Regulation of poly(ADP-ribose) polymerase-1 functions by leukocyte elastase inhibitor/LEI-derived DNase II during caspase-independent apoptosis

Poly(ADP-ribose) polymerase-1 (PARP-1) is an important regulator of apoptosis. Its over-activation at the onset of apoptosis can inhibit the action of apoptotic endonucleases like caspase-activated DNase and DNAS1L3. Therefore, controlled PARP-1 proteolysis during caspase-dependent apoptosis is considered essential to promote DNA degradation. Yet, little is known about the interplay of PARP-1 and endonucleases that operate during caspase-independent cell death. Here we show that in the long-term cultured HeLa cells which undergo caspase-independent death, PARP-1 co-immunoprecipitates with leukocyte elastase inhibitor-derived DNase II (L-DNase II), an acid DNase implicated in this death pathway and activated by serine proteases. Our results indicate that, despite having putative poly(ADP-ribose)-acceptor sites, LEI/L-DNase II is neither significantly poly(ADP-ribosyl)ated nor inhibited by PARP-1 during caspase-independent apoptosis. Unexpectedly, caspase-independent apoptosis induced by hexa-methylene amiloride, LEI/L-DNase II can activate PARP-1 and promote its auto-poly(ADP-ribosyl)ation, thus inhibiting PARP-1 activity. Moreover, overexpression of LEI blocks the pro-survival effect of PARP-1 in this model of cell death. Our results provide the original evidence for a new mechanism of PARP-1 activity regulation in the caspase-independent death pathway involving LEI/L-DNase II.     

Conformational modification of serpins transforms leukocyte elastase inhibitor into an endonuclease involved in apoptosis

The best-characterized biochemical feature of apoptosis is degradation of genomic DNA into oligonucleosomes. The endonuclease responsible for DNA degradation in caspase-dependent apoptosis is caspase-activated DNase. In caspase-independent apoptosis, different endonucleases may be activated according to the cell line and the original insult. Among the known effectors of caspase-independent cell death, L-DNase II (LEI [leukocyte elastase inhibitor]-derived DNase II) has been previously characterized by our laboratory. We have thus shown that this endonuclease derives from the serpin superfamily member LEI by posttranslational modification (A. Torriglia, P. Perani, J. Y. Brossas, E. Chaudun, J. Treton, Y. Courtois, and M. F. Counis, Mol. Cell. Biol. 18:3612-3619, 1998). In this work, we assessed the molecular mechanism involved in the change in the enzymatic activity of this molecule from an antiprotease to an endonuclease. We report that the cleavage of LEI by elastase at its reactive center loop abolishes its antiprotease activity and leads to a conformational modification that exposes an endonuclease active site and a nuclear localization signal. This represents a novel molecular mechanism for a complete functional conversion induced by changing the conformation of a serpin. We also show that this molecular transformation affects cellular fate and that both endonuclease activity and nuclear translocation of L-DNase II are needed to induce cell death.     

Apoptosis induced by Na+/H+ antiport inhibition activates the LEI/L-DNase II pathway

L-DNase II is derived from its precursor leucocyte elastase inhibitor (LEI) by post-translational modification. In vitro, the conversion of LEI into L-DNase II can be induced by incubation of LEI at an acidic pH. In this study, we proposed to analyze the effects of intracellular acidification on this transformation. Amiloride derivatives, like hexamethylene amiloride (HMA), are known to provoke a decrease of cytosolic pH by inhibiting the Na(+)/H(+) antiport. In BHK cells, treatment with HMA-induced apoptosis accompanied by an increase in L-DNase II immunoreactivity and L-DNase II enzymatic activity. Overexpression of L-DNase II precursor led to a significant increase of apoptosis in these cells supporting the involvement of L-DNase II in HMA induced apoptosis. As previously shown in other cells, etoposide-induced apoptosis did not activate L-DNase. On the contrary, LEI overexpression significantly increased cell survival in etoposide-induced apoptosis. Together these results suggest differential roles of LEI and L-DNase II in response to different types of apoptotic inducers.     

Nuclear export of LEI/L-DNase II by Crm1 is essential for cell survival

LEI/L-DNase II is the key protein of a caspase-independent pathway activated by serine proteases. LEI (Leukocyte elastase inhibitor), L-DNase II precursor, is a member of the clade B serpins (also called serpin b1). In its native conformation it inhibits several intracellular proteases and has an anti-apoptotic activity. Following a metabolic stress and the increase of protease activity in the cell, LEI is cleaved and transformed into L-DNase II (LEI-derived DNase II). This transformation is due to a conformational modification that exposes a nuclear localization signal and an endonuclease active site. In this paper we show that LEI can bind the exportin Crm1, and we identify on LEI a nuclear export signal involved in the control of LEI/L-DNase II nuclearization in healthy cells. Point mutation of this site increases the accumulation of the molecule in the nucleus and triggers cell death.     

