Inverse PPARß/δ agonists suppress oncogenic signaling to the ANGPTL4 gene and inhibit cancer cell invasion.

Besides its established functions in intermediary metabolism and developmental processes, the nuclear receptor peroxisome proliferator-activated receptor ß/δ (PPARß/δ) has a less defined role in tumorigenesis. In the present study, we have identified a function for PPARß/δ in cancer cell invasion. We show that two structurally divergent inhibitory ligands for PPARß/δ, the inverse agonists ST247 and DG172, strongly inhibit the serum- and transforming growth factor ß (TGFß)-induced invasion of MDA-MB-231 human breast cancer cells into a three-dimensional matrigel matrix. To elucidate the molecular basis of this finding, we performed chromatin immunoprecipitation sequencing (ChIP-Seq) and microarray analyses, which identified the gene encoding angiopoietin-like 4 (ANGPTL4) as the major transcriptional PPARß/δ target in MDA-MB-231 cells, previously implicated in TGFß-mediated tumor progression and metastatic dissemination. We show that the induction of ANGPTL4 by TGFß and other oncogenic signals is strongly repressed by ST247 and DG172 in a PPARß/δ-dependent fashion, resulting in the inhibition of ANGPTL4 secretion. This effect is attributable to these ligands' ability to induce a dominant transcriptional repressor complex at the site of transcription initiation that blocks preinitiation complex formation through an histone deacetylase-independent, non-canonical mechanism. Repression of ANGPTL4 transcription by inverse PPARß/δ agonists is functionally linked to the inhibition of cancer cell invasion into a three-dimensional matrix, as (i) invasion of MDA-MB-231 cells is critically dependent on ANGPTL4 expression, (ii) recombinant ANGPTL4 stimulates invasion, and (iii) reverses the inhibitory effect of ST247 and DG172. These findings indicate that a PPARß/δ-ANGPTL4 pathway is involved in the regulation of tumor cell invasion and that its pharmacological manipulation by inverse PPARß/δ agonists is feasible.

Compensatory PI3-kinase/Akt/mTor activation regulates imatinib resistance development.

BCR/ABL-kinase mutations frequently mediate clinical resistance to the selective tyrosine kinase inhibitor Imatinib mesylate (IM, Gleevec). However, mechanisms that promote survival of BCR/ABL-positive cells before clinically overt IM resistance occurs have poorly been defined so far. Here, we demonstrate that IM-treatment activated the phosphatidylinositol 3-kinase (PI3K)/Akt/mammalian target of rapamycin (mTor)-pathway in BCR/ABL-positive LAMA-cells and primary leukemia cells in vitro, as well as in a chronic phase CML patient in vivo. In fact, PI3K/Akt-activation critically mediated survival during the early phase of IM resistance development before manifestation of BCR/ABL-dependent strong IM resistance such as through a kinase mutation. Accordingly, inhibition of IM-induced Akt activation using mTor inhibitors and Akt-specific siRNA effectively antagonized development of incipient IM-resistance in vitro. In contrast, IM-resistant chronic myeloid leukemia (CML) patients with BCR/ABL kinase mutations (n=15), and IM-refractory BCR/ABL-positive acute lymphatic leukemia patients (n=2) displayed inconsistent and kinase mutation-independent autonomous patterns of Akt-pathway activation, and mTor-inhibition overcame IM resistance only if Akt was strongly activated. Together, an IM-induced compensatory Akt/mTor activation may represent a novel mechanism for the persistence of BCR/ABL-positive cells in IM-treated patients. Treatment with mTor inhibitors may thus be particularly effective in IM-sensitive patients, whereas Akt-pathway activation variably contributes to clinically overt IM resistance.

Combined transductional and transcriptional targeting of melanoma cells by artificial virus-like particles.

