Objective. Dnase1-deficient mice with the 129 ؋ C57BL/6 genetic background develop symptoms of systemic lupus erythematosus, such as high titers of antinuclear autoantibodies directed against nucleosomes. In this study we analyzed a potential molecular pathomechanism leading to this autoimmunity, by exploring the influence of extracellular Dnase1 present in serum on the breakdown of chromatin in necrotic cells in vitro.Methods. Human breast adenocarcinoma cells (MCF-7) and other cell lines were subjected to necrosis induced by hydrogen peroxide, streptolysin O, or freezethawing. Subsequently, the influence of sera from Dnase1-deficient and wild-type mice as well as the influence of purified enzymes present in the culture medium on the process of necrotic chromatin breakdown was investigated.Results. Necrotic chromatin breakdown resembled apoptotic DNA laddering and was catalyzed by serum Dnase1 in conjunction with plasmin. During necrosis, Dnase1 and plasminogen penetrated the cell and accumulated in the cytoplasm and nucleus. Plasminogen bound to the cytoskeleton and nuclear structures, was activated to plasmin by either tissue-type or urokinase-type plasminogen activator, and degraded histone H1, thereby facilitating internucleosomal DNA cleavage by Dnase1.Conclusion. Our results suggest that serum Dnase1 in cooperation with the plasminogen system guarantees a fast and effective breakdown of chromatin during necrosis by the combined cleavage of DNA as well as of DNA binding proteins. The failure of such a clearance mechanism might lead to antinuclear autoimmunity similar to that observed in the Dnase1-deficient mouse.

Deoxyribonuclease 1 (DNASE1, DNase I) and deoxyribonuclease 1-like 3 (DNASE1L3, DNase gamma, DNase Y, LS-DNase) are members of a DNASE1 protein family that is defined by similar biochemical properties such as Ca2+/Mg2+-dependency and an optimal pH of about 7.0 as well as by a high similarity in their nucleic acid and amino acid sequences. In the present study we describe the recombinant expression of rat Dnase1 and murine Dnase1l3 as fusion proteins tagged by their C-terminus to green fluorescent protein in NIH-3T3 fibroblasts and bovine lens epithelial cells. Both enzymes were translocated into the rough endoplasmic reticulum, transported along the entire secretory pathway and finally secreted into the cell culture medium. No nuclear occurrence of the nucleases was detectable. However, deletion of the N-terminal signal peptide of both nucleases resulted in a cytoplasmic and nuclear distribution of both fusion proteins. Dnase1 preferentially hydrolysed 'naked' plasmid DNA, whereas Dnase1l3 cleaved nuclear DNA with high activity. Dnase1l3 was able to cleave chromatin in an internucleosomal manner without proteolytic help. By contrast, Dnase1 was only able to achieve this cleavage pattern in the presence of proteases that hydrolysed chromatin-bound proteins. Detailed analysis of murine sera derived from Dnase1 knockout mice revealed that serum contains, besides the major serum nuclease Dnase1, an additional Dnase1l3-like nucleolytic activity, which, in co-operation with Dnase1, might help to suppress anti-DNA autoimmunity by degrading nuclear chromatin released from dying cells.

It was previously shown (Paddenberg et al (1996) Eur J Cell Biol 69, 105 - 119) that cells of established lines like NIH3T3 fibroblasts and the human pancreatic adenocarcinoma PaTu 8902 line only degrade their chromatin at internucleosomal sites after an apoptotic stimulus when infected with Mycoplasma hyorhinis. In order to distinguish mycoplasma nucleases (Mr 47 - 54 kDa) from already described eukaryotic apoptotic enzymes, the mycoplasma nucleases were partially purified from serum-free culture supernatants and further characterized. Here we demonstrate directly that the enriched mycoplasma nucleases were able to fragment the DNA of nuclease-negative substrate nuclei at internucleosomal sites. The DNA degradation was accompanied by morphological changes typical of apoptosis like chromatin condensation and margination followed by shrinkage of the nuclei. The biochemical characterization revealed that the mycoplasma nucleases had a neutral to weakly basic pH-optimum. They required both calcium and magnesium in the mM range for maximal activation and were inhibited by zinc chloride, EGTA and EDTA. In two dimensional zymograms they migrated as three spots with isoelectic points between 8.1 and 9.5. They were not inhibited by monomeric actin. Our data also demonstrate that nuclear extracts prepared from nuclei isolated from Mycoplasma hyorhinis infected cells contained the mycoplasma nuclease activities leading to their internucleosomal DNA-degradation after incubation in the presence of calcium and magnesium.

The ability of two different Jurkat sublines, termed standard and JM, to form DNA ladders was investigated after various apoptotic stimuli. Exposure to a broad spectrum of drugs interfering with signal transduction or cellular metabolism revealed distinct differences between both Jurkat sublines with regard to the pattern of DNA degradation. In standard Jurkat cells, internucleosomal DNA cleavage occurred only after treatment with the protein kinase inhibitor staurosporine. In contrast, the JM subline responded with internucleosomal DNA fragmentation to exposure to gemcitabine, cycloheximide or staurosporine. All drugs induced the formation of DNA fragments of about 50 kb in both sublines, as revealed by pulse field electrophoresis, except H2O2, which caused unspecific DNA degradation. The staurosporine-induced DNA ladder formation was accompanied by an increase in caspase-3 activity in both lines which, however, was considerably lower in Jurkat JM cells after gemcitabine or cycloheximide exposure. When the analysis of internucleosomal DNA degradation was carried out after mycoplasma infection, both Jurkat lines responded with DNA ladder formation after exposure to all drugs used (here only shown for the standard subline). Employing the zymogram technique, nuclease activities of 47 kDa and 54 kDa were detected in culture supernatants, cell homogenates and nuclear extracts only when mycoplasma-infected, whereas the samples obtained from mycoplasma-free sublines were nuclease-negative using this technique, indicating that these endonucleases were of mycoplasmal origin. After drug exposure, the mycoplasmal nucleases must have gained access to the cytoplasm and nuclei of their host cells by an unknown mechanism.

No abstract