Metastatic and non-metastatic colorectal cancer (CRC) cells induce host metalloproteinase production in vivo.

Numerous studies have demonstrated the persistent localization of matrix metalloproteinase (MMP) expression to the interface between invading human colorectal cancer (CRC) cells and surrounding stroma supporting a role for MMPs in CRC invasion and metastasis. The present study sought to determine whether CRC cells of varying metastatic potential would have differential effects on host MMP release. Subcutaneous CRC tumors were generated in BALB/c nude mice using three CRC cell lines: SW480, SW620, and the highly metastatic SW620S5 clone. Representative samples from the subcutaneous CRC were then orthotopically implanted on the cecum of recipient nude mice. Subcutaneous and cecal tumors were analyzed for MMP expression via zymography, western blot, and RT-PCR. In vitro, none of the three cell lines expressed MMP-2 nor MMP-9. In contradistinction, the subcutaneous tumors expressed limited amounts of MMP-2 and MMP-9 while the cecal tumors expressed significant amounts of MMP-2 and MMP-9 as well as other smaller members of the MMP family. MMP-9 mRNA and protein was confirmed as host in origin by RT-PCR with mouse specific primers and a mouse MMP-9 molecular weight of 105 kDa as determined by zymography and western blot analysis. In situ hybridization also localized the mRNA for MMP-9 to the host stromal cells. In conclusion, CRC cells appear incapable of producing MMP-2 and MMP-9 in vitro but are capable of up-regulating host MMP production in vivo. Enhanced host MMP-9 production in metastatic CRC cell-derived subcutaneous and cecal tumors suggests that metastatic colon cells may acquire the expression of important MMP regulating factor(s) in vivo.

Bone extracellular matrix induces homeobox proteins independent of androgens: possible mechanism for androgen-independent growth in human prostate cancer cells.

BACKGROUND: Differences in gene expression in prostate cells are believed to be secondary to epithelial-stromal interactions. We theorized that bone matrix may provide a fertile "soil" for prostate cancer by inducing androgen-dependent genes and allowing for androgen-independent growth. METHODS: Human prostate cancer cells (LNCaP) were grown under different conditions and analyzed for differential expression of mRNA. LNCaP cells were grown in the presence of 10 nM dihydrotestosterone (DHT), on extracellular matrix (ECM) derived from bone cells (without exogenous DHT), and on plastic culture dishes without exogenous DHT. A differential display of mRNA produced by LNCaP cells grown in the above conditions was then analyzed. RESULTS: Multiple unique transcripts were present in cells that were grown in the presence of DHT and on bone ECM (without exogenous DHT), but not on plastic culture dishes without exogenous DHT. Nine of these transcripts were then cloned and analyzed. Many (5/9) of these transcripts were found to contain multiple ATTA motifs in their corresponding 3'-untranslated regions. ATTA motifs have been shown to be homeobox protein-binding sites. Homeobox proteins and their target genes are thought to regulate cellular differentiation. Consistent with this, we demonstrated by reverse transcription polymerase chain reaction (PCR) that homeobox genes were differentially expressed in LNCaP cells when the cells were grown in the presence of DHT and on bone ECM (without exogenous DHT), but not on plastic culture dishes without exogenous DHT. Furthermore, we assayed LNCaP/fetal fibroblast chimeric tumors (n = 8) that were grown in male nude mice. Some of these tumors continued to grow in these mice despite treatment with surgical castration. In blinded studies, we were able to determine which tumor samples were androgen independent by their expression of homeobox genes. All samples that were androgen independent (n = 4) expressed the homeobox genes. Finally, gel retardation assay demonstrated that the homeobox proteins were able to bind to our cloned DNA sequences. Furthermore, footprinting analysis showed that the homeobox proteins bound to the ATTA motif in the 3'-region of our target DNA. CONCLUSIONS: Bone ECM, in the absence of DHT, has the ability to regulate androgen-responsive genes. Furthermore, many of these genes contain homeobox binding sites and the expression of homeobox genes may itself be regulated by bone ECM. If so, this may partially explain the clinical observation that bone provides a fertile "soil" for prostate cancer growth and metastasis.

Role of prostatic basal cells in the regulation and suppression of human prostate cancer cells.