BACKGROUND: Artificial virus-like particles (AVPs) represent a novel type of liposomal vector resembling retroviral envelopes. AVPs are serum-resistant and non-toxic and can be endowed with a peptide ligand as a targeting device. The vitronectin receptor, alphavbeta3-integrin, is commonly upregulated on malignant melanoma cells. In the present study we investigated whether AVPs carrying cyclic peptides with an RGD integrin binding motif (RGD-AVPs) are suitable for the specific and efficient transduction of human melanoma cells. METHODS: Plasmid DNA was complexed with low molecular weight non-linear polyethyleneimine and packaged into anionic liposomes. Transduction efficiencies were determined after transient transfection of different cell lines in serum-free medium using green fluorescent protein or luciferase reporter genes. RESULTS: We demonstrated that RGD-AVPs transduced human melanoma cells with high efficiencies of > 60%. Efficient transduction was clearly dependent on the presence of the cyclic RGD ligand and was selective for melanoma cells. The specificity of the vector system could be further enhanced by using the melanocyte-specific tyrosinase promoter to drive transgene expression. CONCLUSION: Our findings suggest that the AVP technology is a useful approach for generating highly efficient and specific non-viral vectors for melanoma targeting, in particular in a setting of combined transductional and transcriptional targeting.

Cationized human serum albumin as a non-viral vector system for gene delivery? Characterization of complex formation with plasmid DNA and transfection efficiency.

Cationized human serum albumin (cHSA) could serve as a potential non-viral vector system for gene delivery. Native human serum albumin was cationized by covalent coupling of hexamethylenediamine to the carboxyl groups resulting in a shift of the isoelectric point from pH 4-5 to 7-9. The cationized albumin underwent spontaneous self-assembly with DNA as demonstrated by retardation of CMV-nlacZ plasmid in agarose gel electrophoresis. Photon correlation spectroscopy showed a decrease of complex size with increasing cHSA/plasmid ratios. Under optimized conditions complexes were formed with 230-260 nm mean diameter and a homogenous, narrow size distribution. At room temperature complexes were stable in 0.9% sodium chloride solution pH 7.4 for 1 h without aggregation. Process parameters such as albumin concentration, incubation time, temperature, pH, order of reagent addition, the presence of bivalent ions and the ionic strength of the complexation medium all influenced the complex size. Confocal laser scanning microscopy showed interactions of a Texas Red labeled cationized albumin with cell membranes of ECV 304 cells and an enhanced endocytic uptake compared to native albumin. The potential for introducing exogeneous DNA into cells was shown using NIH 3T3 fibroblasts. Successful, albeit low reporter gene expression could be achieved in the presence of chloroquine. Under in vitro conditions no toxic effect could be observed. In conclusion, cationized albumin may have promise as a non-toxic vector for gene delivery, especially for DNA vaccination.

Cell surface display of a lysosomal enzyme for extracellular gene-directed enzyme prodrug therapy.

Prodrug conversion is a promising approach to cytotoxic gene therapy if an efficient transfer of the generated drug to adjacent cells can be achieved. To maximize the efficacy of this strategy we sought to develop a system that is based on a human enzyme, acts extracellularly yet in close vicinity of the transduced cell and can be used with multiple prodrugs. Results obtained with a secreted version of human beta-glucuronidase suggested that this enzyme could be a suitable candidate, although a more stringent retention of the enzyme at the site of the producer cell, such as its attachment to the cell surface, would be desirable. Here, we show that the fusion of the transmembrane domain of the human PDGF receptor to a C-terminally truncated form of human beta-glucuronidase results in its surface accumulation at high steady-state levels. Using a doxorubicin prodrug, we demonstrate that this GDEPT system produces a strong bystander effect and has potent antitumor activity in vivo.

Highly efficient transduction of endothelial cells by targeted artificial virus-like particles.

