DNA Degradation In Vivo

Low levels of DNA degradation occur continuously in all cells, from bacteria to humans, in conjunction with various DNA repair processes. Probably 90% of this degradation can be attributed to excision repair, during which only a few nucleotides are released from each damaged site by the actions of repair endo and exonucleases and recycled in the overall process. Within a population of cells, high levels of DNA degradation occur in individual cells that need elimination from the population. This need occurs either because their DNA has become too heavily damaged to be successfully repaired, because the cells have been infected with bacteriophages or viruses or, during development and tissue remodeling in multicellular organisms, because these cells have received signals instructing them to commit suicide. In higher eukaryotes, this cell suicide is known as apoptosis. In this case, the chromatinDNA is fragmented and packaged with other cellular materials into apoptotic bodies that are engulfed by adjacent healthy cells in the tissue and digested in their lysosomes. The digestion products are recycled in these cells.

Most chromatin DNA fragmentation in eukaryotic cells follows an ordered path that yields first 300-kbp double-strand (ds) fragments, then 50-kbp ds fragments and lastly a range of small ds fragments that are multiples of nucleosome -sized DNA, 180–200 bp (1). This fragmentation pattern reflects the different levels of packing of the DNA in the chromatin. Six 50-kbp loops form 300-kbp rosettes that are stacked together. Cleavage between rosettes releases the 300-kbp fragments, and cleavage at the bases of the loops releases 50-kbp linear ds DNA that is readily cleaved between nucleosomes to small fragments that have 3′-OH and 5′-P termini. When separated by gel electrophoresis in agarose, the latter appear as a characteristic “ladder” of DNA. All stages of the fragmentation are Mg²⁺-dependent, but the subsequent cleavage to small fragments is also activated by Ca²⁺. This may indicate that there is more than one nuclease involved in chromatin DNA cleavage. In a few cases, the earlier stages of fragmentation are sufficient for apoptosis to proceed to completion. Random DNA degradation occurs during cell necrosis as a result of very heavy damage to the cell or tissue injury and is accompanied by lysosome disruption and cytolysis.

 Except for a small body of work on Escherichia coli, little is known about cell suicide in bacteria. When DNA bacteriophages that have not adapted to growth in E. coli are cleaved by host restriction enzyme nucleases, the resulting ds DNA fragments are degraded to small oligonucleotides by a nuclease with Mg²⁺- and ATP-dependent endo- and exonuclease activities encoded by the recB, recC, and recD genes, the recBCD nuclease. This nuclease normally acts in recombination and in recombinational ds break repair, a minor DNA repair pathway. Some bacteriophages that have adapted to E. coli contain genes that encode specific inhibitors of the recBCD nuclease. This nuclease is also responsible for degrading genomic DNA damaged beyond repair by UV light or ionizing radiation (2), but the details of the process are not known. The enzyme has homologues in many other species of bacteria.

 The identity of the nucleases that are responsible for chromatin DNA degradation during apoptosis in higher eukaryotes remains uncertain. Two difficulties encountered are that nuclei contain very low levels of active nucleases and that they are readily contaminated during isolation by nucleases from other organelles . The potent mitochondrial nuclease has been shown to contaminate nuclei isolated from calf thymus (3), a tissue that exhibits a high rate of apoptosis, although this nuclease is not known to act in chromatin DNA degradation. DNase II, an acid deoxyribonuclease (DNase(  normally found in the lysosomes, generates a ladder of ds DNA fragments when incubated with isolated nuclei at acidic pH. A small decrease in intracellular pH (0.3 pH units) occurs during apoptosis (4), but this would not be enough to result in appreciable activation of DNase II, even if it had a bona fide nuclear location. DNase II is not metal ion–dependent and makes ds breaks with 3′-P and 5′-OH groups, termini not found on the apoptotic ds DNA fragments. However, the corresponding acid DNase of the flatworm, Caenorhabditis elegans, does play a role in apoptosis, namely, in digesting the apoptotic bodies when engulfed by healthy cells. A loss of function of the nuclease caused by mutation of the nuc1 gene led to accumulation of the apoptotic bodies in the lysosomes (5).

 Considerable attention has been paid to DNase I as a possible candidate for chromatin DNA fragmentation (6). It is Mg²⁺-dependent and Ca²⁺-activated and makes the appropriate ds breaks in DNA. Where expressed, DNase I occurs in an inactive complex with actinthat is activated in vitro by proteolysis. However, the activity is not expressed in many cell types that undergo apoptosis and, when DNase I is added to isolated nuclei, a smear of randomly cleaved DNA results, rather than a ladder of DNA fragments. Overexpression of the bovine DNase I gene in monkey COS cells induced chromatin DNA ladder formation. However, expression of any nuclease activity in the nuclei may “damage” (nick or cleave) DNA sufficiently to trigger activation of the endogenous apoptotic program.

A bewildering array of other Ca²⁺, Mg²⁺-endonucleases, which vary in size from 18 to 97 kDa and generate ladders of DNA in isolated nuclei, have also been proposed to act in chromatin DNA degradation (7). These have been isolated from nuclei of apoptotic cells, but none have been fully characterized, no inactive forms have been reported, and little is known about their possible relationships. The smallest of these is NUC-18 (18 kDa), which has been identified as a cyclophilin (8). Cyclophilins have peptidyl prolyl cis–trans isomerase activity and play a role in protein folding. Since the specific nuclease activity associated with purified cyclophilins is very low, this activity may be associated with a contaminant. The largest (97 kDa) species cross-reacts with antibody raised against purified endo-exonuclease from Neurospora crassa (7), a Mg²⁺-dependent enzyme with homologues in other fungiand yeast that has been shown to act in recombination and in recombinational ds break repair (3). It may be the eukaryotic counterpart of the bacterial recBCD nuclease. It occurs in both active and inactive forms and has endonuclease activity with both DNA and RNA and exonuclease activity with DNA. A mammalian endo-exonuclease, synergistically activated by Ca²⁺, has been isolated from monkey CV-1 cells (9). It is the major degradative nuclease in nuclei of human leukemia cells, present entirely in inactive form. Proteolysis of endo-exonuclease has been detected by immunoblotting in response to different apoptotic agents and yielded polypeptides identical in sizes to the various Ca²⁺, Mg²⁺-endonucleases isolated from apoptotic cells by others. Proteolysis, an essential feature of apoptosis, may result in the activation and turnover of endo-exonuclease. Direct activation of endo-exonuclease, circumventing the apoptotic signaling pathways, could lead to new therapies in eliminating unwanted cells.

References

1. P. R. Walker, S. Pandey, and M. Sikorska (1995) Cell Death Differ. 2, 93–100. 

2. G. R. Smith (1988) Microbiol. Rev. 52, 1–28. 

3. M. J. Fraser and R. L. Low (1993) "Fungal and mitochondrial nucleases", in Nucleases, 2nd ed. R. J. Roberts, S. M. Linn, and S. Lloyd, eds., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, pp. 171–207.

4. M. A. Barry and A. Eastman (1993) Arch. Biochem. Biophys. 300, 440–450. 

5. J. Hevelone and P. S. Hartman (1988) Biochem. Genet. 26, 447–460. 

6. B. Polzar et al. (1993) Eur. J. Cell Biol. 62, 397–405. 

7. M. J. Fraser et al. (1996) J. Cell Sci. 109, 2343–2360. 

8. J. W. Montague, M. L. Gaido, C. Frye, and J. A. Cidlowski (1994) J. Biol. Chem. 269, 18877–18880.

9. C. Couture and T. Y.-K. Chow (1992) Nucl. Acids Res. 20, 1379–1385.