Cytokeratin expression in normal and malignant prostatic tissue indicates a loss of basal epithelial cells in cancer. We investigated the ability of basal-like prostatic epithelial cells to inhibit the growth of prostatic cancer cells. Human prostate LNCaP cells were grown in medium with or without 10 nM dihydrotestosterone (DHT) on plastic culture dishes or on extracellular matrix derived from basal-like epithelial cells (primary cultures derived from normal peripheral zone of the prostate) that were grown with or without 10 nM DHT. Colorimetric 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assays were used to assess the growth of LNCaP cells. On plastic dishes, growth of LNCaP cells was increased 5-10% by the presence of DHT in the medium. On matrix derived from basal-like cells that were grown in the absence of DHT, growth of LNCaP cells with or without DHT was similar to that on plastic. However, on matrix derived from basal-like cells that were grown with DHT, growth of LNCaP cells was suppressed when compared to all other culture conditions (P < 0.01). To determine whether basal-like cells could alter the function of LNCaP cells, we measured prostate-specific antigen (PSA) mRNA expression with the use of comparative RT-PCR. We found a significant decrease in the mature PSA transcript in cells grown on matrix derived from basal-like cells that were grown with DHT. The expression of PSA transcript was not altered in LNCaP cells that were grown on matrix derived from basal-like cells that were grown in the absence of DHT. Furthermore, using differential display of mRNA, we demonstrated that there were induction and suppression of multiple unique transcripts in the LNCaP cells when grown on the various culture conditions. To determine a possible mechanism for these observations. We used a dot blot immunoassay for several known inhibitory factors. We determined that DHT can induce the basal-like cells to secrete transforming growth factor-beta (TGF-beta 1), and that TGF-beta 1 can inhibit the proliferation of LNCaP cells in a dose dependent manner. We conclude that basal-like epithelial cells, in the presence of DHT, secrete an extracellular matrix o matrix associated factor(s), e.g. TGF-beta 1, that suppresses proliferation and function of prostate cancer cells. Our data suggest that the disappearance of the basal cell layer may be a prerequisite for the progression of prostatic neoplasia.

Tumor necrosis factor mediates apoptosis via Ca++/Mg++ dependent endonuclease with protein kinase C as a possible mechanism for cytokine resistance in human renal carcinoma cells.

PURPOSE: Because renal cell cancers have been found to be resistant to numerous chemotherapeutic agents, other agents including tumor necrosis factor are now being considered for clinical use. In this study, we used 2 renal cancer cell lines. SK-RC-42 and SK-RC-49, and determined the cytotoxic effects of tumor necrosis factor (TNF) and the possible mechanism of TNF resistance. MATERIALS AND METHODS: Cytotoxic assays, comparative reverse transcriptase-polymerase chain reaction (RT-PCR) and nuclease digestion analyses were used. RESULTS: Cytotoxic assays with SK-RC-42 demonstrated that TNF at 50 ng./ml. for 24 hours resulted in 19% cytotoxicity of the cells. Similar assay with SK-RC-49 only demonstrated less than 4% cytotoxicity. Based on these results, we defined our TNF-sensitive cells as SK-RC-42 and SK-RC-49 as our TNF-resistant cells. To determine whether protein kinase C (PKC), which is involved in the signal transduction pathway of a cell, could regulate endogenous basal TNF mRNA levels, comparative PCR analyses for TNF expression were used. The PCR results demonstrated that the TNF-resistant cell, SK-RC-49, had a higher basal expression of TNF mRNA than the TNF-sensitive cell SK-RC-42. With PMA, a PKC activator, for various time points, both cell lines demonstrated an induction of endogenous TNF mRNA. To further confirm our findings that PKC may regulate endogenous TNF expression, a PKC inhibitor, staurosporine, was used. When the cells were treated with staurosporine prior to PMA stimulation, no increase in TNF mRNA expression was seen. To determine whether PKC is involved in providing resistance against TNF, we incubated the SK-RC-42 and SK-RC-49 cells with staurosporine at 100 nM. and TNF at 50 ng./ml. for 24 hours. After factoring out the cytotoxicity of staurosporine, the TNF-mediated cytotoxicity increased to 39.3% and 28.7% for the SK-RC-42 and SK-RC-49 cells, respectively. Furthermore, by using a nuclease digestion assay, we demonstrated that TNF activated a Ca++/Mg++ dependent endonuclease responsible for programmed cell death or apoptosis. CONCLUSIONS: Our data suggest that protein kinase C may play a role in protecting renal cancer cells from undergoing cytokine-mediated apoptosis.

Specific sequences of fibronectin activate the protein kinase C signal transduction pathway in invasive bladder cancer.