Targeting the tumor vasculature by gene therapy is a potentially powerful approach, but suitable vectors have not yet been described. We have designed a new type of liposomal vector, based on the composition of anionic retroviral envelopes, that is serum-resistant and nontoxic. These artificial virus-like envelopes (AVEs) were endowed with a cyclic RGD-containing peptide as a targeting device for the a(v)beta3-integrin on tumor endothelial cells (ECs). The packaging of plasmid DNA complexed with low-molecular-weight, nonlinear polyethyleneimine into these AVEs yielded artificial virus-like particles (AVPs) that transduced ECs with efficiencies of up to 99%. In contrast, transduction of a variety of other cell types by these RGD-AVPs was comparably inefficient under the same experimental conditions. This EC selectivity was mediated, in part, but not exclusively, by the RGD ligand, as suggested by the reduced, but still relatively high, transduction efficiency seen with AVPs lacking RGD. The interaction of anionic lipids of the AVPs with ECs may therefore contribute to the observed selective and highly efficient transduction of this cell type. These findings suggest that the targeted AVE technology is a useful approach to create highly efficient nonviral vectors.

Secreted human beta-glucuronidase: a novel tool for gene-directed enzyme prodrug therapy.

A major problem of tumor gene therapy is the low transduction efficiency of the currently available vectors. One way to circumvent this problem is the delivery of therapeutic genes encoding intracellular enzymes for the conversion of a prodrug to a cytotoxic drug which can then spread to neighboring non-transduced cells (bystander effect). One possibility to improve the bystander effect could be the extracellular conversion of a hydrophilic prodrug to a lipophilic, cell-permeable cytotoxic drug. Toward this end, we have used a secreted form of the normally lysosomal human beta-glucuronidase (s-betaGluc) to establish an extracellular cytotoxic effector system that converts an inactivated glucuronidated derivative of doxorubicin (HMR 1826) to the cytotoxic drug. We demonstrate that s-betaGluc-transduced tumor cells convert HMR 1826 to doxorubicin which is taken up by both transduced and non-transduced cells. s-betaGluc in combination with HMR 1826 efficiently induces tumor cell killing both in cell culture and in vivo. This effect is mediated through a pronounced bystander effect of the generated cytotoxic drug. Most notably, this gene therapeutic strategy is shown to be clearly superior to conventional chemotherapy with doxorubicin. Gene Therapy (2000) 7, 224-231.

Inhibition of D1, D2, and A-cyclin expression in HL-60 cells by the lipid peroxydation product 4-hydroxynonenal.

4-Hydroxynonenal (HNE), a product of lipid peroxidation, is an highly reactive aldehyde that, at concentration similar to those found in normal cells, blocks proliferation and induces a granulocytic-like differentiation in HL-60 cells. These effects are accompained by a marked increase in the proportion G0/G1 cells. The mechanisms of HNE action were investigated by analyzing the expression of the cyclins and cyclin-dependent protein kinases (CDKs), controlling the cell cycle progression. Data obtained by exposing cells to dimethyl sulfoxide (DMSO) were used for comparison. 4-Hydroxynonenal downregulated both mRNA and protein contents of cyclins D1, D2, and A until 24 h from the treatments, whereas DMSO inhibited cyclin D1 and D2 expression until the end of experiment (2 days) and induces an increase of cyclin A until 1 day. Cyclins B and E, and protein kinase CDK2 and CDK4 expressions were not affected by HNE, whereas DMSO induced an increase of cyclin E, B, and CDK2 from 8 h to 1 day. These data are in agreement with previous results indicating a different time-course of accumulation in G0/G1 phases of cells treated with HNE and DMSO and suggest that the HNE inhibitory effect on proliferation and cell cycle progression may depend by the downregulation of D1, D2, and A cyclin expression.

Functional domains in cyclin D1: pRb-kinase activity is not essential for transformation.

