Skip Navigation


Carcinogenesis Advance Access originally published online on May 2, 2008
Carcinogenesis 2008 29(8):1509-1518; doi:10.1093/carcin/bgn105
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
29/8/1509    most recent
bgn105v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Yuen, H.-F.
Right arrow Articles by Chan, K.-W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yuen, H.-F.
Right arrow Articles by Chan, K.-W.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

TWIST modulates prostate cancer cell-mediated bone cell activity and is upregulated by osteogenic induction

Hiu-Fung Yuen1,2, Wai-Kei Kwok2, Ka-Kui Chan1, Chee-Wai Chua2, Yuen-Piu Chan1, Ying-Ying Chu3, Yong-Chuan Wong2, Xianghong Wang2 and Kwok-Wah Chan1,*

1 Department of Pathology
2 Department of Anatomy
3 Department of Chemistry, The University of Hong Kong, Pokfulam, Hong Kong, China

* To whom correspondence should be addressed. Department of Pathology, Queen Mary Hospital, Hong Kong, China. Tel: +852 28554874; Fax: +852 28725197; Email: kwchan{at}pathology.hku.hk

Correspondence may also be addressed to Xianghong Wang, Department of Anatomy, The University of Hong Kong, 1/F, Faculty of Medicine Building, 21 Sassoon Road, Hong Kong, China. Tel: +852 28192868; Fax: +852 28170857; Email: xhwang{at}hkucc.hku.hk


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 References
 
TWIST, a helix-loop-helix transcription factor, is highly expressed in many types of human cancer. We have previously found that TWIST confers prostate cancer cells with an enhanced metastatic potential through promoting epithelial–mesenchymal transition (EMT) and a high TWIST expression in human prostate cancer is associated with an increased metastatic potential. The predilection of prostate cancer cells to metastasize to bone may be due to two interplaying mechanisms (i) by increasing the rate of bone remodeling and (ii) by undergoing osteomimicry. We further studied the role of TWIST in promoting prostate cancer to bone metastasis. TWIST expression in PC3, a metastatic prostate cancer cell line, was silenced by small interfering RNA and we found that conditioned medium from PC3 with lower TWIST expression had a lower activity on stimulating osteoclast differentiation and higher activity on stimulating osteoblast mineralization. In addition, we found that these effects were, at least partly, associated with TWIST-induced expression of dickkopf homolog 1 (DKK-1), a factor that promotes osteolytic metastasis. We also examined TWIST and RUNX2 expressions during osteogenic induction of an organ-confined prostate cancer cell, 22Rv1. We observed increased TWIST and RUNX2 expressions upon osteogenic induction and downregulation of TWIST through short hairpin RNA reduced the induction level of RUNX2. In summary, our results suggest that, in addition to EMT, TWIST may also promote prostate cancer to bone metastasis by modulating prostate cancer cell-mediated bone remodeling via regulating the expression of a secretory factor, DKK-1, and enhancing osteomimicry of prostate cancer cells, probably, via RUNX2.

Abbreviations: ALP, alkaline phosphate; ChIP, chromatin immunoprecipitation; DKK-1, dickkopf homolog 1; EMT, epithelial–mesenchymal transition; FBS, fetal bovine serum; PCR, polymerase chain reaction; TRACP, tartrate-resistant acid phosphatase


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Supplementary material
 References
 
Prostate cancer has been estimated to be the most common newly diagnosed cancer and the second leading cause of cancer-related death in men in the USA in 2007 (1). Previous studies have shown that patients with localized prostate cancer have a very high 5-year survival rate and a relatively low mortality to incidence ratio compared with other cancer types (2). Based on these observations, intense treatment toward prostate cancer might not be beneficial to the patients. However, a group of patients with micrometastatic disease at the time of diagnosis, which is mostly unidentified due to lack of sensitive detection method, will eventually develop clinically detectable metastatic disease, at which point the median survival is reduced to only 12–15 months (3).

The primary target of prostate cancer metastasis is bone that accounts for ~85% cases of metastatic prostate cancer. Zoledronic acid, a new form of bisphosphonate that functions to reduce bone resorption, has been recently shown to decrease the risk for and delay the occurrence of skeleton-related events in metastatic prostate cancer (4,5). This suggests that inhibition of bone resorption or remodeling might be an effective treatment for prostate cancer. Prostate cancer cells interact with osteoblasts and osteoclasts in order to survive and thrive in bone environment. They rely on this viscous cycle, cancer cell–microenvironment interaction, to establish overt metastases in bone. Disruption of this cycle has been suggested to be a target for treatment of prostate cancer (6). Although prostate cancer is usually osteoblastic in nature, an early phase of osteolytic event induced by prostate cancer has been suggested to be crucial for the establishment of bone metastasis (7–9). This early osteolytic event may be one of the reasons why bisphosphonate, a bone resorption inhibitor, is effective in suppressing prostate cancer to bone metastasis. By better understanding the interaction among prostate cancer cells, osteoclasts and osteoblasts, we may be able to develop effective therapy against prostate cancer to bone metastasis.

Recent studies have shown that the Wnt signaling pathway to be one of the most important paracrine interactions among prostate cancer cells, osteoclasts and osteoblasts. Expressions of Wnt-1 and beta-catenin are significantly elevated in patients with advanced metastatic prostate cancer (10). dickkopf homolog 1 (DKK-1), an inhibitor of canonical Wnt signaling, has been shown to be highly expressed in osteolytic, but not osteoblastic, prostate and mammary cancer cell lines (11,12). When highly expressed in prostate cancer cells, DKK-1 blocks Wnt-mediated osteoblastic activity and induces osteolytic bone metastasis in vivo (12). On the other hand, bone cells also stimulate growth of prostate cancer cells by Wnt signaling through paracrine action (13).

Osteomimicry, the acquisition of osteoblast-like phenotype through an increased expression of several important bone factors has been described in prostate and breast cancers (14,15). It enables the cancer cells to reside and thrive in bone. Its onset is promoted in part by osteoblasts (16). Factors recently identified to promote osteomimicry in prostate cells are extracellular signal-regulated kinase in Notch signaling (17), osteocalcin and bone sialoprotein (18). RUNX2 is one of the most important factors during acquisition of osteomimicry. RUNX2 has been shown to modulate the expression of bone sialoprotein in human prostate cancer cells (19) and promote osteolytic breast cancer to bone metastasis (20,21). RUNX2 is also expressed in human prostate cancer cell line PC3 and is responsible for the formation of RUNX2–osteoblast specific element 2 complex that promotes expression of bone-related factors (22).

