Autoantibodies to endostatin in patients with breast cancer: correlation to endostatin levels and clinical outcome.

Circulating autoantibodies to self-antigens overexpressed by cancer cells are common in cancer patients. As specific proteins are expressed during neoangiogenesis, a similar phenomenon might occur with particular antigens of tumour vessels. Collagen XVIII, from which endostatin is cleaved, is highly expressed in the perivascular basement membrane of tumour-associated blood vessels and autoantibodies to endostatin have been reported in cancer patients. The present study analyses the incidence of naturally occurring autoantibodies to endostatin in the sera of breast cancer patients and their relation to endostatin serum levels and patient clinical outcome. Serum samples from 36 patients with localised breast cancer and 59 patients with a fully documented history of metastatic breast cancer were used. The immunoreactivity of serum samples was tested against purified recombinant human endostatin and endostatin levels were determined by immunoassay. We could detect anti-endostatin antibodies in the sera of 66% of the patients with localised disease and 42% of the patients with metastatic disease (P=0.03). There was no correlation between the presence of antibodies to endostatin and circulating levels of endostatin. The detection of autoantibodies to endostatin was associated with better prognosis in metastatic breast cancer patients (median survival time: 20 vs 8 months, P = 0.03), as was the presence of low levels of serum endostatin (median survival time: 20 vs 9 months, P = 0.007). These results show that a natural immune reaction against endostatin can occur in breast cancer patients. This could have important therapeutic implications with regard to endostatin therapy and raises the question of a possible role of this humoral reaction against endostatin in the neoplastic process.

The bacterial nucleoside N(6)-methyldeoxyadenosine induces the differentiation of mammalian tumor cells.

Contrary to bacterial DNA, mammalian DNA contains very little if any N(6)-methyldeoxyadenosine (MDA). The possible biological effect of this nucleoside on eukaryotic cells has been studied on different tumor cell lines. Addition of MDA to C6.9 glioma cells triggers a differentiation process and the expression of the oligodendroglial marker 2',3'-cyclic nucleotide 3'phosphorylase (CNP). The biological effects of N(6)-methyldeoxyadenosine were not restricted to C6.9 glioma cells since differentiation was also observed on pheochromocytoma and teratocarcinoma cell lines and on dysembryoplastic neuroepithelial tumor cells. The precise mechanism by which MDA induces cell differentiation remains unclear, but is related to cell cycle modifications. These data point out the potential interest of N(6)-methyldeoxyadenosine as a novel antitumoral and differentiation agent. They also raise the intriguing question of the loss of adenine methylation in mammalian DNA. Furthermore, the finding that a methylated nucleoside found in bacterial DNA induces a biological process might have implications in gene therapy approaches when plasmid DNAs are injected into humans.

Epidermal growth factor modulates the expression of vascular endothelial growth factor in the human prostate.

The growth and dissemination of tumors in the body has been associated with angiogenesis. Vascular endothelial growth factor (VEGF) is an angiogenic factor that stimulates endothelial cell growth and enhances vascular permeability. VEGF exerts its action by binding to specific cell surface receptors. Three receptors, VEGFR-1 (flt-1), VEGFR-2 (flk-1), and VEGFR-3 (flt-4) have been identified. Very little information on the coordinated expression of VEGF and its receptors in normal prostate, benign prostatic hyperplasia (BPH), and prostate carcinoma is available. Therefore, we examined the immunohistochemical localization of VEGF and its receptors in tissues derived from normal human prostate, BPH, and prostatic carcinoma. Immunostaining for VEGF was absent in the normal prostate. Epithelium lining the glands of prostate derived from patients with BPH exhibited strong immunostaining. The intensity of staining was relatively less in prostate carcinoma. It is interesting that VEGFR-1 and VEGFR-3 were strongly expressed in both stromal and epithelial tissues in normal prostate, BPH, and carcinoma. In comparison, VEGFR-2 was not localized to normal prostate and its expression in the stroma of BPH and epithelium of carcinoma was very weak. Because progression of prostate cancer is accompanied by altered expression of epidermal growth factor (EGF) and its receptor (EGFR) in malignant cells, we investigated the effect of EGF on VEGF gene expression by Northern blot analysis in 2 human prostate cancer cell lines that express EGFR. EGF greatly enhanced the expression of VEGF messenger RNA in DU145 and PC3 cell lines in a dose-dependent manner. The EGF induction of VEGF gene expression suggests a mechanism by which angiogenesis could be accelerated in BPH and prostate carcinoma.

p53 Status and gene transfer experiments using CMV enhancer/promoter.