The mechanism of human bladder cancer cell invasion is not clear, but it appears that extracellular matrix components, such as fibronectin, may be involved. To investigate the role of fibronectin in tumor cell invasion and progression, we used an in vitro invasion assay to define the motility stimulating fragment of fibronectin for invasive human bladder cancer T24 cells. Using a modified Boyden chamber assay and purified fragments of fibronectin, we demonstrated that both the 120 kDa chymotrypsin generated fragment of fibronectin (containing the cell attachment RGD motif and additional sequences towards the carboxyl-terminal heparin binding domain), as well as the trypsin generated 60 kDa fragment of fibronectin (containing the carboxyl-terminal heparin binding domain and additional sequences towards the cell attachment RGD motif), were able to stimulate the migration of invasive human bladder cancer T24 cells. Control fragments containing only the amino-terminal gelatin binding region of fibronectin did not stimulate the motility of the human bladder cancer T24 cells. To determine the molecular mechanism in which these fragments may stimulate the migration of the T24 cells, we assayed for intracellular signal transduction pathway protein kinase C (PKC). We demonstrated that both the 120 kDa and the 60 kDa fragments were able to stimulate the activation of protein kinase C. Non-motility stimulating fragments of fibronectin were not able to activate protein kinase C. We conclude that the PKC signal transduction pathway may be involved in matrix mediated motility, and suggest that the inhibition of such pathway(s) may alter the malignant phenotype of human bladder cancer.

Ha-ras oncogene expression abrogates a pH dependent endonuclease activity of apoptosis in normal rat kidney cells.

To investigate the role of oncogene expression in the resistance to tumor necrosis factor-alpha (TNF), we transfected the mutated T24-Ha-ras oncogene into the murine kidney cell line NRK and an alternative murine cell line C127 cells. The resulting transfectants, NRK-Ha and HC127, were assayed for TNF mediated cytotoxicity. Cellular cytotoxicity of 45% over 48 h occurred with the NRK cells. However, ras transfectant NRK-Ha cells demonstrated 0% cytotoxicity over the same period. Both C127 cells and the ras transfectant HC127 demonstrated 40% and 25% cytotoxicity, respectively, over 48 h when incubated with TNF. Furthermore, DNA isolated from NRK, C127, HC127, but not NRK-Ha cells revealed the presence of DNA fragmentation 'ladders' indicative of successful apoptosis when the cells were incubated with TNF. To determine the possible mechanism in which the ras oncogene may have protected the NRK-Ha cells from TNF mediated cytotoxicity and apoptosis, total nuclear endonucleases from the NRK cells and the ras transfectant NRK-Ha cells were isolated. We determined that the endonuclease activity in the NRK and the ras transfectant NRK-Ha cells was a pH dependent endonuclease. Significant degradation of the target DNA was observed only in pH 4-6 buffers containing the endonuclease. Furthermore, preliminary intracellular pH analysis suggested that while the NRK cells have an intracellular pH of 6.0, the ras transfectant NRK-Ha cells have an intracellular pH of 7.2 and may have abrogated its pH dependent endonuclease. Both the C127 cells and the ras transfectant HC127 cells did not express a pH dependent endonuclease but rather a Ca2+/Mg2+ dependent endonuclease. Furthermore, preliminary intracellular pH analysis suggested that both the C127 and HC127 cells have the same intracellular pH. Our results indicate that in normal rat kidney cells, ras oncogene transfection may cause a disruption in the endonuclease activation involved in apoptosis.

LAK cell mediated apoptosis of human bladder cancer cells involves a pH-dependent endonuclease system in the cancer cell: possible mechanism of BCG therapy.

Intravesical bacillus Calmette-Guerin (BCG) is an effective treatment for superficial bladder cancer. However, its mechanism has been only partially elucidated. We studied whether LAK cell killing of human bladder cancer cells occurs via apoptosis (programmed cell death) or necrosis. Fluorescent dye labeled T24 cells were observed to undergo morphologic changes associated with apoptosis in the presence of LAK cells when analyzed under a fluorescence microscope. Furthermore, analysis of the DNA isolated from the cytotoxic assay confirmed that the LAK cell induced death of the T24 cells occurred via apoptosis. By pretreating the LAK cells with antifibronectin antibodies, we were able to significantly inhibit the LAK cell killing of the T24 cells. The percentage of cytotoxicity was reduced from 50% to 13% (p = 0.001), and the apoptotic pattern seen on agarose gel electrophoresis was significantly diminished. There was no significant change in the viability of the LAK cells following treatment with the antibodies. Endonuclease isolation from human bladder cancer T24 cells demonstrated that these cells express a pH-dependent and not a Ca++/Mg++ dependent endonuclease. Significant degradation of a target DNA was observed only in pH 4 to pH 5.6 buffers containing endonuclease from T24 cells and not in pH 6 to pH 8 buffers containing endonuclease from T24 cells. The presence or absence of Ca++/Mg++ in the various pH buffers did not alter the endonuclease activity. Finally, we demonstrated that death of T24 cells can be induced by altering the intracellular pH of the cells to 5.6 or lower with the proton ionophore nigericin. We conclude that LAK cells induce T24 cells to undergo apoptosis and that this process involves the fibronectin molecule present on the LAK cell membrane. Furthermore, the cleavage postulate that, in vivo, LAK cells activated by IL-2 produced by BCG activated CD4+ cells may induce bladder cancer cells to undergo apoptosis. This may partially explain the mechanism whereby BCG achieves its therapeutic effect.