Studies have shown that TWIST has an important role on bone development. TWIST is expressed in primary osteoblastic cells (23) and preosteoblasts (24). Its expression is reduced with maturation of these cells. TWIST expression level, when altered, modifies the differentiation state of human osteoblast cells (25). These data suggest that TWIST may play a role during early phase osteogenic differentiation and may act as a key switch for bone cell differentiation. In addition, reduced TWIST expression in osteoblasts decreases the expression of RUNX2 (26). TWIST has also been shown to interact with and inhibit RUNX2 transcription activity to inhibit osteoblast differentiation and bone formation (27). On the other hand, TWIST co-operates with homeobox, msh-like 2 to promote proliferation and differentiation of skeletogenic mesenchyme (28). Although most studies suggested that TWIST might function to maintain osteogenic lineage by preserving more osteoprogenitor-like features and inhibit terminal differentiation of osteoblasts, its exact functions during osteoblast differentiation still remain elusive and may be cell-type specific (29).

Recently, we have demonstrated that TWIST confers taxol resistance to prostate cancer cells (30) and promotes prostate cancer progression through epithelial–mesenchymal transition (EMT) (31). Others have also shown that TWIST enhances the metastatic potential of breast cancer (32), another cancer that has a strong predilection to undergo bone metastasis. TWIST has been shown to control the expression of a microRNA, which leads to increased metastatic ability of cancer cells (33). In addition, our recent analysis of 115 human prostate cancer specimens shows that a higher TWIST expression in the primary cancer is associated with higher risks for the development of metastasis, 95% of which were to bone in our patient cohort (34). These results led us to hypothesize that TWIST expression in prostate cancer cells might play a positive role in the development of bone metastasis. This study aimed to elucidate the role of TWIST in prostate cancer to bone metastasis. Specifically, we determined how TWIST expression in prostate cancer cells had modulated prostate cancer cell-mediated osteolytic or osteoblastic bone metastasis, and what role TWIST had to play in osteomimicry of prostate cancer cells.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Supplementary material
 References
 
Cell culture conditions
Human prostate cancer cell lines 22Rv1 (from Prof. Franky Chan, Department of Anatomy, The Chinese University of Hong Kong), LNCaP (American type culture collection, Manassas, VA) and PC3 (American type culture collection) were maintained in RPMI 1640 (Invitrogen, Carlsbad, CA) supplemented with 10, 5 and 5% fetal bovine serum (FBS) (Invitrogen), respectively. C42B (from Prof. Franky Chan), a highly metastatic derivative of LNCaP, was maintained in T-medium (Invitrogen) supplemented with 10% FBS. Murine macrophage cell line RAW264.7 and murine preosteoblast cell line MC3T3-E1 clone 14 (both from American type culture collection) were maintained in Dulbecco's modified Eagle's medium and alpha minimum essential media (Invitrogen), respectively, and supplemented with 10% FBS. PC3 pSuppressor stable transfectants were generated previously (31) and maintained in RPMI 1640 supplemented with 5% FBS and intermittently selected with addition of 600 µg/ml G418 (Sigma, St. Louis, MO). Human embryonic kidney cell line 293 was maintained in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% FBS.

Plasmids
FLAG-tagged TWIST expression plasmid, pcDNA3.1 FLAG-tagged TWIST was a generous gift from Prof. L.Kedes. DKK-1 expression plasmid, pcDNA3.1-DKK1, was constructed by cloning the DKK-1 coding region, amplified using PC3 complementary DNA and primers Dkk1F–BamH1 (5'-TAGGGATCCATGATGGCTCTGGGCGCAGC-3') and Dkk1R–EcoR1 (5'-GCCGAATTCTTAGTGTCTCTGACAAGTGT-3'), into pcDNA3.1 (Invitrogen). Dkk-1 reporter luciferase plasmid, pDkk-1pro, was constructed by cloning –1000 to +141 promoter region of Dkk-1, amplified by using PC3 genomic DNA with primers Dkk1proF (5'-ACTGGTACCGAAGCGTTGCGATGTGATA-3') and Dkk1proR (5'-TCGAAGCTTAAGGACTCAAGAGGGAGAA-3'), into pGL-3 (Promega, Maison, WI). Amplification of DKK-1 coding sequence and promoter region was performed using high fidelityPLUS PCR System (Roche Molecular Biochemicals, Indianapolis, IN) and sequencing was performed to confirm correct sequence of the insert. RUNX2-responsive reporter plasmid, p147-luc, was a generous gift from Prof. G.Karsenty. The BglII–HindIII fragment of this p147-luc containing the two OSE elements from OG2 promoter was cloned into pGL3 to form a pGL3-OG2 construct. pSIREN-siLUC, siTWIST1 and siTWIST2 plasmids were constructed according to the manufacturer's instruction (Clontech Laboratory, Palo Alto, CA; KnockoutTM RNAi-Ready System) by using linearized pSIREN and a short self-complementary oligo [control siLuc: provided by the manufacturer, siTWIST sequences shTWIST1 and shTWIST2 were designed according to descriptions given by Valsesia-Wittmann et al. (35)].

Western blotting
Western blot was performed as described previously (31). TWIST antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). DKK-1 and BMP-6 antibodies were both from R&D Systems (Minneapolis, MN). RUNX2, FLAG and actin antibodies were from Sigma.

Reverse transcription–polymerase chain reaction
Reverse transcription–polymerase chain reaction (PCR) was performed as described previously (36). Primer sequences are listed in supplementary Table I (available at Carcinogenesis online).

Collection of conditioned medium
For stable PC3 transfectants, we seeded 1.5 x 106 cells into a T-75 flask and then replaced the medium with RPMI 1640 with 0.5% FBS after 24 h. The conditioned medium was collected after 48 h, stored at –80°C and normalized by the content of genomic DNA through adding RPMI 1640 as described by Dai et al. (37) before use. For transient transfectants in DKK-1 rescue analysis, 1.5 x 105 cells were seeded into a six-well plate. After 24 h, transfection of DKK-1 expression plasmid was performed by Fugene 6 (Roche Molecular Biochemicals). The medium was changed into RPMI 1640 with 0.5% FBS 24 h after transfection. Conditioned medium was then collected after 48 h and stored at –80°C before use. The conditioned medium from transient transfectants was not normalized by the DNA content because the activity of the cells might be dependent on the transfection efficiency and therefore, the protein level of transiently transfected DKK-1 was checked by western blot every time before the use of the conditioned medium.

In vitro osteoclastogenesis assay
The 50% conditioned medium was a mixture of equal volume of conditioned medium from PC3 transfectants and normal medium (Dulbecco's modified Eagle's medium–10% FBS) of RAW264.7 cells. Negative control was a mixture of 50% RPMI 1640 (medium used for PC3 cells culture) and 50% normal medium of RAW264.7 cells. The 50% conditioned medium as well as the negative control was supplemented with 5 ng/ml receptor activator of nuclear factor B ligand (R&D Systems) as described (38). Five thousand RAW264.7 cells were seeded into 96-well plate. Medium was then replaced with 50% conditioned medium on day 1. Medium was changed on day 4. The number of differentiated osteoclasts was determined by tartrate-resistant acid phosphatase (TRACP) staining assay kit (Sigma). When counted under a light microscope, only a red-stained cell with three or more nuclei was counted as a differentiated osteoclast.