Comparison of transfection efficiencies between different commercial reagents or methods is a matter of major concern in the field of gene therapy. Transfection efficiencies are usually evaluated by the quantification of a reporter gene expression (i.e., luciferase or lacZ) whose expression is usually driven by the CMV promoter. However, this experimental approach does not consider the possible effects of the transfection on the activity of the promoter used to drive reporter genes expression. Using p53 null fibroblasts we show that transfection efficiency estimated by the use of pCMV-luc or pCMV-betagal plasmids may be dramatically affected by the cell p53 status. These data highlight the fact that differences in p53 levels may be one of the parameters involved in the variation of transfection efficiencies observed with different cell lines. Furthermore, they point to the fact that comparison of transfection efficiencies should distinguish differences in the efficiency of transfection from differences in the level of transcription of the transgene. Finally they suggest that the known p53 down-regulation of the CMV promoter should be considered in order to avoid the nonintentional construction of transfer vectors in which the expression of a transgene down-modulates its own promoter.

An enzymatic procedure for the purification of DNA restriction fragments without gel electrophoresis and ethidium bromide staining.

A rapid and simple enzymatic method for the purification of a DNA fragment from a restriction digest was developed. The method is based on the two features of exonuclease III activity: digestion of DNA from a 3'-OH at blunt or recessed ends and failure to initiate digestion at DNA ends with four-base 3' overhangs. Herein, we establish a method for purification of a DNA restriction fragment without any physical separation via gel electrophoresis. The elimination of the ethidium bromide staining and ultraviolet irradiation steps should increase the quality and the safety of the purified DNA, a matter of major concern in the perspective of human gene therapy. In addition, since the method described does not use the visualization of the restriction fragments or their difference in size it can be used to purify a DNA fragment from a pool of DNA fragments with the same size even when microquantities of material are available.

Bacterial DNA methylation and gene transfer efficiency.

The necessary amplification step in bacteria of any plasmid currently used in DNA immunization or gene therapy introduces modification in the nucleotide sequence of plasmid DNA used in gene transfer. These changes affect the adenine and the internal cytosine in respectively all of the GATC and CC(A/T)GG sequences. These modifications which introduce 6-methyladenine and 5-methylcytosine in plasmidic DNA are the consequence of the existence of the bacterial modification systems Dam and Dcm. In eucaryotes, the presence of 5-methylcytosine at dinucleotides -CG- is involved in silencing gene expression, but the possible consequences of the presence of the bacterial G(m)ATC and C(m)C(A/T)GG sequences in the plasmids used in gene transfer experiments are presently unknown. Since the possibility exists to obtain plasmid DNA lacking this specific bacterial pattern of methylation by using (dam(-), dcm(-)) bacteria we performed experiments to compare in vitro and in vivo gene transfer efficiency of a pCMV-luc reporter plasmid amplified either in the JM109 (dam(+), dcm(+)) or JM110 (dam(-), dcm(-)) bacteria. Data obtained demonstrated that the presence of 6-methyladenine in GATC sequences and 5-methylcytosine in the second C of CC(A/T)GG motifs does not reduce the levels of luciferase activity detected following in vitro or in vivo gene transfer. On the contrary, gene transfer with a pCMV-luc amplified in JM109 (dam(+), dcm(+)) bacteria gives greater amounts of luciferase than the same transfection performed with a plasmid amplified in the mutated JM110 (dam(-), dcm(-)) counterpart. Therefore, these data do not suggest that the use of (dam(-), dcm(-)) bacteria to amplify plasmid DNA may increase gene transfer efficiency. However, the persistence of the use of (dam(+), dcm(+)) bacteria in order to amplify plasmid DNA raises the question of the possible biological consequences of the introduction of the bacterial G(m)ATC and C(m)C(A/T)GG sequences in eukaryotic cells or organisms.