Although cyclin D1 plays a major role during cell cycle progression and is involved in human tumourigenesis, its domain structure is still poorly understood. In the present study, we have generated a series of cyclin D1 N- and C-terminal deletion constructs. These mutants were used to define the domains required for transformation of rat embryonal fibroblasts (REF) in cooperation with activated Ha-ras and and to establish correlations with defined biochemical properties of cyclin D1. Protein binding and REF assays showed that the region of the cyclin box required for the interaction with CDK4 as well as C-terminal sequences determining protein stability were crucial for transformation. Surprisingly, however, the N-terminal deletion of 20 amino acids which impaired pRb kinase activity did not affect the transforming ability of cyclin D1. Likewise, no effect on transformation was observed with mutants defective in p21CIP interaction. These observations argue against a crucial role of pRb inactivation or p21CIP squelching in cyclin D1-mediated transformation.

CDF-1 mediated repression of cell cycle genes targets a specific subset of transactivators.

The cdc25C, cyclin A and cdc2 genes are regulated during the cell cycle through two contiguous repressor binding sites, the CDE and CHR, located in the region of transcription initiation and interacting with a factor termed CDF-1. The target of this repression seems to be transcriptional activation of these promoters by transcription factors bound upstream. The majority of these factors falls into the class of glutamine-rich activators, suggesting that CDF-1-mediated repression might be activation domain specific. In the present study we have used chimeric promoter constructs to demonstrate that the cdc25C UAS, but not the core promoter, is crucial for repression. In addition, we show that only specific transcription factors and activation domains are responsive to CDE-CHR-mediated cell cycle regulation. These observations clearly indicate that CDF-1 interferes with activation of transcription by a specific subset of transactivators. The repressible activation domains belong to the same class of glutamine-rich activators, pointing to specific interactions of CDF-1 with components of the transcription machinery. In agreement with this conclusion we find that a simple inversion of the CDF-CHR module completely abrogates cell cycle-regulated repression.

Cell cycle-independent induction of apoptosis by the anti-tumor drug Flavopiridol in endothelial cells.

The anti-tumor drug Flavopiridol is a potent inhibitor of cyclin-dependent kinases (cdks). As a consequence, Flavopiridol-treated cells arrest in both G1 and G2, but Flavopiridol has also been shown to be cytotoxic for some tumor cell lines. The underlying molecular events are, however, unclear. We now show that Flavopiridol induces apoptosis in human umbilical vein endothelial cells (HUVECs), as judged by the occurrence of classical apoptotic markers, including chromatin condensation, internucleosomal cleavage, DNA fragmentation (TUNEL assay), annexin V binding and poly(ADP-ribose) polymerase (PARP)-cleavage. Such induction of apoptosis occurs with equal efficiency in both proliferating and G0/G1-arrested cells. Because growth-arrested HUVECs lack cdk2 activity and contain high levels of the cdk inhibitor p27, our observations suggest that cell cycle regulated cdks may not be the only critical target for Flavopiridol-induced apoptosis. Surprisingly, A549 lung carcinoma cells were clearly dependent on cell proliferation for the induction of cell death, pointing to cell type-related differences in the mechanism of Flavopiridol action.

Abrogation of c-Raf expression induces apoptosis in tumor cells.

Signal transduction pathways involving the c-Raf protein kinase are frequently activated in tumor cells. We have addressed the relevance of this activation by a loss-of-function approach. An anti-sense phosphorothioate oligonucleotide (ODN) specifically targeted against c-raf mRNA (Monia et al., 1996a) was used to block c-Raf protein expression in four different cell lines derived from lung, cervical, prostate and colon carcinomas. Concomitant with the abrogation of c-Raf expression we observed the occurrence of classical apoptotic markers, including chromatin condensation, inter-nucleosomal DNA cleavage, annexin V binding and cleavage of PARP, which was followed by cell death, affecting most of the cell population. This induction of apoptosis occurred independent of the p53 status of the cell. These findings demonstrate that c-Raf can protect tumor cells from undergoing programmed cell death, and suggest that the interference with c-Raf expression or function by ODNs or specific drugs could represent a powerful means for improving the efficacy of anti-cancer therapy.