Bacillus Calmette-Guérin interacts with the carboxyl-terminal heparin binding domain of fibronectin: implications for BCG-mediated antitumor activity.

Intravesical bacillus Calmette-Guérin (BCG) has been shown to be an effective treatment for superficial transitional cell carcinoma of the bladder. The mechanisms by which BCG achieves this effect remain unclear. Reports have attributed an important role to fibronectin both in the initial attachment of BCG to bladder surfaces and in the limitation of tumor cell motility. In the present study, using limited protease cathepsin B degradation followed by Western blot analyses with antibodies to various domains of the fibronectin molecule, we showed that BCG appears to bind to fibronectin near the carboxyl terminal and adjacent to the heparin binding domain. Furthermore a 51-chromium release assay with human bladder cancer cell line T24 as target cells and lymphokine activated killer (LAK) cells as effector cells showed that fibronectin was needed for tumor cytotoxicity by the LAK cells. By using antibodies and peptides to various domains of the fibronectin molecule, the heparin binding domain, but not the cell binding domain, carboxyl terminal region, or the amino terminal region of the fibronectin molecule, was identified as essential to tumor cell lysis by the LAK cells. Flow cytometric analysis showed that both peripheral blood lymphocytes and the LAK cells express fibronectin receptors VLA-3, VLA-4 and VLA-5 on their surfaces. However, the numbers of receptors are not significantly different in the two cell populations. We conclude that, by binding near the carboxyl terminal region and adjacent to the heparin-binding domain of the fibronectin molecule, BCG may protect this region of the molecule from tumor proteases, and may thus allow the antitumor activity of the host immune cells to take place.

Altered extracellular matrices influence cellular processes and nuclear matrix organizations of overlying human bladder urothelial cells.

A central issue in tumor biology is the understanding of the interactions between tumor cells and their environment. Using normal and ras oncogene transfected rat fibroblast cells, we now demonstrate that the transfected cells make altered extracellular matrices (ECM) and that their resulting ECM influence the proliferation and genetic regulation of human bladder cancer EJ cells. Using Western blot analyses, we observed that the ras transfected fibroblast cells lacked the ability to produce extracellular matrix component laminin whereas the normal parental fibroblast cells were able to produce intact laminin. Both transfected and nontransfected fibroblast cells were able to synthesize other extracellular matrix molecules such as type IV collagen and fibronectin. Human bladder tumor EJ cells were grown on ECM derived from normal and transfected rat fibroblast cells, and the proliferation rate and type IV collagen mRNA expression of EJ cells were determined. We observed that EJ cells, when grown on ECM derived from the ras transfected fibroblast cells, had a higher growth rate than when grown on ECM derived from the normal fibroblast cells (P < 0.037). Furthermore, EJ cells grown on ECM derived from transfected fibroblast cells showed up-regulation of type IV collagen mRNA expression when compared with EJ cells grown on ECM derived from nontransfected fibroblast cells. Finally EJ cells grown on purified laminin but not on collagen IV coated flasks showed the same level of type IV collagen mRNA expression as when grown on ECM derived from nontransfected parental fibroblast cells. Haptotactic/motility assays with EJ cells and ECM derived from ras transfected and nontransfected fibroblast cells demonstrated that ECM of ras transfected fibroblast cells, but not the parental fibroblast cells, provided a permissive or fertile soil for EJ tumor cell invasion. Finally, two-dimensional gel electrophoresis of 35S-labeled nuclear matrix proteins of EJ cells cultured on ECM derived from ras transfected fibroblast cells revealed expression of proteins in the molecular weight range of M(r) 35,000-45,000 and isoelectric focusing pH range of 5.5 to 6.0. These proteins were not present in EJ cells cultured on ECM derived from parental nontransfected fibroblast cells. We conclude that extracellular matrices derived from transformed stroma producing cells may influence the proliferation, genetic regulation, and maintenance of the overlying urothelial tumor cells. The mechanism by which the ECM may influence cellular behavior and phenotype may be in their ability to modulate the nuclear matrix proteins of the overlying cell.

