Genetic factors and suppression of metastatic ability of prostatic cancer

Progression of prostatic cancer from nonmetastatic to high metastatic ability may involve the loss of a metastasis suppressor gene. To test this possibility, nonmetastatic and highly metastatic Dunning rat prostatic cancer cells were fused. Hybrid clones were isolated which conserved the chromosomes from their parental cells. When these hybrids were injected into animals, none developed distant metastases. When these nonmetastatic primary tumors were passaged in vivo, occasional animals developed distant metastases. Cytogenetic analysis of eight of these metastatic revertants demonstrated a consistent loss of a copy of a normal chromosome 2. Although previous studies have demonstrated that specific chromosomes can inhibit tumorigenicity in cell fusion experiments, this is the first study to show that prostatic cancer metastasis is associated with the loss of a specific chromosome. Furthermore, these studies suggest that a metastasis suppressor gene for rat prostatic cancer is located on chromosome 2.     

Metastatic phenotype of mouse-human melanoma cell hybrids is associated with the presence of chromosome 17 from highly metastatic human melanoma cells

We have previously shown that in interspecific mouse-human melanoma cell hybrids obtained by fusion of nonmetastatic mouse with metastatic human melanoma cells, the metastatic phenotype predominates. The purpose of this study was to identify human chromosome(s) which could be responsible for conferring metastatic potential upon nonmetastatic mouse melanoma cells and therefore harbor the genes important for the metastatic properties of human melanoma cells. Seven mouse-human melanoma hybrids were examined; five were derived from the fusion of nonmetastatic (C19) and metastatic (C3) mouse K-1735 melanoma clones with highly metastatic clone (C15) of human melanoma A375 and the two others had as a human partner a nonmetastatic clone (Cls) of the A375 melanoma. The hybrids were examined during segregation of human chromosomes in vitro and in vivo for metastatic properties in nude mice and for the retaining human chromosomes. In the hybrid H7, which demonstrated the highest metastatic potential, the presence of human chromosomes was studied by GTG-banding and by fluorescence in situ hybridization (FISH) analysis. In the other hybrids, only FISH detection of human chromosomes was applied. All hybrids derived from nonmetastatic mouse and metastatic human melanoma cells demonstrated metastatic properties from early passages, when they contained the majority of the human chromosomes. Their metastatic phenotype remained stable during further segregation of most of the human chromosomes except for 17. Chromosome 17 was retained most consistently in all examined hybrids. However, the metastatic phenotype of the hybrids was associated only with the presence of chromosome 17 from the metastatic human donor cells. This chromosome was also found in almost 100% of cells recovered from lung metastases derived from the hybrid cells. In one lung metastasis developed from the H7 hybrid, chromosome 17 was detected as the sole human chromosome and these cells preserved the acquired high metastatic properties. Based on these results we conclude that human chromosome 17 from metastatic melanoma cells (A375 C15), when functional in the mouse genetic background, can be sufficient to render the recipient nonmetastatic mouse cells to a fully malignant phenotype. Additional data suggest that this ability might be related to the expression of the mutated human p53 gene.     

Functional evidence for a metastasis suppressor gene for rat prostate cancer within a 60-kilobase region on human chromosome 8p21-p12.