In vitro osteoblast mineralization
The 50% conditioned medium was a mixture of equal volume of conditioned medium from PC3 transfectants and normal medium of MC3T3-E1 cells. Negative control was 50% RPMI 1640 and 50% normal medium of MC3T3-E1 cells, whereas positive control was 50% RPMI 1640 and 50% normal medium of MC3T3-E1 cells with 100 ng/ml BMP-6 supplementation. The 50% conditioned medium and the control were then supplemented with 10 mM β-glycerophosphate and 50 µg/ml ascorbic acids. Five thousand MC3T3-E1 cells were seeded into 24-well plate. The culture medium was changed every three successive days. On day 9, medium was replaced with 50% conditioned medium and changed every three following successive days. On day 18, cells were stained for calcium deposition by alizarin red S staining and von Kossa method.

Alizarin red S staining
Cells were air-dried for 10 min, fixed in 50% ethanol at room temperature for 10 min thrice, stained in 10 mg/ml alizarin red for 5 min and then washed with phosphate-buffered saline until all precipitates were cleared.

von Kossa method
Cells were fixed in 95% ethanol at 37°C for 5 min, rehydrated with graded ethanol and then washed with water for five times. Cells were then stained in 5% silver nitrate under bright light until they turned black. They were washed five times with water. The action of silver nitrate was stopped by 5% sodium thiosulphate and the cells were then washed again with water for five times.

Determination of alkaline phosphatase and TRACP activities
Similar cultivation conditions were used as those for in vitro osteoclastogenesis and mineralization assays. Cells in a 96-well plate were lysed. The activities of alkaline phosphate (ALP) and TRACP were determined using TRACP & ALP assay kit (TaKaRa Bio, Dalian, China) according to the manufacturer's instruction.

Chromatin immunoprecipitation
Chromatin immunoprecipitation (ChIP) analysis was performed according to the manufacturer's instruction (Upstate Biotechnology, Lake Placid, NY). Briefly, cells were fixed by formaldehyde for 48 h after transfection, washed with phosphate-buffered saline and lysed by sodium dodecyl sulfate lysis buffer. Lysate was then agitated gently with anti-FLAG antibody (Sigma) overnight at 4°C. The antibody–protein–DNA complex was pulled down by Protein A agarose/Salmon Sperm DNA. Cross-links in the complex were reversed by 0.1 M NaCl and DNA was recovered by chloroform purification and ethanol precipitation. The DNA was then subjected to PCR amplification of the Dkk-1 promoter region containing E-box-1 (forward primer: 5'-CCTCCCTCTCTAAACTTCCCA-3' and backward primer: 5'-GGGTGCAAGTTGCTCATTAACCCT-3') and E-box-2 (forward primer: 5'-AGGACCTCAAAGCCGGGGATGTA-3' and backward primer: 5'-ACAGAGCCGAGGGGTGATA-3').

Luciferase reporter assay
Luciferase assay was performed as described previously (36).

Osteogenic induction
22Rv1 cells (1 x 104) were seeded into a six-well plate and allowed to grow for 24 h. Medium was replaced by osteogenic medium (RPMI 1640 supplemented with 10% FBS, 10 mM β-glycerophosphate and 50 µg/ml ascorbic acids) and changed every three successive days. After 10 days of osteogenic induction, cells were harvested for RNA and protein level analyses.

Generation of 22Rv1 siTWIST transfectant
pSIREN-siLUC, siTWIST1 and siTWIST2 constructs were transfected to retroviral packaging cell line by Fugene 6 and the viral particles in the culture medium were harvested. The medium was filtered and then applied to 22Rv1 cells. The viral containing medium was then replaced by RPMI 1640 with 10% FBS after 24 h. Cells were subjected to osteogenic induction as described above.

Statistical analysis
Statistical analysis was performed using SPSS 13.0 software (SPSS, Chicago, IL). Mann–Whitney U test was used to test the significance of the difference between samples and P < 0.05 was considered significant.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Supplementary material
 References
 
TWIST was expressed in osteolytic but not osteoblastic prostate cancer cell lines
We performed reverse transcription–PCR and western blotting to study if there was a differential TWIST expression between osteolytic and osteoblastic prostate cancer cell lines. As shown in Figure 1, TWIST was highly expressed in PC3, a highly osteolytic prostate cancer cell line. It was weakly expressed in 22Rv1, an organ-confined prostate cancer cell line, which was established from primary prostate carcinoma through serial propagation of a xenograft CWR22 in mice (39). Its expression was not detected in the two osteoblastic cell lines, LNCaP and its highly metastatic derivative, C42B. Previous study in our laboratory has demonstrated that TWIST expression is high in DU145, another osteolytic prostate cancer cell line (31). DKK-1 has been described as a mediator that controls the transition from osteolytic to osteoblastic cycle in bone metastasis of prostate cancer and is highly expressed in osteolytic prostate cancer cells (11,12). In this study, we found that the level of DKK-1 was positively associated with TWIST expression. RUNX2, on the other hand, was highly expressed in two metastatic prostate cancer cell lines, C42B and PC3, but it was low in LNCaP and 22Rv1 cells. The results suggest that increased expression of TWIST and DKK-1 was correlated with osteolytic ability of prostate cancer cells.


Figure 1
View larger version (50K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Differential expression of TWIST, RUNX2 and DKK-1 in prostate cancer cell lines. The left panel shows RNA level determined by reverse transcription (RT)–PCR. The right panel shows protein level determined by western blot analysis. TWIST was highly expressed in PC3, an osteolytic prostate cancer cell line, and weakly expressed in 22Rv1, an organ-confined prostate cancer cell line. It was not expressed in C42B, an osteoblastic prostate cancer cell line, and LNCaP, a lymph node-metastasized prostate cancer cell line. RUNX2 was highly expressed in PC3 and C42B, both of which have high bone metastatic ability. Its expression was lower in 22Rv1 and barely detectable in LNCaP. DKK-1 expression was high in osteolytic (PC3) but low in osteoblastic (C42B) prostate cancer cell lines. Note that TWIST and DKK-1 expressions in the four prostate cancer cell lines studied were correlated. A ratio of the target band intensity to reference (PC3) band intensity is shown below the results. A relative ratio (to PC3) of the target band intensity adjusted by actin is shown below the results.