Chemical evidence for the existence of activated G-actin.

Globular actin (G-actin) will polymerize to form filamentous actin (F-actin) under physiological ionic conditions, and is known to be regulated by univalent and bivalent cations, such as K+ and Mg2+. The current concept of this process involves four steps: activation, nucleation, elongation and annealing. Evidence for the existence of activated G-protein has been suggested by changes in the resistance to proteolysis [Rich & Estes (1976) J. Mol. Biol. 104, 777-792] and u.v.-light absorption [Rouayrenc & Travers (1981) Eur. J. Biochem. 116, 73-77]. More recently we [Liu et al. (1990) Biochem. J. 266, 453-459] have provided direct chemical evidence for extensive conformational changes during the transformation of G-actin into F-actin. In this study we now present direct chemical evidence for the existence of a short-lived species, an activated form of G-actin, which can be detected by changes in the accessibility of the free thiol groups on the G-actin molecule when modified by a specific thiol-group-targeted reagent, 7-dimethylamino-4-methyl-3-N-maleimidylcoumarin (DACM). The presence of K+ and/or Mg2+ ions caused a large increase in the accessibility of the thiol groups of Cys-217 and Cys-374, but not those of Cys-10 and Cys-257. Mg2+ effected relatively faster changes than did K+ ions. The results suggest that the function of these ions is to convert G-actin into an activated form, and further suggest that the change in conformation is mainly confined to the large domain. Such changes at least involve certain portions of the G-actin molecule that contain Cys-217 and Cys-374. On the other hand, little or no significant change could be observed in the small domain of G-actin as reflected by the accessibility of Cys-10. The bound nucleotide remained as ATP during the activation of G-actin and was hydrolysed to ADP on polymerization. The activated G-actin had a life-time of about 8 min or less depending on the concentration of G-actin. At higher protein concentration, its life-time was much shorter, probably owing to the earlier onset of polymerization, which apparently is governed by the concentration of the activated form. The life-time of this new species can be extended by lowering the temperature and is less affected by actin concentration. This new species is considered to be an activated form of G-actin, since polymerization renders all the thiol groups on actin inaccessible to the reagent DACM.

Surgical treatment of primary tumors of the sacrum.

Fifty-four patients were surgically treated for primary tumors of the sacrum. The ratio of men to women was 2:1, and most patients were from 30-50 years of age. Nearly half the tumors were chordomas, and the others were mostly giant cell tumors and neurofibromas. Proven malignancy was present in less than 10%. The operative and postoperative mortality was 11%, wound infection 11%, delayed wound healing 11%, and the immediate success rate 89%. A follow-up study of 33 cases for more than two years showed that 25 patients were living and well, providing a survival rate of 75.7%. Except for malignant tumors, preservation of upper sacral nerve roots was possible and necessary in the majority of low grade and benign tumors. Ligation of both internal iliac arteries and a temporary block of the common iliac arteries or aorta minimized bleeding. Injury to veins must be carefully avoided. Postoperative X-ray treatment appeared helpful in reducing recurrences, especially in upper sacral tumors where nerve root compression had to be relieved.

Giant-cell tumor of bone: analysis of two hundred and eight cases in Chinese patients.

Giant-cell tumor of bone seems to occur more frequently in Chinese people than in those residing in Western countries. The estimated incidence is about 20 per cent of all primary tumors of bone. Of 208 surgically treated and pathologically proved giant-cell tumors, 194 were benign. We excluded patients with primary or secondary amputation unrelated to recurrence and those followed for less than two years or lost to follow-up. Of the remaining 111 patients who were followed for more than two years, twenty-nine had a recurrence, giving a recurrence rate of 26.1 per cent. The rate of recurrence was highest following curettage and bone-grafting (41.2 per cent) and was much lower in patients who were treated by resection and fusion (7.1 per cent). Since resection of this tumor with reconstructive procedures, either by massive homogenous bone-grafting or artificial joint replacement, is complicated and might cripple the patient if it fails, we propose excision and curettage with bone-grafting as the most suitable method of treatment in the majority of patients with giant-cell tumor of bone.