We recently demonstrated that the human chromosome 8p21-p12 region encodes a metastasis suppressor gene for rat prostate cancer. The presence of this region suppresses the metastatic ability of rat prostate cancer cells (N. Nihei et al., Genes Chromosomes Cancer, 17: 260-268, 1996). To define further the region harboring the metastasis suppressor gene, a truncated human chromosome 8 containing this region was transferred into highly metastatic AT6.3 rat prostate cancer cells by microcell-mediated chromosome transfer. The region of human chromosome 8 retained in each microcell hybrid was determined by a PCR analysis with sequence-tagged site markers, and this analysis placed the metastasis suppressor gene in the interval between D8S2249 and D8S2244 on human chromosome 8p21-p12. One of the metastasis-suppressed microcell hybrids was used for construction of representative yeast/bacterial artificial chromosome (YAC/BAC) library covering the candidate region using a transformation-associated recombination technology (N. Kouprina et al., Genomics, 53: 21-28, 1998). The final contig corresponding to the candidate region was assembled by four YAC/BAC clones. Each clone was transfected into the AT6.3 cells, and the resultant transfectants were tested for their metastatic ability in athymic nude mice. Introduction of a 60-kb YAC/BAC clone resulted in significant suppression of the metastatic ability without suppression of the tumorigenicity. In contrast, other YAC/BAC clones in the contig had neither metastasis nor tumor suppressor activity. This demonstrates that the 60-kb fragment from the human chromosome 8p21-p12 region contains the metastasis suppressor gene for the AT6.3 cells. Frequent loss of heterozygosity of this region is detected in human prostate cancer, which suggests that our target metastasis suppressor gene may also play an important role in the progression of human prostate cancer.     

A molecular role for lysyl oxidase in breast cancer invasion

We identified previously an up-regulation in lysyl oxidase (LOX) expression,an extracellular matrix remodeling enzyme, in a highly invasive/metastatic human breast cancer cell line, MDA-MB-231, compared with MCF-7, a poorly invasive/nonmetastatic breast cancer cell line. In this study, we demonstrate that the mRNA expression of LOX and other LOX family members [lysyl oxidase-like (LOXL), LOXL2, LOXL3, and LOXL4] was observed only in breast cancer cells with a highly invasive/metastatic phenotype but not in poorly invasive/nonmetastatic breast cancer cells. LOX and LOXL2 showed the strongest association with invasive potential in both highly invasive/metastatic breast cancer cell lines tested (MDA-MB-231 and Hs578T). To determine whether LOX is directly involved in breast cancer invasion, LOX antisense oligonucleotides were transfected into MDA-MB-231 and Hs578T cells, and found to inhibit invasion through a collagen IV/laminin/gelatin matrix in vitro compared with LOX sense oligonucleotide-treated and untreated controls. In addition, treatment of MDA-MB-231 and Hs578T cells with beta-aminopropionitrile (an irreversible inhibitor of LOX enzymatic activity) decreased invasive activity. Conversely, MCF-7 cells transfected with the murine LOX gene demonstrated a 2-fold increase in invasiveness that was reversible by the addition of beta-aminopropionitrile in a dose-dependent manner. In addition, endogenous LOX mRNA expression was induced when MCF-7 cells were cultured in the presence of fibroblast conditioned medium or conditioned matrix, suggesting a role for stromal fibroblasts in LOX regulation in breast cancer cells. Moreover, the correlation of LOX up-regulation and invasive/metastatic potential was additionally demonstrated in rat prostatic tumor cell lines, and human cutaneous and uveal melanoma cell lines. These results provide substantial new evidence that LOX is involved in cancer cell invasion.     

Decreased expression of E-cadherin in the progression of rat prostatic cancer.

Cadherins represent a family of Ca(2+)-dependent cell adhesion molecules involved in homotypic, homophilic cell-cell interactions. Recent studies have shown that the cadherins can play a role in invasive and metastatic behavior. Using the established Dunning R-3327 model system of serially transplantable rat prostate cancers, the expression of E- and P-cadherin in rat prostatic cancer was studied. Analysis within this system demonstrated that whereas E-cadherin was expressed in the normal rat prostate and the well- or moderately differentiated, noninvasive Dunning tumors, no expression, either at the mRNA or at the protein level, could be detected in the invasive sublines. Since not all invasive Dunning tumors studied have metastatic ability, these results suggest that a decreased expression of E-cadherin is correlated with invasive behavior rather than with metastatic ability. Recently, genetic instability occurred in an animal bearing the well differentiated, androgen-responsive, slow growing, nonmetastatic Dunning R-3327-H rat prostate cancer resulting in the progression to an anaplastic, androgen-independent, fast growing, highly metastatic state. This spontaneously arising tumor, termed the AT6 subline, in its original host was heterogeneously composed of both a well differentiated and an anaplastic population of cancer cells in which areas of squamous cell differentiation were occasionally observed. The original animal bearing this heterogeneous AT6 cancer developed multiple metastases, the lung metastases being heterogeneously composed of anaplastic and squamous cell populations. Cytogenetic analysis demonstrated that the lung metastases were derived from a specific subpopulation of cancer cells present in the original AT6 primary tumor. Immunohistochemical studies demonstrated that only the area of lung metastases displaying squamous morphology were positive for E-cadherin. In contrast, the anaplastic areas of the lung metastases and the metastases in other organs were E-cadherin negative. By the first passage of the AT6 tumor only the anaplastic cells were present and no detectable E-cadherin mRNA or protein was found in the primary tumor and metastatic deposits. These results suggest that a decreased expression of E-cadherin is associated with the progression of prostatic cancer.     