 
Suppression of TWIST in PC3 inhibited PC3-mediated osteoclast differentiation
To investigate the significance of the positive correlation between TWIST and osteolytic ability of prostate cancer cells, we went on to study how changes in TWIST expression in prostate cancer cell could affect the extent of prostate cancer cell-induced osteoclast differentiation. Previously, three PC3 sublines were generated that showed differential TWIST expression. While PC3-siCon expressed a high level of TWIST, PC3-siTWIST c1 and c2 cell lines expressed much lower levels of TWIST because each of them had a stably transfected vector containing a TWIST-targeting small interfering RNA sequence (31). In this study, conditioned media from these three cell lines were used to treat a murine macrophage cell line, RAW264.7, and the ability of the conditioned medium to induce osteoclast-like differentiation of RAW264.7 was studied. As shown in Figure 2a, in the RAW264.7 cells treated with conditioned medium from PC3-siCon, differentiated osteoclast-like cells were observed (red arrows, left square) but the number of osteoclast-like cells was lower when RAW264.7 cells were treated with conditioned media from PC3-siTWIST c1 and c2 cell lines (two squares in the middle) or the negative control (right square). Quantitative analysis showed significantly lower numbers of differentiated osteoclast-like cells when RAW264.7 cells were treated with conditioned media from PC3-siTWIST c1 and c2 cells (light gray column, Figure 2b) than when RAW264.7 cells were treated with PC3-siCon-conditioned medium (dark gray column, P < 0.05, Figure 2b). These results show that conditioned medium from PC3 cells could enhance the receptor activator of nuclear factor B ligand-induced osteoclast differentiation and by suppressing TWIST expression in PC3 cells, the conditioned medium was less able to induce osteoclast differentiation. A similar conclusion is derived by examining the activity of TRACP, an enzyme indicative of osteoclast differentiation, in RAW264.7 cells treated with conditioned media from the above three cell lines. As shown in Figure 2c, RAW264.7 cells cultured in conditioned media from PC3-siTWIST c1 and c2 had lower TRACP activities (light gray column; decreased by 29 and 36%, respectively) than those cultured in PC3-siCon-conditioned medium (dark gray column, P < 0.05). These results suggest that PC3 cells with low level of TWIST had a lower ability to induce TRACP activity in RAW264.7 cells. Taken together, the data show that TWIST plays a positive role in prostate cancer-mediated osteoclast activity.


Figure 2
View larger version (49K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. TWIST expression in PC3-modulated PC3-mediated bone cell activity. TWIST expression was reduced in PC3 through small interfering RNA technology. Conditioned media from three PC3 sublines with differential TWIST expression level were collected and their ability to induce osteoclast and osteoblast activities or differentiations was measured. The ability to induce osteoclast activity was determined by the number of differentiated osteoclasts formed and TRACP activities in RAW264.7 after treatment with conditioned medium. To measure the ability to induce osteoblast activity, mineralization and ALP activity in MC3T3-E1 after treatment with conditioned medium were determined. Note that suppression of TWIST expression in PC3 reduced the ability of its conditioned medium to induce osteoclast differentiation and enhanced the ability of its conditioned medium to promote osteoblast mineralization. (a) Differentiated osteoclasts with red staining and three or more nuclei were formed more frequently when RAW264.7 cells were treated with conditioned medium from PC3-siCon comparing with those from PC3-siTWIST c1 or c2. Photos were taken under x200 magnifications in light microscope (differentiated osteoclasts were indicated by red arrows). (b) Number of differentiated osteoclasts and (c) TRACP activities were significantly reduced (both P < 0.05) when RAW264.7 cells were treated with conditioned media from PC3-siTWIST c1 and c2 (light gray columns) compared with that from PC3-siCon (dark gray column). (d) Calcium deposition in MC3T3-E1 cells was demonstrated by alizarin red S staining (upper panel) and von Kossa method (lower panel). MC3T3-E1 cells treated with conditioned media from PC3-siTWIST c1 and c2 (wells 2, 3, 6 and 7) showed more intense staining than those treated with PC3-siCon-conditioned medium (wells 1 and 5). (e) ALP activity in MC3T3-E1 cells treated with conditioned medium from PC3-siTWIST transfectants (light gray column) was higher compared with PC3-siCon (dark gray column). Error bars represent ±1 SD from at least three independent experiments and asterisk represents significant difference (P < 0.05) compared with PC3-siCon.

 
Suppression of TWIST in PC3 promoted PC3-mediated osteoblast mineralization
To investigate the effect of TWIST in prostate cancer cell-mediated osteoblast activity, we studied the ability of conditioned media from PC3-siCon, siTWIST c1 and c2 to induce osteoblast mineralization in a murine preosteoblast cell line, MC3T3-E1, cultured in a mineralization medium. Calcium deposition was detected by alizarin red S staining and von Kossa method. As shown in Figure 2d, calcium deposition of MC3T3-E1 cells treated with conditioned medium from PC3-siCon (wells 1 and 5) was much less compared with those treated with conditioned media from PC3-siTWIST c1 (wells 2 and 6) and c2 (wells 3 and 7) in alizarin red S staining (red stained, upper panel) and von Kossa method (black stained, lower panel). The results suggest that by suppressing TWIST expression in prostate cancer cells, their ability to induce osteoblast mineralization would be increased. A similar conclusion is obtained when we consider the results of our study on ALP activity. As shown in Figure 2e, MC3T3-E1 cells had a lower activity of ALP when treated with conditioned medium from PC3-siCon (dark gray column) compared with those treated with conditioned media from PC3 transfectants with lower TWIST expression (PC3-siTWIST c1 and c2, light gray column, P < 0.05). Taken together, these results indicate that suppression of TWIST in prostate cancer cells would increase their ability to induce osteoblast activity.

Effect of TWIST on expression levels of several bone cell factors in PC3 cells
TWIST is a transcription factor that functions through altering expression level of its downstream target genes. A recent study using ChIP-chip analysis has revealed that TWIST target genes in Drosophila are far more than predicted previously (40). Since TWIST itself is not a secretory factor, it is probable that the effect of TWIST in prostate cancer on bone cell activity is mediated through its transcriptional regulation of bone cell factors. To test this hypothesis, we analyzed whether TWIST could alter the expression levels of several bone cell factors in PC3-siCon, PC3-siTWIST c1 and c2. As shown in Figure 3, messenger RNA level of bone cell factors was estimated by reverse transcription–PCR (left panel) and protein levels of some of them were tested by western blotting (right panel). We found that while DKK-1, BMP-6, connective tissue growth factor and fibroblast growth factor 2 levels were lower in cells with a low level of TWIST, the expression of RUNX2 and endothelin 1 was higher in PC3-siTWIST cells. These results suggest that TWIST is involved in regulating the expression of the bone factors.


Figure 3
View larger version (47K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. TWIST altered the expression of factors in PC3 that were shown to be involved in prostate cancer to bone metastasis. Messenger RNA and protein were extracted from PC3-siCon, PC3-siTWIST c1 and c2. Left and right panels show results of reverse transcription–PCR and western blotting, respectively. Expression of DKK-1, bone morphogenetic protein-6, connective tissue growth factor and fibroblast growth factor 2 was decreased, whereas RUNX2 and endothelin 1 were increased in PC3 with reduced TWIST expression. Due to TWIST having many transcription targets, alterations of its expression might lead to changes in the expression of many genes, which contributed, synergistically or additionally, to the overall effect of TWIST in prostate cancer-mediated bone cell activity. Reverse transcription–PCR primers are listed in supplementary Table I (available at Carcinogenesis Online). A relative ratio (to PC3-siCon) of the target band intensity adjusted by actin is shown below the results.

