Tumorigenicity of simian virus 40-hepatocyte cell lines: effect of in vitro and in vivo passage on expression of liver-specific genes and oncogenes.

Five simian virus 40 (SV40)-hepatocyte cell lines were examined for tumorigenicity and the effect of in vitro passage on the expression of four liver-specific genes (albumin, transferrin, alpha 1-antitrypsin, and phosphoenolpyruvate carboxykinase), two oncogenes (c-Ha-ras and c-raf), and two genes associated with hepatocarcinogenesis (alpha-fetoprotein and placental-type glutathione-S-transferase). At low passage (12 to 22), all five cell lines expressed the four liver-specific genes at levels similar to those in the liver and were not tumorigenic or were weakly tumorigenic. At high passage (33 to 61), the cell lines formed carcinomas, and four out of five cell lines produced primary tumors that metastasized. At least two cell lines produced well-differentiated hepatocellular carcinomas that expressed liver-specific RNAs. Levels of expression of liver-specific genes changed with time in culture. Some of the changes in liver-specific gene expression in the tumor tissue (such as for the phosphoenolpyruvate carboxykinase gene) paralleled those that occurred with in vitro passage, while other changes (such as for the albumin gene) did not parallel those that occurred with in vitro passage. Correlations between enhanced expression of c-Ha-ras and tumorigenic potential and between the process of SV40 immortalization and induced expression of c-raf and glutathione-S-transferase-P were observed. Induction of alpha-fetoprotein was detected with in vitro and in vivo passage only in the CWSV14 cell line and was paralleled by diminished albumin expression. In conclusion, we developed a model system with five SV40-hepatocyte cell lines, tumors induced by them, and tumor cell lines to examine changes in gene expression that accompany the progression from a normal cell to a hepatocellular carcinoma. Because the SV40-hepatocyte cell lines and tumor cell lines remain highly differentiated and vary in the magnitude of expression of specific genes, they can be used to study the molecular mechanisms regulating gene expression, in particular those regulating specific genes associated with differentiation.     

Expression of heat shock proteins and heat shock protein messenger ribonucleic acid in human prostate carcinoma in vitro and in tumors in vivo

Heat shock proteins (HSPs) are thought to play a role in the development of cancer and to modulate tumor response to cytotoxic therapy. In this study, we have examined the expression of hsf and HSP genes in normal human prostate epithelial cells and a range of prostate carcinoma cell lines derived from human tumors. We have observed elevated expressions of HSF1, HSP60, and HSP70 in the aggressively malignant cell lines PC-3, DU-145, and CA-HPV-10. Elevated HSP expression in cancer cell lines appeared to be regulated at the post-messenger ribonucleic acid (mRNA) levels, as indicated by gene chip microarray studies, which indicated little difference in heat shock factor (HSF) or HSP mRNA expression between the normal and malignant prostate cell lines. When we compared the expression patterns of constitutive HSP genes between PC-3 prostate carcinoma cells growing as monolayers in vitro and as tumor xenografts growing in nude mice in vivo, we found a marked reduction in expression of a wide spectrum of the HSPs in PC-3 tumors. This decreased HSP expression pattern in tumors may underlie the increased sensitivity to heat shock of PC-3 tumors. However, the induction by heat shock of HSP genes was not markedly altered by growth in the tumor microenvironment, and HSP40, HSP70, and HSP110 were expressed abundantly after stress in each growth condition. Our experiments indicate therefore that HSF and HSP levels are elevated in the more highly malignant prostate carcinoma cells and also show the dominant nature of the heat shock-induced gene expression, leading to abundant HSP induction in vitro or in vivo.     

Heterogeneity of anticancer drug sensitivity in squamous cell carcinoma of the tongue