Invasive potential and substrate dependence of attachment in the dunning R-3327 rat prostate adenocarcinoma model

Cancer cell attachment to and invasion of the extracellular matrix has been associated with the metastatic potential of cell lines of the Dunning R-3327 rat prostatic adenocarcinoma model. We investigated the cell-matrix interactions of prostate tumor cells by comparing the invasive ability through reconstructed extracellular matrix and attachment upon EHS NATRIX (natural extracellular matrix), fibronectin, laminin, and collagen Type IV. We observed a correlation between metastatic potential and substrate dependence of attachment in prostate cancer cells. Nonmetastatic AT-1 cells possessed a higher adhesive potential to extracellular matrix components than the highly metastatic cells (ML, MLL and AT-3). It was also found that the invasive potential of the three highly metastatic cell lines was significantly higher than that of the nonmetastatic cell line. Here, it is reported that the ability to traverse a matrigel matrix correlates with their metastatic potential. These observations suggest that the extracellular matrix components are highly involved in influencing prostate cancer cell activities. In addition, we investigated the effects of two differentiation agents, retinoic acid (RA) and difluoromethylornithine (DFMO), on the adhesive and invasive profiles of the tumor cells. After treatment with both agents, adhesion was increased to levels not different from nonmetastatic cells. Furthermore, the ability of highly metastatic cells to traverse a matrigel barrier was significantly reduced after treatment with both differentiation agents. These results suggest that RA and DFMO are capable in reversing the metastatic potential of prostate cancer cells in vitro and may give a possible insight into their role as potential therapeutic agents in vivo.     

Inserting chromosome 18 into pancreatic cancer cells switches them to a dormant metastatic phenotype.

We demonstrated previously that restoration of chromosome 18 suppressed growth of pancreatic cancer cells in vitro, as well as that of tumors inoculated into nude mice. We also demonstrated that loss of 18q was associated with poor prognosis. Hence there is the possibility that the 18q arm harbors a gene(s) implicated in tumor progression and/or metastasis. In this study, we evaluated the effect of restoring chromosome 18 on metastasis in a few human pancreatic cancer cell lines with and without inactivation of SMAD4. After microcell-mediated chromosome 18 transfer, hybrid cells showed more than a 10-fold weaker metastatic ability than corresponding parental cells; mice injected with 1.25 x 10(6)/250 micro l hybrid clones via tail vein had less than one-tenth of the number of macroscopic metastases in the lung when compared with the control cells. Microscopic examination confirmed the decrease in the number of metastatic lesions. After inoculation of hybrid cells, more than 80% of the high-power fields showed no micrometastases, contrasting with their abundance after using the parental cells. Hybrid cells restored maspin expression irrespective of SMAD4 status in corresponding parental cells. On the other hand, significantly lower vascular endothelial growth factor and matrix metalloproteinase 2 secretion was observed by measuring levels in the conditioned media (CM); the averages were 22% and 20%, respectively. Angiogenesis assays using in vivo Matrigel plugs demonstrated that less neovascularization was observed in nude mice with hybrid cells than with corresponding parental cells. When cells were treated with CM from hybrids, the migration of human umbilical vascular endothelial cells was decreased, but it was partially restored with anti-vascular endothelial growth factor neutralizing antibody, as compared with CM from parental cells. These data represent the first functional evidence suggesting that chromosome 18q encodes a gene that strongly suppresses metastatic activity, possibly through dormancy.     