 
TWIST positively regulated DKK-1 expression at transcriptional level
TWIST and DKK-1 expressions were positively correlated among the four prostate cancer cell lines we had tested (Figure 1). Furthermore, suppression of TWIST expression in PC3 cells resulted in decreased expression of Dkk-1 (Figure 3). We also detected a decrease in the protein level of DKK-1 in the conditioned medium from PC3 cells expressing lower levels of TWIST (supplementary Figure 1 is available at Carcinogenesis Online). Next, we analyzed whether TWIST could regulate Dkk-1 expression at transcriptional level. Promoter region of Dkk-1, from –1000 to +141 bp relative to the transcriptional start site, was cloned into pGL3 plasmid for luciferase reporter analysis (Figure 4a). The length of the promoter selected for analysis was based on a previous Dkk-1 promoter study (41). We transfected the Dkk-1 reporter construct into PC3-siCon, PC3-siTWIST c1 and c2 and measured the luciferase activity in the transient transfectants. As shown in Figure 4b, the luciferase activity was reduced by 58 and 52% in PC3-siTWIST c1 and c2 cells, respectively (P < 0.001, upper graph, light gray columns), relative to that in PC3-siCon cells (upper graph, dark gray columns), suggesting that downregulation of TWIST expression had decreased promoter activity of the Dkk-1 gene. To further confirm this finding, we performed cotransfection of a TWIST expression vector with the Dkk-1 reporter construct into 293 cells. We found that there was ~3.14-fold increase in luciferase activity (lower graph, open column) when compared with cotransfection of a mock vector and the Dkk-1 reporter into 293 cells (filled column). This result indicates that overexpression of TWIST resulted in increased Dkk-1 promoter activity. Because of the presence of two E-box elements on the Dkk-1 promoter region, we proceeded to examine whether TWIST was able to bind to Dkk-1 promoter region by ChIP analysis. Our results showed that two fragments of DNA containing either one of the two E-box elements could be amplified from FLAG-TWIST-transfected 293 cells but not from 293 cells transfected with a control vector (Figure 4c). Future mutational analysis on these two E-box elements may be able to answer whether TWIST binds to DKK-1 promoter through these two E-box elements. Nonetheless, all these results suggest that TWIST positively regulates Dkk-1 expression at transcriptional level possibly through interaction with the Dkk-1 promoter region.


Figure 4
View larger version (30K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4. Dkk-1 is a transcriptional target of TWIST. Ability of TWIST to transcriptionally regulate Dkk-1 was determined by luciferase reporter and ChIP assay. (a) A schematic diagram to show the pDkk-1pro construct. Two E-boxes, putative binding sequences of TWIST, were identified in –1000 bp upstream region of Dkk-1 structural gene; –1000 to +141 bp Dkk-1 promoter region was cloned into pGL-3 luciferase reporter plasmid. The length of the promoter selected for analysis was based on a previous Dkk-1 promoter study (41). (b) Reduction in TWIST expression decreased Dkk-1 promoter activity (upper panel) and cotransfection of TWIST expression vector with pDkk-1pro increased Dkk-1 promoter activity (lower panel). (c) The DNA pulled down by ChIP was subjected to PCR amplification of Dkk-1 promoter region containing E-box-1 (forward primer: 5'-CCTCCCTCTCTAAACTTCCCA-3' and backward primer: 5'-GGGTGCAAGTTGCTCATTAACCCT-3') and E-box-2 (forward primer: 5'-AGGACCTCAAAGCCGGGGATGTA-3' and backward primer: 5'-ACAGAGCCGAGGGGTGATA-3'). Dkk-1 promoter region containing the two E-boxes was pulled down by anti-FLAG antibody in ChIP analysis suggesting that TWIST might bind to the Dkk-1 promoter region. Error bars represent ±1 SD from at least three independent experiments and asterisks represent significant difference (P < 0.05) compared with dark gray column.

 
Effect of TWIST in prostate cancer cell-mediated bone activity might be regulated through DKK-1
To further investigate the role of DKK-1 in TWIST-modulated prostate cancer cell-mediated bone activity, DKK-1 expression was restored in the PC3-siTWIST c1 and c2 cells by transfecting them with a DKK-1 expression vector. Western blotting showed that DKK-1 expression in siTWIST transfectants transiently transfected with DKK-1 expression vector was similar to levels found in PC3-siCon cells or PC3-siCon cells transfected with mock vector (Figure 5a). Conditioned media were collected from PC3-siCon cells transfected with mock vector and PC3-siTWIST cells transfected with DKK-1 expression vector. The conditioned media were then applied to RAW264.7 cells. As shown in Figure 5b, conditioned media from PC3-siTWIST transfectants with DKK-1 expression restored were able to induce RAW264.7 cells to undergo osteoclast differentiation at a level similar to that of conditioned medium from PC3-siCon transfected with mock vector (red arrows indicate differentiated osteoclasts). Quantitative analysis (Figure 5c) showed that the number of differentiated osteoclasts was similar between RAW264.7 cells treated with conditioned media from PC3-siCon transfected with mock vector (dark gray column) and from PC3-siTWIST transfectants transfected with a DKK-1 expression vector (light gray columns). In addition, overexpression of DKK-1 in PC3-siTWIST transfectants also enhanced the ability of their conditioned media to induce TRACP activity in RAW264.7 cells (Figure 5d, light gray column) to a level similar to that of PC3-siCon (dark gray column). Moreover, we found that conditioned media from PC3-siTWIST cells transfected with DKK-1 expression vector (Figure 5e, wells 5, 6, 11 and 12) and conditioned medium from PC3-siCon transfected with mock vector (wells 4 and 10) had similar ability to induce mineralization in MC3T3-E1 cells. ALP activity in MC3T3-E1 cells treated with conditioned media from PC3-siCon transfected with mock vector (dark gray column, Figure 5f) and PC3-siTWIST transfectants transfected with DKK-1 expression vector (light gray column) was also similar. These results demonstrate that the ability of the PC3 cells to stimulate osteoclast and osteoblast activities modified by TWIST suppression could be reversed by DKK-1 overexpression. The results suggest that DKK-1 may be a major downstream effector responsible for TWIST-modulated prostate cancer-mediated bone cell activity.


Figure 5
View larger version (64K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5. Overexpression of DKK-1 in siTWIST stable transfectants rescued their phenotype on prostate cancer-mediated bone cell activity. (a) Protein level of DKK-1 was restored in siTWIST transfectants, by transient transfection of DKK-1 expression plasmid, to its normal level in PC3-siCon. (b) Differentiated osteoclasts (indicated by red arrows) were observed when RAW264.7 cells were treated with conditioned media from PC3-siCon transfected with mock vector and PC3-siTWIST transfectants transiently expressed DKK-1. (c) Number of differentiated osteoclasts and (d) TRACP activity was similar when RAW264.7 cells were treated with conditioned media from PC3-siCon transfected with mock vector (dark gray column) and PC3-siTWIST transfectants transiently expressed DKK-1 (light gray column). (e) Calcium deposition of MC3T3-E1 cells was also similar when cells were treated with conditioned media from PC3-siCon transfected with mock vector (4 and 10) and PC3-siTWIST transfected with DKK-1 (5, 6, 11 and 12) as shown by alizarin red S (upper panel) and von Kossa (lower panel) assay. (f) ALP activity was also not significantly different between MC3T3-E1 cells treated with conditioned media from PC3-siCon transfected with mock vector (dark gray column) and PC3-siTWIST transfected with DKK-1 (light gray column). Error bars represent ±1 SD from at least three independent experiments. A relative ratio (to PC3-siCon) of the target band intensity adjusted by actin is shown below the results. An asterisk represents significant difference (P < 0.05) compared with PC3-siCon VC.