Acid DNases and their interest among apoptotic endonucleases

Apoptosis is characterized by cell shrinkage, nuclear condensation and internucleosomal DNA cleavage. Besides the central role of caspases and other proteases, cell death triggers DNA degradation so that DNases have an active role in apoptotic cell death. The best-characterized apoptotic DNase is CAD, a neutral Mg-dependent endonuclease. Its activity is regulated by its inhibitor, ICAD, which is cleaved by caspases. Other neutral DNases have been shown to cleave nuclear DNA in apoptotic conditions: endonuclease G, GADD. In cells, the cytosolic pH is maintained to 7.2, mostly due to the activity of the Na(+)/H(+) exchanger. In many apoptotic conditions, a decrease of the intracellular pH has been shown. This decrease may activate different acid DNases, mostly when pH decreases below 6.5. Three acidic DNases II are so far known: DNase II alpha, DNase II beta and L-DNase II, a DNase II, derived from the serpin LEI (Leukocyte Elastase Inhibitor). Their activation during cell death is discussed in this review.     

Nuclear translocation of a leukocyte elastase Inhibitor/Elastase complex during staurosporine-induced apoptosis: role in the generation of nuclear L-DNase II activity

Using L1210 murine leukemia cells, we have previously shown that in response to treatment with drugs having different targets, apoptotic cell death occurs through at least two different signaling pathways. Here, we present evidence that nuclear extracts from staurosporine-treated cells elicit DNase II activity that is not detected in nuclear extracts from cisplatin-treated cells. This activity correlates with the accumulation of two nuclear proteins (70 and 30 kDa) which are detected by an anti-L-DNase II antibody. Partial purification of this DNase II activity suggests that the 30-kDa protein could be the nuclease responsible for staurosporine-induced DNA fragmentation. The 70-kDa protein is also recognized by an anti-elastase antibody, suggesting that it carries residues belonging to both L-DNase II and elastase. Since previous findings showed that L-DNase II was generated from the leukocyte inhibitor of elastase, we propose that the 70-kDa protein results from an SDS-stable association between these two proteins and is translocated from the cytoplasm to the nucleus during staurosporine-induced apoptosis.     

On the use of an appropriate TdT-mediated dUTP–biotin nick end labeling assay to identify apoptotic cells

Apoptosis is an essential cellular mechanism involved in many processes such as embryogenesis, metamorphosis, and tissue homeostasis. DNA fragmentation is one of the key markers of this form of cell death. DNA fragmentation is executed by endogenous endonucleases such as caspase-activated DNase (CAD) in caspase-dependent apoptosis. The TUNEL (TdT-mediated dUTP–biotin nick end labeling) technique is the most widely used method to identify apoptotic cells in a tissue or culture and to assess drug toxicity. It is based on the detection of 3′-OH termini that are labeled with dUTP by the terminal deoxynucleotidyl transferase. Although the test is very reliable and sensitive in caspase-dependent apoptosis, it is completely useless when cell death is mediated by pathways involving DNA degradation that generates 3′-P ends as in the LEI/L-DNase II pathway. Here, we propose a modification in the TUNEL protocol consisting of a dephosphorylation step prior to the TUNEL labeling. This allows the detection of both types of DNA breaks induced during apoptosis caspase-dependent and independent pathways, avoiding underestimating the cell death induced by the treatment of interest.     

Interaction of Leukocyte Elastase Inhibitor/L-DNase II with BCL-2 and BAX

Leukocyte Elastase Inhibitor (LEI, also called serpin B1) is a protein involved in apoptosis among other physiological processes. We have previously shown that upon cleavage by its cognate protease, LEI is transformed into L-DNase II, a protein with a pro-apoptotic activity. The caspase independent apoptotic pathway, in which L-DNase II is the final effector, interacts with other pro-apoptotic molecules like Poly-ADP-Ribose polymerase (PARP) or Apoptosis Inducing Factor (AIF). The screening of LEI/L-DNase II interactions showed a possible interaction with several members of the BCL-2 family of proteins which are known to have a central role in the regulation of caspase dependent cell death. In this study, we investigated the regulation of LEI/L-DNase II pathway by two members of this family of proteins: BAX and BCL-2, which have opposite effects on cell survival. We show that, in both BHK and HeLa cells, LEI/L-DNase II can interact with BCL-2 and BAX in apoptotic and non-apoptotic conditions. These proteins which are usually thought to be anti-apoptotic and pro-apoptotic respectively, both inhibit the L-DNase II pro-apoptotic activity. These results give further insight in the regulation of caspase independent pathways and highlight the involvement of the intracellular environment of a given protein in the determinism of its function. They also add a link between caspase-dependent and independent pathways of apoptosis.     