CDF-1-mediated repression of cell cycle genes targets a specific subset of transactivators.

The cdc25C , cyclin A and cdc2 genes are regulated during the cell cycle through two contiguous repressor binding sites, the CDE and CHR, located in the region of transcription initiation and interacting with a factor termed CDF-1. The target of this repression seems to be transcriptional activation of these promoters by transcription factors bound upstream. The majority of these factors falls into the class of glutamine-rich activators, suggesting that CDF-1-mediated repression might be activation domain specific. In the present study we have used chimeric promoter constructs to demonstrate that the cdc25C UAS, but not the core promoter, is crucial for repression. In addition, we show that only specific transcription factors and activation domains are responsive to CDE-CHR-mediated cell cycle regulation. These observations clearly indicate that CDF-1 interferes with activation of transcription by a specific subset of transactivators. The repressible activation domains belong to the same class of glutamine-rich activators, pointing to specific interactions of CDF-1 with components of the transcription machinery. In agreement with this conclusion we find that a simple inversion of the CDE-CHR module completely abrogates cell cycle-regulated repression.

Cell proliferation and cell cycle progression are not impaired in fibroblasts and ES cells lacking c-Fos.

The transcription factor AP-1 is thought to play an important role in the control of cell proliferation, but the function of individual Fos and Jun family members is a largely unresolved issue. To directly analyse the function of c-Fos in the control of cell proliferation we have used embryonic stem (ES) cells and fibroblasts lacking c-Fos due to a disruption of the c-fos gene by homologous recombination. Our results demonstrate that proliferation of normally cycling cells and reentry of quiescent cells into the cell cycle following serum stimulation are not c-Fos-dependent and occur with similar efficiency in c-fos-/- and control cells. We also show that there is no compensatory overexpression or activation of other known Fos or Jun family members. On the contrary, the c-fos-/- cells showed a reduced induction of fra-1 after serum stimulation which is in agreement with the previous identification of fra-1 as a c-Fos target gene. Comparison of the AP-1 binding and transactivation activities in c-fos-/- and +/+ fibroblasts by electrophoretic mobility antibody supershift and CAT assays suggests that c-Fos is not a major component of AP-1 complexes in these cells. It is therefore conceivable that the lack of any detectable effect on cell proliferation in c-fos-/- cells might be due to a functional redundancy among the different AP-1 family members.

A novel member of human tissue inhibitor of metalloproteinases (TIMP) gene family is regulated during G1 progression, mitogenic stimulation, differentiation, and senescence.

We have identified in the human diploid fibroblast cell line WI-38 a novel serum-inducible gene, mitogen-inducible gene 5 (mig-5), of the delayed-early class, which represents a new member of the family of human tissue inhibitors of metalloproteinases (TIMPs). The deduced Mig-5 protein shares the highest degree of homology with chicken TIMP-3 (74% identity) and is more distantly related to human TIMP-1 and TIMP-2 (30-38% identity), indicating that mig-5 may represent the human homolog of chicken TIMP-3. In contrast to TIMP-1 and TIMP-2, mig-5 mRNA expression is not only induced in response to mitogenic stimulation but also is subject to cell cycle regulation in normally proliferating WI-38 fibroblasts and HL-60 myeloid cells, showing a clear peak around mid-G1. In agreement with this observation, differentiation of HL-60 cells to either granulocytic or macrophage-like cells leads to increased levels of mig-5 mRNA concomitant with a block in G1. In contrast, mig-5 expression is decreased in senescent human fibroblasts, suggesting that these cells may be blocked at a stage in G1 before or after the phase of maximum mig-5 expression. Since in contrast to the vast majority of other known mitogen-inducible genes, mig-5 expression is periodically up-regulated in G1, this gene should represent an invaluable tool for the analysis of cell cycle progression, terminal differentiation, and replicative senescence.