 
Suppression of TWIST reduced RUNX2 induction during osteogenic induction in 22Rv1
It has been suggested that prostate cancer cells acquire osteomimicry as they progress to develop bone metastasis (15). RUNX2, which promotes osteoblast differentiation, is upregulated in prostate cancer cells during osteogenic induction (17). In the present study, we investigated the role of TWIST in osteomimicry of an organ-confined prostate cancer cell line, 22Rv1. We found that after 12 days of osteogenic induction, the expression of TWIST together with RUNX2 was upregulated at both transcriptional (left panel, Figure 6a) and protein level (right panel) in 22Rv1 cells. Reports on the effect of TWIST on RUNX2 transcription were controversial, as inactivation of TWIST in osteoblasts has been correlated with reduced messenger RNA and protein levels of RUNX2 (26) yet another study has shown that TWIST does not transcriptionally regulate RUNX2 expression (27). In view of these conflicting reports, we went on to analyze the effect of altering TWIST expression on RUNX2 expression in 22Rv1 cells upon osteogenic induction. We found that TWIST and RUNX2 were both induced in 22Rv1 siLuc cells upon osteogenic induction (Figure 6b, lanes 1 and 2). However, silencing of TWIST in 22Rv1 cells by short hairpin RNA resulted in reduced induction of RUNX2 expression during osteogenic induction (lanes 3 and 4). These results suggest that TWIST expression may positively regulate the expression of RUNX2 during osteogenic induction in 22Rv1 cells. To further study the effect of TWIST overexpression on the transcriptional activity of RUNX2, we cotransfected a TWIST expression vector and pGL3-OG2, a luciferase reporter containing two OSE elements for the detection of RUNX2 transcriptional activity, into 293 cells. As shown in Figure 6c, the transcriptional activity of OSE elements present in the promoter of pGL3-OG2 was higher in cells transfected with TWIST expression vector (filled column). These results suggest that TWIST might promote osteomimicry in organ-confined prostate cancer cells through enhancement of transcriptional activity of RUNX2. However, further investigations are required to elucidate the significance of TWIST upregulation during osteogenic induction in terms of the survival and growth characteristic of the prostate cancer cells in bone.


Figure 6
View larger version (28K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 6. Osteogenic induction of 22Rv1, an organ-confined prostate cancer, upregulated TWIST. (a) 22Rv1 cells were cultured in an osteogenic medium, which contained 10 mM β-glycerophosphate and 50 µg/ml ascorbic acid for 10 days. TWIST expression was upregulated together with RUNX2 upon osteogenic induction in 22Rv1. (b) Reduced TWIST expression in 22Rv1, by short hairpin RNA technology, reduced TWIST and RUNX2 expression induced upon osteogenic induction. A relative (to 22Rv1 siLUC) ratio of the target band intensity adjusted by actin was displayed below the results. (c) By using pGL3-OG2, a RUNX2-responsive plasmid, we found that RUNX2 transcription activity was enhanced in 293 cells by cotransfection of a TWIST-expressing plasmid. Error bars represent ±1 SD from at least three independent experiments and asterisks represent significant difference (P < 0.05) compared with dark gray column. A relative (to 22Rv1 siLUC) of the target band intensity adjusted by actin is shown below the results.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Supplementary material
 References
 
Our previous study has shown that upregulation of TWIST in human prostate cancer specimens correlates with an increase in metastatic potential (34) while 95% of the metastatic prostate cancer in the patient cohort studied established their secondary lesion in bone. In this study, using PC3 with differential TWIST expression, firstly, we found that TWIST might promote prostate cancer to bone metastasis through enhancing prostate cancer cell-mediated osteolytic bone lesion, which has been suggested as an important initial phase of prostate cancer to bone metastasis (7,8). Secondly, we demonstrated that TWIST modulates PC3-mediated bone cell activity through transcriptional activation of Dkk-1. Finally, we showed that suppression of TWIST reduced the level of RUNX2 induction during osteogenic induction in an organ-confined prostate cancer cell line, 22Rv1. Taken together, our findings presented here provide evidence that inactivation of TWIST in prostate cancer cells might decrease the ability of prostate cancer cells to induce initial osteolytic phase during establishment of bone metastasis and to undergo osteomimicry, implying that TWIST inactivation might decrease the ability of prostate cancer cells to thrive in bone environment.

TWIST has been shown to regulate osteoblast differentiation (reviewed in ref. 29). However, these studies on TWIST and differentiation of bone cells are mostly about the intrinsic role of TWIST in bone cells. Our present study examined how TWIST expression modulates the paracrine action of prostate cancer cells on bone cell activity in vitro and thus affects bone remodeling. Recently, TWIST in perichondrium has been described to promote maturation of chondrocytes right next to this structure (42). In that study, TWIST has been shown to suppress the expression and secretion of fibroblast growth factor 18 in perichondrium during chondrogenesis in a RUNX2-dependent manner. Our present study demonstrated that TWIST in prostate cancer cells regulated the expression of a secretory factor, DKK-1, in order to modulate prostate cancer cell-mediated bone cell activity in a paracrine action. We further confirmed this through an immunoprecipitation experiment showing that DKK-1 protein level in the conditioned medium was decreased in PC3 cells expressing a lower level of TWIST (supplementary Figure 1 is available at Carcinogenesis Online).

Bone remodeling is a tightly controlled process including bone resorption by osteoclasts and bone formation by osteoblasts, and failure of which may cause severe bone disease (43). Bone resorption and bone formation are two coupled processes. When bone resorption increases, the osteoclasts recruited will increase bone resorption and activate osteoblasts, which in turn increase bone formation thereafter (44). Cancer cells metastasize to bone, disrupt bone remodeling and lead to an unbalanced bone gain or bone loss, thereby causing severe morbidity and high mortality. Our present study found that a high TWIST expression in prostate cancer cells was associated with enhanced cancer cell-mediated osteoclast differentiation and subdued cancer cell-mediated osteoblast mineralization. Cellular interactions occurring in cancer to bone metastasis are far more complex in vivo, as they involve concurrently cancer cells, osteoblasts and osteoclasts. Further investigation in vivo is required to elucidate the exact role of TWIST in prostate cancer to bone metastases.