Multiple apoptotic pathways induced by p53-dependent acidification in benzo[a]pyrene-exposed hepatic F258 cells

Polycyclic aromatic hydrocarbons (PAH), such as benzo[a]pyrene (B[a]P), are ubiquitous genotoxic environmental pollutants. Their DNA-damaging effects lead to apoptosis induction, through similar pathways to those identified after exposure to other DNA-damaging stimuli with activation of p53-related genes and the involvement of the intrinsic apoptotic pathway. However, at a low concentration of B[a]P (50 nM), our previous results pointed to the involvement of intracellular pH (pHi) variations during B[a]P-induced apoptosis in a rat liver epithelial cell line (F258). In the present work, we identified the mitochondrial F0F1-ATPase activity reversal as possibly responsible for pHi decrease. This acidification not only promoted executive caspase activation, but also activated leucocyte elastase inhibitor/leucocyte-derived DNase II (LEI/L-DNase II) pathway. p53 appeared to regulate mitochondria homeostasis, by initiating F0F1-ATPase reversal and endonuclease G (Endo G) release. In conclusion, a low dose of B[a]P induced apoptosis by recruiting a large panel of executioners apparently depending on p53 phosphorylation and, for some of them, on acidification.     

Differential involvement of DNases in HeLa cell apoptosis induced by etoposide and long term-culture

We have applied to human HeLa cells two different stimuli of apoptosis: the antitumoral drug etoposide, and a more 'physiological' death condition, obtained by growing cells in the same medium for long time periods, for up to 10 days. Analysis of different parameters demonstrated that in both experimental systems the same apoptotic features are visible. However, the DNA degradation pattern appeared to be different, suggesting the involvement of different DNases. In this view, we have analyzed the activity and expression of Ca2+-Mg2+-dependent and acid DNases. We have observed that DNase I is not modulated during apoptosis. In contrast, the acid L-DNase II (derived from Leukocyte Elastase Inhibitor by post-translational modification), recently identified in our laboratory, is mainly active in the apoptotic pathway induced by long term-culture. Furthermore, we have provided evidence that while caspase 3 is activated by both inducers, caspase 1 is essential only for the etoposide-induced apoptosis.     

DNases and apoptosis.

Here we review the different apoptotic DNases. From a functional point of view, DNases implicated in apoptosis may be classified into three groups: the Ca2+/Mg2+ endonucleases, the Mg2+-endonucleases, and the cation-independent endonucleases. The first group includes DNase I which has no specificity for the linker region, DNase gamma which has some homology with DNase I, and other DNases which cleave DNA in the linker region. Both DNase I and DNase gamma have been cloned. The other nucleases of this category have dispersed molecular weights. Their sequences are unknown and it is difficult to determine their role(s) in apoptosis. It seems that different pathways are present and that these nucleases may be activated either by caspases or serine proteases. The caspase 3 activated DNase (CAD, CPAN, or DFF40) belongs to the Mg2+-dependent endonucleases. DNase II belongs to the third group of acid endonucleases or cation-independent DNases. We have shown the involvement of DNase II in lens cell differentiation. Recently, the molecular structure of two different enzymes has been elucidated, one of which has a signal peptide and appears to be secreted. The other, called L-DNase II, is an intracellular protein having two enzymatic activities; in its native form, it is an anti-protease, and after posttranslational modification, it becomes a nuclease.     

Leukocyte Elastase Inhibitor, the precursor of L-DNase II, inhibits apoptosis by interfering with caspase-8 activation

LEI (Leukocyte Elastase Inhibitor), the precursor of the pro-apoptotic molecule L-DNase II, belongs to the ovalbumin subgroup of serpins. Several serpins can inhibit apoptosis: the viral serpin Crm A inhibits Fas or TNFalpha-induced apoptosis, and overexpression of PAI-2 or PI-9 protects cells from TNFalpha or granzyme B induced apoptosis. We have previously shown that LEI overexpression protects cells from etoposide-induced apoptosis. The molecular reason of this anti-apoptotic activity is now investigated. We show that, in BHK-21 and HeLa cells, LEI anti-protease activity is essential for its anti-apoptotic effect. The protease inhibited is cathepsin D, released from the lysosome during etoposide treatment. Cathepsin D enhances caspase activity in the cell by cleaving procaspase-8 and LEI overexpression slows down this cleavage, protecting cells from apoptosis. This let us presume that high expression of LEI in tumor cells may reduce the efficiency of etoposide as a chemotherapeutic agent.     