Suppression of tumorigenicity of A549 lung adenocarcinoma cells by human chromosomes 3 and 11 introduced via microcell-mediated chromosome transfer

To map tumor suppressor genes for lung adenocarcinomas, we introduced normal human chromosomes 3, 7, and 11 into the A549 tumor cell line by microcell-mediated chromosome transfer to test which chromosomes had the ability to suppress tumorigenicity. These human chromosomes, which contain the neomycin gene as a selectable marker, were transferred into A549 lung adenocarcinoma cells at frequencies of 0.3–1.8 × 10^?6. Two microcell hybrid clones with an introduced chromosome 3, two with an introduced chromosome 7, and six with an introduced chromosome 11 were isolated and examined for their growth properties and tumorigenicity in nude mice. Whereas parental A549 cells formed tumors with an average latency of 68 d, both microcell hybrids with an introduced chromosome 3 failed to form tumors for over 360 d. Similar tumorigenicity results were obtained when the clones were implanted into denuded tracheas, a more orthotopic transplantation site. The two clones with an introduced chromosome 7 were still tumorigenic; they formed tumors within 100–123 d after injection and grew progressively, although the tumors grew slightly slower than the parental cells did. Among the six clones with an introduced chromosome 11, one clone was still highly tumorigenic but did not contain an extra copy of an intact introduced chromosome 11. Three clones with a single intact copy of introduced chromosome 11 formed tumors with latency periods significantly longer than those of the parental cells. Two clones had two copies of the introduced chromosome 11, and both failed to form tumors within 1 yr of injection. These results indicate that chromosomes 3 and 11 can suppress the tumorigenicity of A549 lung adenocarcinoma cells.     

Correlation between reduction of metastasis in the MDA-MB-435 model system and increased expression of the Kai-1 protein

Using microcell-mediated transfer of a normal chromosome 11 into the highly metastatic MDA-MB-435 human breast carcinoma cell line, we previously showed that human chromosome 11 contains a metastasis-suppressor gene for breast cancer. A known metastasis-suppressor gene, kai-1, and a related family member, tapa-1, have been mapped to chromosome 11p11.2 and 11p15.5, respectively. To determine if these genes are responsible for the metastasis suppression seen in our microcell hybrids, we examined their expression by western blot analysis. Although tapa-1 expression did not significantly correlate with metastasis suppression, kai-1 production was dramatically increased in the metastasis-suppressed chromosome 11 microcell hybrids and unchanged in the metastatic chromosome 6 controls. Transfection of full-length kai-1 cDNA into MDA-MB-435 cells resulted in clones that did not have a significantly decreased in vivo incidence of lung metastases. However, western blot analysis showed that the primary tumors and the metastatic lesions of the transfectants had decreased levels of kai-1 protein compared with the inoculated cells. Furthermore, several of the transfectant clones expressed heavily modified kai-1 protein compared with that of the microcell hybrids. Our data indicate that protein modification may affect the normal function of kai-1 in vivo and that a threshold level of kai-1 protein expression may be necessary for suppression of the metastatic phenotype. Mol. Carcinog. 21:111–120, 1998. © 1998 Wiley-Liss, Inc.     

Genetic instability and the acquisition of metastatic ability by rat mammary cancer cells following v-H-ras oncogene transfection