TWIST has been recently demonstrated to alter Wnt signaling or itself acts as Wnt signaling-responsive gene in different cell contexts. Wnt1, which activates beta-catenin/TCF transcription activity, upregulates TWIST in mammary cells to inhibit lactogenic differentiation (45). Canonical Wnt signaling induced by Wnt3a in cultured chondrocytic cells and differentiating limb-bud mesenchyme activates TWIST expression to inhibit chondrogenesis (46). Overexpression of secreted Frizzled-related protein, another Wnt antagonist, decreases the expression of TWIST in prostate cancer cell line PC3 (47). WntD is a newly identified Wnt signaling member that is activated by TWIST in Drosophila embryos (48). In the present study, we showed for the first time that TWIST regulated the expression of DKK-1, a canonical Wnt signaling antagonist, probably at the transcriptional level. Our ChIP analysis showed that TWIST might bind to the promoter region of Dkk-1 probably through the E-box elements. However, it is not known whether this binding is a direct one or not. TWIST has been shown to form a complex with Smad4 and Histone deacetylase 1 before it can alter the expression of some other genes (49). Hence, it is possible that TWIST might also be recruited to the Dkk-1 promoter region through binding to other proteins.

Osteomimicry has been suggested to be a prerequisite for prostate cancer to survive and thrive in bone (15). The role played by TWIST in osteoblast differentiation is controversial and evidence suggests that it may be cell-type dependent (29). In the present study, we found that suppression of TWIST downregulated RUNX2, a potent osteoblast transcription factor, in an organ-confined prostate cancer cell line 22Rv1 during osteogenic induction. This suggests that TWIST might mediate prostate cancer osteomimicry. TWIST has also been shown to promote EMT (31,32). As prostate cancer cells are epithelial in origin, whereas osteoblasts are derived from mesenchymal cells, we speculate that EMT might be a prerequisite for prostate cancer cells to acquire osteomimicry and that TWIST would be a mediator of this sequence of events. TWIST has been described to upregulate RUNX2 in an osteoblast cell line (26) yet confer no effect on another (27). These data show that TWIST affects RUNX2 expression differently in different cell lines. It is probably that TWIST interacts with other proteins to modify RUNX2 expression.

In summary, TWIST has been shown previously to promote metastasis through EMT (31,32) and confer taxol resistant in prostate cancer (30,31). The present study identified other possible mechanisms of TWIST on prostate cancer progression via its ability to modulate prostate cancer cell-mediated bone cell activity and osteomimicry. It was shown that TWIST functioned upstream of DKK-1 to promote osteolytic metastasis in prostate cancer. More studies are required to provide further information on the role of TWIST in promoting prostate cancer to bone metastasis and the feasibility of therapeutic intervention through targeting at this pathway.


   Supplementary material
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
References
 
Supplementary Table I and Figure 1 can be found at http://carcin.oxfordjournals.org/


   Acknowledgments
 
We would like to thank Prof. F.Chan for prostate cancer cell line 22Rv1 and C42B, Prof. Gerard Karsenty for plasmid p147-Luc, Prof. Larry Kedes for plasmid pcDNA3.1 FLAG-tagged TWIST and Dr Colm Morrissey for his technical advice on in vitro osteoclast differentiation and osteoblast mineralization assay.

Conflict of Interest Statement: None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 References
 