Differential induction of 'metabolic genes' after mitogen stimulation and during normal cell cycle progression.

Mitogenic stimulation of quiescent cells not only triggers the cell division cycle but also induces an increase in cell volume, associated with an activation of cellular metabolism. It is therefore likely that genes encoding enzymes and other proteins involved in energy metabolism and biosynthetic pathways represent a major class of mitogen-induced genes. In the present study, we investigated in the non-established human fibroblast line WI-38 the induction by mitogens of 17 genes whose products play a role in different metabolic processes. We show that these genes fall into 4 different categories, i.e. non-induced genes, immediate early (IE) primary genes, delayed early (DE) secondary genes and late genes reaching peak levels in S-phase. In addition, we have analysed the regulation of these genes during normal cell cycle progression, using HL-60 cells separated by counterflow elutriation. A clear cell cycle regulation was seen with those genes that are induced in S-phase, i.e. thymidine kinase, thymidylate synthase and dihydrofolate reductase. In addition, two DE genes showed a cell cycle dependent expression. Ornithine decarboxylase mRNA increased around mid-G1, reaching maximum levels in S/G2, while hexokinase mRNA expression was highest in early G1. In contrast, the expression of other DE and IE genes did not fluctuate during the cell cycle, a result that was confirmed with elutriated WI-38 and serum-stimulated HL-60 cells. These observations suggest that G0-->S and G1-->S transition are distinct processes, exhibiting characteristic programmes of gene regulation, and merging around S-phase entry.

Identification of serum-inducible genes: different patterns of gene regulation during G0-->S and G1-->S progression.

We have identified, by differential cDNA library screening, 15 serum inducible genes in the human diploid fibroblast cell line WI-38. The genes fall into two classes that are distinguished by their dependence on protein synthesis for the induction by serum, i.e., primary and secondary genes. While 11 of these genes encode known proteins, 4 other genes have not been described to date. The former genes encode proteins of diverse functions, including the monocyte-derived neutrophil chemotactic factor (MONAP), calmodulin, tropomyosin, tenascin, collagenase, plasminogen activator inhibitor-2a, the 'sperm-specific' cleavage signal-1 protein, metallothionein IIa and the mitochondrial chaperonin hsp-60. Interestingly, one of the unknown genes contains a large open reading frame for a polypeptide that is highly homologous to a previously unidentified long open reading frame in the opposite strand of the gene coding for the transcription factor HTF-4. We also studied the regulation of these serum-induced genes during cell cycle progression in normally cycling WI-38 and HL-60 cells separated by counterflow elutriation as well as in serum-stimulated HL-60 cells. Our results clearly show that, in contrast to the prevailing opinion, the expression of most genes induced after mitogen stimulation is not subject to a significant regulation in normally proliferating cells. This supports the hypothesis that the progression into S from either G0 or G1 are distinct processes with specific patterns of gene expression.

Suppression of the growth factor-mediated induction of c-fos and down-modulation of AP-1-binding activity are not required for cellular senescence.

It has been suggested that the impaired response of the c-fos gene to serum growth factors and the concomitant loss of AP-1 activity may be a crucial step in the process of cellular senescence. In the present study, we provide evidence arguing against such a conclusion. Data obtained in five independent experiments showed that both c-fos RNA and protein expression were similar in 'young' and in senescent serum-stimulated WI-38 cells, suggesting that the previously reported suppression of c-fos induction is not an obligatory event in the process of cellular senescence. Likewise, expression of fra-1, c-jun and junB continued to be high in serum-stimulated senescent cells, while induction of fosB was reduced approximately fivefold. Among all genes tested fosB thus seems to be the most suitable marker for the detection of senescent cells. Stimulated senescent cells showed only a approximately twofold reduction of AP-1-binding activity, and senescent cells continuously exposed to serum exhibited normal AP-1-binding activity. These observations support the conclusion that a down-modulation of AP-1 is not crucial for human fibroblasts to enter the senescent state.