When nonmetastatic dimethylbenzanthracene-induced rat mammary cancer cells (RMC1) were transfected with a control plasmid containing the neomycin resistance gene (i.e., Neo/Only), none of the five clonal Neo/Only transfectants isolated and characterized was metastatic, only one of five had a single structural chromosomal abnormality, and only one of five had a single numerical chromosomal change not present in the untransfected parental RMC1 cells. In contrast, when RMC1 cells were transfected with a plasmid containing both the neo resistance and mutated v-H-ras oncogene (i.e., Neo/Ras), four of nine clonal Neo/Ras transfectants isolated and characterized were highly metastatic, and all nine had multiple structural and/or additional numerical chromosomal abnormalities. The frequency of transfectants which had structural and/or additional numerical chromosomal changes in Neo/Ras transfectants was significantly higher than that in Neo/Only transfectants (P less than 0.05). In addition, seven of nine Neo/Ras transfectants had structural abnormalities in chromosome 1, whereas none of five Neo/Only transfectants had such abnormalities (P less than 0.05). All four Neo/Ras transfectants that were highly metastatic had structural aberrations involving a gain in chromosome 4. In contrast, none of the three Neo/Ras transfectants which were of low metastatic ability had a similar aberration involving chromosome 4. Correlation between a gain in chromosome 4 and a gain of high metastatic ability was significantly (P less than 0.05). These results demonstrate that, when RMC1 cells are transfected with v-H-ras, transfectants expressing the mutated v-H-ras p21 become genetically unstable and undergo chromosomal changes. These studies suggest that, if the appropriate chromosomal changes occur, these v-H-ras transfectants acquire high metastatic ability.     

Metastatic potential of cloned murine melanoma cells transfected with activated c-Ha-ras

We sought to determine whether the transfection of tumorigenic but not metastatic cells with the activated c-Ha-ras oncogene was invariably associated with acquisition of the metastatic phenotype. Three clonally derived lines of the K-1735 murine melanoma, characterized as nonmetastatic or poorly metastatic, were transfected with plasmids containing the 6.6-kilobase BamHI fragment of the mutant human c-Ha-ras gene and the neo gene, that confers resistance to neomycin (pSV2neoEJ). Cells transfected with pSV2neo, a plasmid containing the neo gene, served as controls for the procedure of Polybrene-mediated transfection. All cell lines were injected into syngeneic C3H/HeN and into athymic mice, and the results were compared with those produced by highly metastatic K-1735 M-2 cells. Although the pSV2neoEJ-transfected cells produced more rapidly growing s.c. tumors than the control cell lines did, the incidence of spontaneous metastasis was not increased. Following i.v. inoculation, the c-Ha-ras transfectants were retained in lung vasculature in greater proportions than pSV2neo counterpart transfectants were. The c-Ha-ras transfectants also produced significantly more lung tumor colonies, which grew faster than the few lung tumor colonies in mice given injections of control melanoma cells. We concluded that transfection of the activated c-Ha-ras oncogene into nonmetastatic K-1735 melanoma cells leads to accelerated tumor growth in vivo and can confer the ability to form lung colonies after i.v. injection but not the ability to metastasize from a primary s.c. tumor.     

In vitro motility of BSp73 rat tumor cells with different metastatic ability

The highly metastatic and nonmetastatic variants of the rat tumor BSp73 have been tested for motility in vitro by a phagokinetic assay. Surprisingly, the nonmetastatic variant is locomotional, whereas the highly metastatic variant exhibited stationary motility only. Soluble factors of normal rat serum and medium conditioned by regenerating normal rat lung, or solid constituents of the extracellular matrix material which was exudated by normal rat lung fibroblasts or bovine cornea endothelial cells as well as collagen type III could not stimulate locomotory ability in these tumor cells. However, by contact to rat lung cells they acquired passive mobility. The significance of these cellular properties for their metastatic behavior is discussed.     

Suppression of the tumorigenic phenotype of a rat liver epithelial tumor cell line by the p11.2–p12 region of human chromosome 11