Muscle transfection by electroporation with high-voltage and short-pulse currents provides high-level and long-lasting gene expression.

Gene transfer into muscle by electroporation with low-voltage and long-pulse (LV/LP, 100 V/50 msec) currents was shown to be more efficient than simple intramuscular DNA injection. Nevertheless, transgene expression declined from day 7 and only reached 10% of the maximum 3 weeks after electroporation. We have optimized electroporation conditions including voltage, pulse number, and the amount of injected luciferase-encoding plasmid DNA in the tibialis anterior muscle. Using high-voltage and short-pulse (HV/SP, 900 V/100 microsec) currents, we observed an average 500-fold increase in luciferase expression, in comparison with nonelectroporated muscle. Moreover, sustained and long-lasting gene expression was observed for at least 6 months. When we compared HV/SP currents with LV/LP currents, luciferase expression was similar 24 hr after electroporation. One month later, whereas luciferase expression was stable in muscle electroporated with HV/SP currents, it decreased 600-fold in muscle electroporated with LV/LP currents. In conclusion, electroporation with high-voltage and short-pulse currents provides high-level and long-lasting gene expression in muscle.

Mycoplasma: a new potential vector for gene therapy?

The introduction of foreign genetic material in living cells is the basis of the current protocols of gene therapy. The concept that the de novo synthesis of a foreign therapeutic protein requires the entrance of the corresponding gene into target cells via virus or synthetic vectors is directly inherited from experiments on bacterial transduction or transformation. However, the difficulties inherent in the penetration and the expression of foreign DNA into eucaryotic cells are probably responsible for the low efficiency of this therapeutic approach. In this paper, we explore the possibility of avoiding these limiting critical steps by expressing the foreign gene on the surface rather than inside the target cells by the use of mycoplasma, the smallest reported living cell. The absence of transfer of genetic information between this vector and eucaryotic target cells, the sensibility of mycoplasmas to antibiotics and their cytadherance are among the interesting features of this potential vector. The interest of this new vector in the case, e.g. of the gene-directed enzyme prodrug therapy of solid tumors, is discussed.

Epigenetic control of programmed cell death: inhibition by 5-azacytidine of 1,25-dihydroxyvitamin D3-induced programmed cell death in C6.9 glioma cells.

In mammalian DNA cytosine methylation occurs specifically at CpG dinucleotide. Although the full array of function of DNA methylation is yet to be elucidated, it is well established that DNA methylation is an important mechanism involved in gene expression, DNA replication and cancer. Rat glioma C6.9 cells undergo programmed cell death (PCD) after treatment with 1,25-dihydroxyvitamin D3 (1,25-D3). Hence, these cells were used to study whether DNA methylation was involved in the control of PCD. We found that 1,25-D3-mediated PCD of C6.9 cells was suppressed by exposure of the cells to the DNA demethylating agents 5-azacytidine (5-AzaC) and 5-aza-2'-deoxycytidine. This effect remains detectable several cell divisions following removal of 5-AzaC and, therefore, involves DNA methylation as an epigenetic regulatory mechanism of PCD. Accordingly, internucleosomal fragmentation, a feature of apoptosis that is detected in 1,25-D3-treated cells, is no longer observable after treatment of these cells with 5-AzaC. However, 5-AzaC does not totally suppress the responsiveness of C6.9 cells to 1,25-D3 since the induction of the c-myc gene remains unaffected. These results suggest that a change in DNA methylation pattern could suppress 1,25-D3-mediated PCD through the expression of previously hypermethylated genes such as proto-oncogenes with death-repressor activity, endogenous virus sequences or even genes inducing change in the differentiated state of these cells.

Vitamin D receptor stable transfection restores the susceptibility to 1,25-dihydroxyvitamin D3 cytotoxicity in a rat glioma resistant clone.