Heterogeneity is known to be present to varying degrees in cancer cell groups. There have been no reports, however, of studies in which a single cell clone was prepared from a cancer cell group to examine heterogeneity with respect to anticancer drug sensitivity. Thus, the authors herein report an investigation into the heterogeneity of cancer cells within the same tumor with respect to anticancer drug sensitivity. Anticancer drug sensitivity was investigated in primary tumors, metastatic lymph node tumors, recurrent tumors and established cell lines obtained from four cases of tongue cancer using an oxygen electrode apparatus. As differences were observed in anticancer drug sensitivity from one case to another, even though all four were of the same pathological tissue type, the individual differences were apparently significant. Moreover, primary tumors and recurrent tumors demonstrated different sensitivities to the anticancer drugs even in the same patient. When single cell clones were prepared from primary tumors and anticancer drug sensitivity testing was carried out, sensitivity to anticancer drugs that was not seen in the primary tumors was observed. We performed RT-PCR on cell groups derived from this single cell using MDR1, MRP1, MRP2 and ERCC1, which are primary genes that are resistant to anticancer drugs. Expression of MDR and ERCC1 was not observed in single cell clones nos. 1-10. MRP1 and MRP2, on the other hand, were expressed in all of these single cell clones. Because cells with different sensitivity levels were initially present in the cancer cell groups, even when large numbers of cancer cells died in response to anticancer drug therapy, the results suggest the possibility that recurrence and metastasis occur based on cells with differing sensitivities. After examining anticancer drug sensitivity at the single cell level, we believe that anticancer drug-resistant genes may be involved in the heterogeneity of anticancer drug sensitivity with respect to cancer cell groups.     

A transforming activity not detected by DNA transfer to NIH 3T3 cells is detected by JB6 mouse epidermal cells.

Transfection of four different mouse epidermal tumor cell DNAs into NIH 3T3 cells yielded neither morphologically altered foci nor anchorage independence. However, promotion-sensitive, but not promotion-insensitive, JB6 mouse epidermal cell lines were permissive for the expression of anchorage independence after transfection of DNA from three of these tumor cell lines. This transforming activity and the promotion-sensitive activity that confers sensitivity to promotion of transformation show differences in restriction enzyme sensitivity. In view of this difference and the differences in both recipient cells and 12-O-tetradecanoyl-phorbol-13-acetate dependence of expression, it appears that the transforming activity and the promotion-sensitive activity are specified by different genes. The JB6 promotion-sensitive cell lines may be useful for detecting and cloning transforming genes that escape detection in the NIH 3T3 cell focus assay.     

Colorectal Cancer Migration and Invasion Initiated by microRNA-106a

MicroRNAs have been implicated in the regulation of several cellular signaling pathways of colorectal cancer (CRC) cells. Although emerging evidence proves that microRNA (miR)-106a is expressed highly in primary tumor and stool samples of CRC patients; whether or not miR-106a mediates cancer metastasis is unknown. We show here that miR-106a is highly expressed in metastatic CRC cells, and regulates cancer cell migration and invasion positively in vitro and in vivo. These phenotypes do not involve confounding influences on cancer cell proliferation. MiR-106a inhibits the expression of transforming growth factor-β receptor 2 (TGFBR2), leading to increased CRC cell migration and invasion. Importantly, miR-106a expression levels in primary CRCs are correlated with clinical cancer progression. These observations indicate that miR-106a inhibits the anti-metastatic target directly and results in CRC cell migration and invasion.     

Identification of the potential prognostic genes of human melanoma

The melanoma is one of the most dangerous forms of skin diseases. It may spread to other parts of the body and cause serious illness and death. Early detection and diagnosis are crucial. However, the systemic expression analysis for the different staging of melanoma is still lacking to date. In this study, we analyzed the gene expression profiles of the different staging of melanoma by the differential expression analysis and random forest analysis. First, the results of the principal component analysis showed that the clustering of primary tumor samples, normal samples, and pigment nevus samples got closer, while the clustering of tumor metastatic samples and normal samples was far away. Moreover, the gene expression of tumor metastasis stage and the initial stage had obvious differences. Almost 426 genes identified had differential expression. The functional enrichment of differentially expressed genes was associated with the epidermal cell differentiation, epidermis development, and the keratinocyte differentiation. Taken together, our findings identified the differentially expressed signatures between primary melanoma and metastatic melanoma. Our results would provide the potential mechanisms of melanoma.     

Gene expression profile of human bone marrow stromal cells determined by restriction fragment differential display analysis

Using an in vitro osteogenic culture system, we carried out a restriction fragment differential display (RFDD-PCR) to identify genes expressed by these cells in their undifferentiated stage and not expressed, or expressed at a lower level, in a closely related but distinct cell type: bone marrow stromal cells (BMSC)-derived osteoblasts (BDO). Forty-seven candidate regulated genes, selected by RFDD, were analyzed by RT-PCR analysis in three cell clones and in primary cultures from seven different donors. A subset of three genes were confirmed as upregulated in BMSC relative to BDO in every primary culture and cloned population examined: betaIG-h3, IGFbp3, and LOXL2. Their differential expression was confirmed by Northern analysis and the corresponding proteins were detected by immunolocalization in BMSC.     