Comparative chromosomal mapping studies and investigations of tumor-associated chromosomal abnormalities suggest that the development of hepatic tumors in humans and rats may share a common molecular mechanism that involves inactivation of the same tumor suppressor genes or common genetic loci. We investigated the potential of human chromosomes 2 and 11 to suppress the tumorigenic phenotype of rat liver epithelial tumor cell lines. These tumor cell lines (GN6TF and GP7TB) display elevated saturation densities in culture, efficiently form colonies in soft agar, and produce subcutaneous tumors in 100% of syngeneic rat hosts with short latency periods. Introduction of human chromosome 11 by microcell fusion markedly altered the tumorigenicity and the transformed phenotype of GN6TF cells. In contrast, the tumorigenic potential and phenotype of GP7TB cells was unaffected by the introduction of human chromosome 11, indicating that not all rat liver tumor cell lines can be suppressed by loci carried on this chromosome. Introduction of human chromosome 2 had little or no effect on the tumorigenicity or cellular phenotype of either tumor cell line, suggesting the involvement of chromosome 11–specific loci in the suppression of the GN6TF tumor cell line. The GN6TF-11^neo microcell hybrid cell lines displayed significantly reduced saturation densities in monolayer cultures, and their ability to grow in soft agar was completely inhibited. Although GN6TF-11^neo cells ultimately formed tumors in 80–100% of syngeneic rat hosts, the latency period for tumor formation was much longer. Molecular characterization of GN6TF-11^neo microcell hybrid cell lines indicated that some of the clonal lines had spontaneously lost significant portions of the introduced human chromosome, partially delineating the chromosomal location of the putative tumor suppressor locus to the region between the centromere and 11p12. Molecular examination of microcell hybrid–derived tumor cell lines further defined the minimal portion of human chromosome 11 capable of tumor suppression in this model system to the region 11p11.2-p12. © 1995 Wiley-Liss, Inc.     

Cancer cell motility-inhibitory protein in the Dunning adenocarcinoma model.

Cell motility has been associated with metastatic ability in the Dunning R3327 rat prostatic adenocarcinoma model. Cancer cell motility promoters but not inhibitors have been described by many investigators. Serum-containing and serum-free media conditioned by the nonmotile, nonmetastatic G Dunning subline inhibited the motility of the highly motile, highly metastatic MAT-LyLu subline. Motility inhibition by the G subline-conditioned serum-free media was lost upon heating to 100 degrees C and by treatment with trypsin. The motility inhibitory protein(s) had a molecular weight exceeding 50,000 as determined by diafiltration. G subline-conditioned RPMI 1640 contained several proteins with molecular weights of between approximately 53,000 and 116,000 when analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. One of these bands may represent the first inhibitor of cancer cell motility identified.     

Suppression of tumorigenicity in somatic cell hybrids. II. Human chromosomes implicated as suppressors of tumorigenicity in hybrids with Chinese hamster ovary cells

Nontumorigenic diploid human cells were fused with tumorigenic Chinese hamster ovary cells (CHO), and the hybrids were tested for tumorigenicity to determine if specific human chromosomes are associated with suppression of tumorigenicity in cell hybrids. Chromosome complements of cells of 62 nontumorigenic and 45 tumorigenic hybrids (divided into those of low, medium, and high tumorigenicity) as well as 44 tumors derived from the tumorigenic hybrids were determined by both analysis of banded chromosomes and assays of gene markers. Although no single human chromosome was consistently associated with the suppressed phenotype, chromosome 2 was never found in tumor cells, and chromosomes 9, 10, 11, and 17 were found at very low incidences in tumor cells, which suggested that they carry tumorigenicity suppressor information. Since not all suppressed hybrids contained these chromosomes, it is likely that they suppressed tumorigenicity only in combination with each other or other chromosomes. Nine chromosomes in 12 pairwise combinations of nonhomologous chromosomes were not found in tumor cells and were found at an incidence of 5% or less in hybrids of both medium and high tumorigenicity. Other experiments implicated 11 of these combinations involving only 8 chromosomes (chromosomes 4, 7, 8, 9, 10, 11, 13, and 17) as those primarily involved in suppression. Whether chromosome 2 requires another chromosome to effect suppression could not be determined. Further evaluations of the implicated suppressors, including selection of tumorigenic segregants from a panel of suppressed hybrids, again implicated the same chromosomes and their combinations in suppression. Oncogenes have been mapped to many of these chromosomes, and they are frequently involved in tumor-type-specific numerical or structural abnormalities in human neoplasias. The combined evidence suggests that specific human chromosomes of a normal cell carry genes that can regulate several cell phenotypes necessary for the expression of tumorigenicity.