCa 2b cells. Moreover, coculturing osteoblasts with prostate cancer cells resulted in a decreased expression of OPG by the osteoblasts. This dramatic decrease of OPG in the bone microenvironment might initially induce activation of osteoclasts, which is thought to be a prerequisite for tumor growth in bone. These results suggest that OPG may be a candidate chemopreventive drug for bone metastases from prostate cancer. Evidence that OPG may inhibit tumor growth in the bone was previously provided in vivo (46) , and recent publications have also shown that OPG prevented the establishment of prostate cancer tumors in bone (47) . Expression of OPG and RANKL in human prostate cancer was recently reported to be significantly increased in bone metastases when compared with primary prostate cancer or nonskeletal metastases (48) . The OPG/RANKL ratio may therefore be an important factor influencing the degree of osteolysis in a given bone metastasis.
The results of this study demonstrate that reproducible molecular events occur in this model of bone metastases from prostate cancer. Our data thus provide a rationale for developing therapeutic strategies designed to target these molecular changes in an effort to improve the morbidity and mortality rates associated with bone metastases of prostate cancer. Indeed, strategies to inhibit the main targets identified in this study (namely IGF-I and mdm2) and to use OPG as a therapeutic agent have been developed at the bench, and some early clinical trials are currently under way (24 , 49, 50, 51, 52) .
|
FOOTNOTES |
|---|
1 This work was supported by Association for the Cure of Cancer of the Prostate, DAMD17-991-9028, NIH Grants CA75499-04, DK50583, AR45548, and the Ligue contre le cancer. 
2 K. F. and J. Y. contributed equally to this work.
3 To whom requests for reprints should be addressed, at Department of GU Medical Oncology, M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Box 013, Houston, TX 77030. Phone: (713) 792-2898; Fax: (713) 792-5696; E-mail: nnavone{at}mail.mdanderson.org
4 The abbreviations used are: PSA, prostate-specific antigen; ET-1, endothelin-1; PMO, primary mouse osteoblast; IGF, insulin-like growth factor; IGFBP-3, IGF-binding protein 3; OPG, osteoprotegerin; IGF-IR, IGF-I receptor; IGFBP, IGF-binding protein; PTHrP, parathyroid hormone-related protein; MMP, matrix metalloproteinase; RANKL, nuclear factor-B ligand; FBS, fetal bovine serum; ECM, extracellular matrix; SCID, severe combined immunodeficient disease; Rb, retinoblastoma; MDM2, mouse double minute 2.
Received 10/ 3/02; revised 1/ 6/03; accepted 2/14/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Fizazi, P. Beuzeboc, J. Lumbroso, V. Haddad, C. Massard, M. Gross-Goupil, M. Di Palma, B. Escudier, C. Theodore, Y. Loriot, et al. Phase II Trial of Consolidation Docetaxel and Samarium-153 in Patients With Bone Metastases From Castration-Resistant Prostate Cancer J. Clin. Oncol., May 20, 2009; 27(15): 2429 - 2435. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Fizazi, A. Lipton, X. Mariette, J.-J. Body, Y. Rahim, J. R. Gralow, G. Gao, L. Wu, W. Sohn, and S. Jun Randomized Phase II Trial of Denosumab in Patients With Bone Metastases From Prostate Cancer, Breast Cancer, or Other Neoplasms After Intravenous Bisphosphonates J. Clin. Oncol., April 1, 2009; 27(10): 1564 - 1571. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Kearns, S. Khosla, and P. J. Kostenuik Receptor Activator of Nuclear Factor {kappa}B Ligand and Osteoprotegerin Regulation of Bone Remodeling in Health and Disease Endocr. Rev., April 1, 2008; 29(2): 155 - 192. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Fizazi The role of Src in prostate cancer Ann. Onc., November 1, 2007; 18(11): 1765 - 1773. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-J. Cheng, X.-c. Ye, F. Vakar-Lopez, J. Kim, S.-M. Tu, D.-T. Chen, N. M. Navone, L.-Y. Yu-Lee, S.-H. Lin, and M. C-T. Hu Bone Microenvironment and Androgen Status Modulate Subcellular Localization of ErbB3 in Prostate Cancer Cells Mol. Cancer Res., July 1, 2007; 5(7): 675 - 684. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X.-H. Liu, A. Kirschenbaum, S. Yao, G. Liu, S. A. Aaronson, and A. C. Levine Androgen-Induced Wnt Signaling in Preosteoblasts Promotes the Growth of MDA-PCa-2b Human Prostate Cancer Cells Cancer Res., June 15, 2007; 67(12): 5747 - 5753. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. A. Samani, S. Yakar, D. LeRoith, and P. Brodt The Role of the IGF System in Cancer Growth and Metastasis: Overview and Recent Insights Endocr. Rev., February 1, 2007; 28(1): 20 - 47. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Vessella and E. Corey Targeting factors involved in bone remodeling as treatment strategies in prostate cancer bone metastasis. Clin. Cancer Res., October 15, 2006; 12(20): 6285s - 6290s. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Dai, J. Keller, J. Zhang, Y. Lu, Z. Yao, and E. T. Keller Bone Morphogenetic Protein-6 Promotes Osteoblastic Prostate Cancer Bone Metastases through a Dual Mechanism Cancer Res., September 15, 2005; 65(18): 8274 - 8285. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
0.05 respect to controls (PMOs grown alone); **, P < 0.05 respect to PMOs in coculture with MDA PCa 2b cells. These findings were reproduced in an independent experiment.