Establishment of murine cell lines by constitutive and conditional immortalization

Mouse cell lines were immortalized by introduction of specific immortalizing genes. Embryonic and adult animals and an embryonal stem cell line were used as a source of primary cells. The immortalizing genes were either introduced by DNA transfection or by ecotropic retrovirus transduction. Fibroblasts were obtained by expression of SV40 virus large T antigen (TAg). The properties of the resulting fibroblast cell lines were reproducible, independent of the donor mouse strains employed and the cells showed no transformed properties in vitro and did not form tumors in vivo. Endothelial cell lines were generated by Polyoma virus middle T antigen expression in primary embryonal cells. These cell lines consistently expressed relevant endothelial cell surface markers. Since the expression of the immortalizing genes was expected to strongly influence the cellular characteristics fibroblastoid cells were reversibly immortalized by using a vector that allows conditional expression of the TAg. Under inducing conditions, these cells exhibited properties that were highly similar to the properties of constitutively immortalized cells. In the absence of TAg expression, cell proliferation stops. Cell growth is resumed when TAg expression is restored. Gene expression profiling indicates that TAg influences the expression levels of more than 1000 genes that are involved in diverse cellular processes. The data show that conditionally immortalized cell lines have several advantageous properties over constitutively immortalized cells.     

Dipeptidyl peptidase-IV inhibits glioma cell growth independent of its enzymatic activity

Malignant gliomas exhibit abnormal expression of proteolytic enzymes that may participate in the uncontrolled cell proliferation and aberrant interactions with the brain extracellular matrix. The multifunctional membrane bound serine aminopeptidase dipeptidyl peptidase (DPP)-IV has been linked to the development and progression of several malignancies, possibly both through the enzymatic and nonenzymatic mechanisms. In this report we demonstrate the expression of DPP-IV and homologous proteases fibroblast activation protein, DPP8 and DPP9 in primary cell cultures derived from high-grade gliomas, and show that the DPP-IV-like enzymatic activity is negatively associated with their in vitro growth. More importantly, the DPP-IV positive subpopulation isolated from the primary cell cultures using immunomagnetic separation exhibited slower proliferation. Forced expression of the wild as well as the enzymatically inactive mutant DPP-IV in glioma cell lines resulted in their reduced growth, migration and adhesion in vitro, as well as suppressed glioma growth in an orthotopic xenotransplantation mouse model. Microarray analysis of glioma cells with forced DPP-IV expression revealed differential expression of several candidate genes not linked to the tumor suppressive effects of DPP-IV in previous studies. Gene set enrichment analysis of the differentially expressed genes showed overrepresentation of gene ontology terms associated with cell proliferation, cell adhesion and migration. In conclusion, our data show that DPP-IV may interfere with several aspects of the malignant phenotype of glioma cells in great part independent of its enzymatic activity.     

Expression of matrix metalloproteinase 8 (MMP-8) and tyrosinase-related protein-1 (TYRP-1) correlates with the absence of metastasis in an isogenic human breast cancer model

The multi-step nature of metastasis poses difficulties in both design and interpretation of experiments to unveil the mechanisms causing the process. In order to facilitate such studies, we have previously derived a pair of breast tumor cell lines that originate from the same breast tumor but which have diametrically opposite metastatic capabilities. In this system, the monoclonal cell line M-4A4 is metastatic to the lungs of athymic mice, whereas clone NM-2C5 is equally tumorigenic but non-metastatic. Here, we report that representational difference analysis (RDA) of cDNA obtained from the two clonal populations revealed an increased expression of tyrosinase-related protein-1 (TYRP-1) and the matrix metalloproteinase-8 (MMP-8) genes in the non-metastatic cell line. RNA and protein analyses in cultured cells and in primary xenograft tissues confirmed that the non-metastatic cell line expresses TYRP-1 and MMP-8 at levels that are at least 20-fold higher than the metastatic counterpart. Other members of the MMP family (MMP-9 and MMP-2) and the tissue inhibitor of metalloproteinase-2 (TIMP-2) were found to be expressed at similar levels in both populations. The effects of MMP-8 and TYRP-1 on in vitro invasion and migration were assessed in cells whose expression of these genes was altered by stable transduction with sense and antisense constructs. Specific down-regulation of MMP-8 in non-metastatic NM-2C5 cells resulted in a 2.5-fold increased capacity to invade through Matrigel. Unlike other members of the matrix metalloproteinase family, MMP-8 has not previously been implicated in the processes of tumorigenesis or metastasis. The successful identification of two proteins that are differentially expressed in these matched clonal cell lines and the tumors that they produce demonstrates the feasibility of using this approach to search for genes that are associated with aberrant differentiation toward metastatic behavior.     