  1. Jemal A, et al. Cancer statistics, 2007. CA Cancer J. Clin. (2007) 57:43–66.[Abstract/Free Full Text]
  2. Klotz L. Active surveillance for prostate cancer: for whom? J. Clin. Oncol. (2005) 23:8165–8169.[Abstract/Free Full Text]
  3. Denmeade SR, et al. Development of prostate cancer treatment: the good news. Prostate (2004) 58:211–224.[CrossRef][Web of Science][Medline]
  4. Saad F, et al. A randomized, placebo-controlled trial of zoledronic acid in patients with hormone-refractory metastatic prostate carcinoma. J. Natl Cancer Inst. (2002) 94:1458–1468.[Abstract/Free Full Text]
  5. Saad F, et al. Long-term efficacy of zoledronic acid for the prevention of skeletal complications in patients with metastatic hormone-refractory prostate cancer. J. Natl Cancer Inst. (2004) 96:879–882.[Abstract/Free Full Text]
  6. Loberg R, et al. A paradigm for the treatment of prostate cancer bone metastases based on an understanding of tumor cell-microenvironment interactions. J. Cell. Biochem. (2005) 96:439–446.[CrossRef][Web of Science][Medline]
  7. Keller ET, et al. Prostate cancer bone metastases promote both osteolytic and osteoblastic activity. J. Cell. Biochem. (2004) 91:718–729.[CrossRef][Web of Science][Medline]
  8. Hall CL, et al. Role of Wnts in prostate cancer bone metastases. J. Cell. Biochem. (2006) 97:661–672.[CrossRef][Web of Science][Medline]
  9. Guise TA, et al. Basic mechanisms responsible for osteolytic and osteoblastic bone metastases. Clin. Cancer Res. (2006) 12:6213s–6216s.[Abstract/Free Full Text]
  10. Chen G, et al. Up-regulation of Wnt-1 and beta-catenin production in patients with advanced metastatic prostate carcinoma: potential pathogenetic and prognostic implications. Cancer (2004) 101:1345–1356.[CrossRef][Web of Science][Medline]
  11. Schwaninger R, et al. Lack of noggin expression by cancer cells is a determinant of the osteoblast response in bone metastases. Am. J. Pathol. (2007) 170:160–175.[Abstract/Free Full Text]
  12. Hall CL, et al. Prostate cancer cells promote osteoblastic bone metastases through Wnts. Cancer Res. (2005) 65:7554–7560.[Abstract/Free Full Text]
  13. Liu XH, et al. Androgen-induced Wnt signaling in preosteoblasts promotes the growth of MDA-PCa-2b human prostate cancer cells. Cancer Res. (2007) 67:5747–5753.[Abstract/Free Full Text]
  14. Dorai T, et al. Therapeutic potential of curcumin in prostate cancer–V: interference with the osteomimetic properties of hormone refractory C4-2B prostate cancer cells. Prostate (2004) 60:1–17.[CrossRef][Web of Science][Medline]
  15. Koeneman KS, et al. Osteomimetic properties of prostate cancer cells: a hypothesis supporting the predilection of prostate cancer metastasis and growth in the bone environment. Prostate (1999) 39:246–261.[CrossRef][Web of Science][Medline]
  16. Knerr K, et al. Bone metastasis: osteoblasts affect growth and adhesion regulons in prostate tumor cells and provoke osteomimicry. Int. J. Cancer (2004) 111:152–159.[CrossRef][Web of Science][Medline]
  17. Zayzafoon M, et al. Notch signaling and ERK activation are important for the osteomimetic properties of prostate cancer bone metastatic cell lines. J. Biol. Chem. (2004) 279:3662–3670.[Abstract/Free Full Text]
  18. Huang WC, et al. Human osteocalcin and bone sialoprotein mediating osteomimicry of prostate cancer cells: role of cAMP-dependent protein kinase A signaling pathway. Cancer Res. (2005) 65:2303–2313.[Abstract/Free Full Text]
  19. Barnes GL, et al. Osteoblast-related transcription factors Runx2 (Cbfa1/AML3) and MSX2 mediate the expression of bone sialoprotein in human metastatic breast cancer cells. Cancer Res. (2003) 63:2631–2637.[Abstract/Free Full Text]
  20. Barnes GL, et al. Fidelity of Runx2 activity in breast cancer cells is required for the generation of metastases-associated osteolytic disease. Cancer Res. (2004) 64:4506–4513.[Abstract/Free Full Text]
  21. Javed A, et al. Impaired intranuclear trafficking of Runx2 (AML3/CBFA1) transcription factors in breast cancer cells inhibits osteolysis in vivo. Proc. Natl Acad. Sci. USA (2005) 102:1454–1459.[Abstract/Free Full Text]
  22. Brubaker KD, et al. Prostate cancer expression of runt-domain transcription factor Runx2, a key regulator of osteoblast differentiation and function. Prostate (2003) 56:13–22.[CrossRef][Web of Science][Medline]
  23. Murray SS, et al. Expression of helix-loop-helix regulatory genes during differentiation of mouse osteoblastic cells. J. Bone Miner. Res. (1992) 7:1131–1138.[Web of Science][Medline]
  24. Rice DP, et al. Integration of FGF and TWIST in calvarial bone and suture development. Development (2000) 127:1845–1855.[Abstract]
  25. Lee MS, et al. TWIST, a basic helix-loop-helix transcription factor, can regulate the human osteogenic lineage. J. Cell. Biochem. (1999) 75:566–577.[CrossRef][Web of Science][Medline]
  26. Yousfi M, et al. TWIST inactivation reduces CBFA1/RUNX2 expression and DNA binding to the osteocalcin promoter in osteoblasts. Biochem. Biophys. Res. Commun. (2002) 297:641–644.[CrossRef][Web of Science][Medline]
  27. Bialek P, et al. A twist code determines the onset of osteoblast differentiation. Dev. Cell (2004) 6:423.[CrossRef][Web of Science][Medline]
  28. Ishii M, et al. Msx2 and Twist cooperatively control the development of the neural crest-derived skeletogenic mesenchyme of the murine skull vault. Development (2003) 130:6131–6142.[Abstract/Free Full Text]
  29. Komori T. Regulation of osteoblast differentiation by transcription factors. J. Cell. Biochem. (2006) 99:1233–1239.[CrossRef][Web of Science][Medline]
  30. Wang X, et al. Identification of a novel function of TWIST, a bHLH protein, in the development of acquired taxol resistance in human cancer cells. Oncogene (2004) 23:474–482.[CrossRef][Web of Science][Medline]
  31. Kwok WK, et al. Up-regulation of TWIST in prostate cancer and its implication as a therapeutic target. Cancer Res. (2005) 65:5153–5162.[Abstract/Free Full Text]
  32. Yang J, et al. Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell (2004) 117:927–939.[CrossRef][Web of Science][Medline]
  33. Ma L, et al. Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature (2007) 449:682–688.[CrossRef][Web of Science][Medline]
  34. Yuen HF, et al. Significance of TWIST and E-cadherin expression in the metastatic progression of prostatic cancer. Histopathology (2007) 50:648–658.[CrossRef][Web of Science][Medline]
  35. Valsesia-Wittmann S, et al. Oncogenic cooperation between H-Twist and N-Myc overrides failsafe programs in cancer cells. Cancer Cell (2004) 6:625–630.[CrossRef][Web of Science][Medline]
  36. Ling MT, et al. Overexpression of Id-1 in prostate cancer cells promotes angiogenesis through the activation of vascular endothelial growth factor (VEGF). Carcinogenesis (2005) 26:1668–1676.[Abstract/Free Full Text]
  37. Dai J, et al. Bone morphogenetic protein-6 promotes osteoblastic prostate cancer bone metastases through a dual mechanism. Cancer Res. (2005) 65:8274–8285.[Abstract/Free Full Text]
  38. Corey E, et al. Osteoprotegerin in prostate cancer bone metastasis. Cancer Res. (2005) 65:1710–1718.[Abstract/Free Full Text]
  39. Sramkoski R, et al. A new human prostate carcinoma cell line, 22Rv1. In Vitro Cell. Dev. Biol. Anim. (1999) 35:403–409.[Web of Science][Medline]
  40. Zeitlinger J, et al. Whole-genome ChIP-chip analysis of Dorsal, Twist, and Snail suggests integration of diverse patterning processes in the Drosophila embryo. Genes Dev. (2007) 21:385–390.[Abstract/Free Full Text]
  41. Chamorro M, et al. FGF-20 and DKK1 are transcriptional targets of beta-catenin and FGF-20 is implicated in cancer and development. EMBO J. (2005) 24:73–84.[CrossRef][Web of Science][Medline]
  42. Hinoi E, et al. Runx2 inhibits chondrocyte proliferation and hypertrophy through its expression in the perichondrium. Genes Dev. (2006) 20:2937–2947.[Abstract/Free Full Text]
  43. Robling AG, et al. Biomechanical and molecular regulation of bone remodeling. Annu. Rev. Biomed. Eng. (2006) 8:455–498.[CrossRef][Web of Science][Medline]
  44. Hadjidakis DJ, et al. Bone remodeling. Ann. N. Y. Acad. Sci. (2006) 1092:385–396.[CrossRef][Web of Science][Medline]
  45. Howe LR, et al. Twist is up-regulated in response to Wnt1 and inhibits mouse mammary cell differentiation. Cancer Res. (2003) 63:1906–1913.[Abstract/Free Full Text]
  46. Reinhold MI, et al. The Wnt-inducible transcription factor Twist1 inhibits chondrogenesis. J. Biol. Chem. (2006) 281:1381–1388.[Abstract/Free Full Text]
  47. Zi X, et al. Expression of Frzb/secreted Frizzled-related protein 3, a secreted Wnt antagonist, in human androgen-independent prostate cancer PC-3 cells suppresses tumor growth and cellular invasiveness. Cancer Res. (2005) 65:9762–9770.[Abstract/Free Full Text]
  48. Ganguly A, et al. Drosophila WntD is a target and an inhibitor of the Dorsal/Twist/Snail network in the gastrulating embryo. Development (2005) 132:3419–3429.[Abstract/Free Full Text]
  49. Hayashi M, et al. Comparative roles of Twist-1 and Id1 in transcriptional regulation by BMP signaling. J. Cell Sci. (2007) 120:1350–1357.[Abstract/Free Full Text]
Received December 20, 2007; revised March 26, 2008; accepted April 22, 2008.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
29/8/1509    most recent
bgn105v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Yuen, H.-F.
Right arrow Articles by Chan, K.-W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yuen, H.-F.
Right arrow Articles by Chan, K.-W.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?