Recently, 1,25-dihydroxyvitamin D3 (1,25-D3) and less hypercalcemic analogs were shown to exert a delayed cytotoxic effect on rat C6 glioma cells. 1,25-D3 induces in these cells a programmed cell death, accompanied by the induction of c-myc, p53 and gadd 45 genes. The involvement of the intracellular vitamin D receptor (VDR) remained to be determined. In this lethal process, we have investigated its role in a subclone of C6 cells, which was isolated on the basis of its resistance to 1,25-D3, and in which VDR expression was not detected either at the mRNA or protein levels. The stable transfection of a rat VDR cDNA into this clone restored its susceptibility to the cytotoxic effects of 1,25-D3. This phenomenon was accompanied by a dramatic upregulation of c-myc mRNA expression, as already described in a C6-sensitive clone. These results provide the first evidence that VDR expression, if not sufficient, is necessary to mediate 1,25-D3 cytotoxic effect in C6 glioma cells. Since VDR mRNA expression has been already reported in human brain tumors, our data imply that the identification of VDR expression could become a prerequisite in any strategy of glioma treatment with vitamin D analogs.

Differentially expressed genes in C6.9 glioma cells during vitamin D-induced cell death program.

C6.9 rat glioma cells undergo a cell death program when exposed to 1, 25-dihydroxyvitamin D3 (1,25-D3). As a global analytical approach, we have investigated gene expression in C6.9 engaged in this cell death program using differential screening of a rat brain cDNA library with probes derived from control and 1,25-D3-treated cells. Using this methodology we report the isolation of 61 differentially expressed cDNAs. Forty-seven cDNAs correspond to genes already characterized in rat cells or tissues. Seven cDNAs are homologous to yeast, mouse or human genes and seven are not related to known genes. Some of the characterized genes have been reported to be differentially expressed following induction of programmed cell death. These include PMP22/gas3, MGP and beta-tubulin. For the first time, we also show a cell death program induced up-regulation of the c-myc associated primary response gene CRP, and of the proteasome RN3 subunit and TCTP/mortalin genes. Another interesting feature of this 1,25-D3 induced-cell death program is the down-regulated expression of transcripts for the microtubule motor dynein heavy chain/MAP 1C and of the calcium-binding S100beta protein. Finally 15 upregulated cDNAs encode ribosomal proteins suggesting a possible involvement of the translational apparatus in this cell program. Alternatively, these ribosomal protein genes could be up-regulated in response to altered rates of cellular metabolism, as has been demonstrated for most of the other isolated genes which encode proteins involved in metabolic pathways. Thus, this study presents to our knowledge the first characterization of genes which are differentially expressed during a cell death program induced by 1, 25-D3. Therefore, this data provides new information on the fundamental mechanisms which participate in the antineoplastic effects of 1,25-D3 and on the machinery of a cell death program in a glioma cell line.

Was the formation of 1,25-dihydroxyvitamin D3 initially a catabolic pathway?

Following solar ultraviolet radiation, epidermal 7-dehydrocholesterol is converted to previtamin D3, which then undergoes a thermal isomerization into vitamin D3. The metabolism of vitamin D3, which is usually considered as an inactive compound, gives rise to the active hormone 1,25-dihydroxyvitamin D3, following two hydroxylation steps occurring in liver and kidney. Here, we propose that this anabolic pathway can also be interpreted as a catabolic one leading to the degradation of the photoproducts of 7-dehydrocholesterol, for which a specific biological role in the skin is proposed.

Noradrenaline inhibits the programmed cell death induced by 1,25-dihydroxyvitamin D3 in glioma.

The rat glioma cell line C6.9 has been recently reported to respond to 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) by the induction of a programmed cell death. Since, in vivo, glial cells are thought to be exposed to several neurotransmitters, we investigated the possibility of a neurotransmitter-mediated inhibition of this active cell death process. Noradrenaline and the beta-adrenoceptor agonist isoproterenol showed significant inhibition of the 1,25(OH)2D3-induced programmed cell death. The beta-adrenoceptor antagonist propanolol reversed this inhibition, while the alpha-adrenoceptor antagonist yohimbin was devoid of any effect. This suggests that the efficiency of antiproliferative vitamin D-related therapies could be influenced by endogenous levels of noradrenaline.