DMD and IL1RAPL1: two large adjacent genes localized within a common fragile site (FRAXC) have reduced expression in cultured brain tumors

Common fragile sites (CFSs) are large regions of profound genomic instability found in all individuals. Spanning the center of the two most frequently expressed CFS regions, FRA3B (3p14.3) and FRA16D (16q23.2), are the 1.5 Mb FHIT gene and the 1.0 Mb WWOX gene. These genes are frequently deleted and/or altered in many different cancers. Both FHIT and WWOX have been demonstrated to function as tumor suppressors, both in vitro and in vivo. A number of other large CFS genes have been identified and are also frequently inactivated in multiple cancers. Based on these data, several additional very large genes were tested to determine if they were derived from within CFS regions, but DCC and RAD51L1 were not. However, the 2.0 Mb DMD gene and its immediately distal neighbor, the 1.8 Mb IL1RAPL1 gene are CFS genes contained within the FRAXC CFS region (Xp21.2-->p21.1). They are abundantly expressed in normal brain but were dramatically underexpressed in every brain tumor cell line and xenograft (derived from an intracranial model of glioblastoma multiforme) examined. We studied the expression of eleven other large CFS genes in the same panel of brain tumor cell lines and xenografts and found reduced expression of multiple large CFS genes in these samples. In this report we show that there is selective loss of specific large CFS genes in different cancers that does not appear to be mediated by the relative instability within different CFS regions. Further, the inactivation of multiple large CFS genes in xenografts and brain tumor cell lines may help to explain why this type of cancer is highly aggressive and associated with a poor clinical outcome.     

Influence of Burkitt's lymphoma and primary B cells on latent gene expression by the nonimmortalizing P3J-HR-1 strain of Epstein-Barr virus.

The Epstein-Barr virus (EBV) genes expressed in B lymphocytes immortalized in vitro or in Burkitt's lymphoma (BL) cells infected in vivo have been characterized previously; however, the viral products which are essential for immortalization or for establishment of EBV latency are still not known. To approach this question, we compared the kinetics of expression of EBV nuclear antigens and the two EBV-encoded small RNAs, EBER1 and EBER2, after infection of primary B cells or EBV genome-negative BL cells with either an immortalizing EBV strain (B95-8) or the nonimmortalizing deletion mutant (HR-1). Following infection of primary cells with B95-8 virus, EBV nuclear antigen (EBNA)-2 was expressed first, followed by EBNA-1, -3, and -4 (also called leader protein [LP]) and the two small RNAs. Infection of EBV genome-negative BL cells with the same strain of virus resulted in a similar pattern of gene expression, except that the EBNAs appeared together and more rapidly. EBERs were not apparent in one BL cell line converted by B95-8. The only products detected after infection of primary B lymphocytes with the HR-1 deletion mutant were the EBNA-4 (LP) family and trace amounts of EBER1. Although HR-1 could express neither EBNA-1, EBNA-3, nor EBER2 in primary cells, all these products were expressed rapidly after HR-1 infection of EBV genome-negative BL cell lines. The results indicate that the mutation in HR-1 virus affects immortalization not only through failure to express EBNA-2, a gene which is deleted, but also indirectly by curtailing expression of several other EBV genes whose coding regions are intact in the HR-1 virus and normally expressed during latency. The pattern of latent EBV gene expression after HR-1 infection is dependent on the host cell, perhaps through products specific for the cell cycle or the state of B-cell differentiation.     

Heterogeneity among human nasopharyngeal carcinoma cell lines for inflammatory cytokines mRNA expression levels.

Using polymerase chain reaction (PCR), we confirmed the expression of interleukin-1 alpha (IL-1 alpha) by the human nasopharyngeal carcinoma (NPC) cell line C15 without contribution of either human IL-1 beta or mouse IL-1 alpha in the biological activity previously found in C15. However we showed that IL-1 alpha was not expressed in all NPCs. IL-1 beta and/or tumor necrosis factor (TNF)-alpha genes could also be activated, independently from the number of Epstein Barr Virus (EBV) copies harbored by the cells. Interestingly, the primary tumor C15 showed a profile of TNF-sensitive tumor while C17, C18 and C19 which were derived from metastasis have a typical profile of TNF-resistant cells. Furthermore, the inflammatory cytokines whose genes are classically induced by IL-1 and TNF were found expressed only in C17 and C19 suggesting another level of heterogeneity among NPCs.