The activation of the atypical PKC zeta in light‐induced retinal degeneration and its involvement in L‐DNase II control

Light‐induced retinal degeneration is characterized by photoreceptor cell death. Many studies showed that photoreceptor demise is caspase‐independent. In our laboratory we showed that leucocyte elastase inhibitor/LEI‐derived DNase II (LEI/L‐DNase II), a caspase‐independent apoptotic pathway, is responsible for photoreceptor death. In this work, we investigated the activation of a pro‐survival kinase, the protein kinase C (PKC) zeta. We show that light exposure induced PKC zeta activation. PKC zeta interacts with LEI/L‐DNase II and controls its DNase activity by impairing its nuclear translocation. These results highlight the role of PKC zeta in retinal physiology and show that this kinase can control caspase‐independent pathways.     

Utilization of an in vitro assay to evaluate chromatin degradation by candidate apoptotic nucleases

Apoptosis is commonly associated with the catabolism of the genome in the dying cell. The chromatin degradation occurs in essentially two forms: (1) internucleosomal DNA cleavage to generate oligonucleosomal-length fragments (180-200 bp and multiples thereof), and (2) cleavage of higher order chromatin structures to generate approximately 30-50 Kb fragments. To investigate this component of apoptosis and identify the nuclease(s) responsible, we have developed and utilized an in vitro assay that recapitulates the genomic destruction seen during apoptosis in vivo and allows the simultaneous analysis of both forms of DNA degradation from the same sample. Using this assay we evaluated the digestion patterns of several candidate apoptotic nucleases: DNase I, DNase II, and cyclophilin (NUC18) as well as the bacterial enzyme micrococcal nuclease (not thought to be involved in apoptosis). Chromatin degraded by DNase I formed a smear of DNA on conventional static-field agarose gels and approximately amp;30 - 50 Kb DNA fragments on pulsed field gels. In contrast, DNase II, at a physiologically relevant pH, had no effect on the integrity of HeLa chromatin in either analysis. Similar to DNase I, cyclophilin C produced only approximately 30-50 Kb DNA fragments but did not generate internucleosomal fragments. In contrast, micrococcal nuclease generated both oligonucleosomal and approximately 30-50 Kb DNA fragments. Nuclear extracts from glucocorticoid-treated apoptotic thymocytes generated oligonucleosomal DNA fragments and the larger approximately 30-50 Kb DNA fragments, fully recapitulating both types of apoptotic DNA degradation. Previously, differential sensitivity of nucleases to inhibition by Zn2+ was used to argue that two distinct enzymes mediate approximately 30-50 Kb DNA cleavage and internucleosomal DNA degradation. While, the nuclease activity present in thymocyte nuclear extracts was differentially sensitive to inhibition by Zn2+ during short term incubations it was not during prolonged digestions, suggesting that differences in DNA detection are likely to account for previous results. Together our studies show that none of the nucleases commonly associated with apoptosis could fully recapitulate the DNA degradation seen in vivo.     

The chromatin of differentiating erythroblasts is cleaved into large size fragments independent of caspase activated DNase and apoptosis inducing factor

Erythroblast cell differentiation involves self-controlled and limited nuclear proteolysis prior nucleus loss. Early evidence suggests that apoptotic-like pathways are activated during this process. The chromatin of developing erythroblasts becomes fragmented in vivo, however, the exact mechanisms and molecules involved remain elusive. In this study, erythroblasts were differentiated in culture from CD34-enriched umbilical cord blood progenitor cells and the characteristics of DNA fragmentation were examined. This analysis shows that the chromatin of differentiating erythroblasts is cleaved into discrete fragments of 50-200 kb. This process most likely involves one or several endonucleases as we detect in vivo double strand DNA cleavage. However, major players of the apoptotic DNA degradation, caspase activated DNase and apoptosis inducing factor, are not activated in these cells. Therefore, our data suggests that erythroblast chromatin degradation may involve enzymes distinct form those active in apoptotic cells.     

Degradation of chromosomal DNA during apoptosis

Apoptosis is often accompanied by degradation of chromosomal DNA. CAD, caspase-activated DNase, was identified in 1998 as a DNase that is responsible for this process. In the last several years, mice deficient in the CAD system have been generated. Studies with these mice indicated that apoptotic DNA degradation occurs in two different systems. In one, the DNA fragmentation is carried out by CAD in the dying cells and in the other, by lysosomal DNase II after the dying cells are phagocytosed. Several other endonucleases have also been suggested as candidate effectors for the apoptotic degradation of chromosomal DNA. In this review, we will discuss the mechanism and role of DNA degradation during apoptosis.     

DNase II and the Chk2 DNA damage pathway form a genetic barrier blocking replication of horizontally transferred DNA

We have previously shown that DNA from dying tumor cells may be transferred to living cells via the uptake of apoptotic bodies and may contribute to tumor progression. DNA encoding H-ras(V12) and c-myc oncogenes may be transferred to the nucleus of the phagocyte but will only integrate and propagate in p53- and p21-deficient mouse embryonic fibroblasts, whereas normal cells are resistant to transformation. Here, we show that this protective mechanism (activation of p53 and p21 after uptake of apoptotic bodies) is dependent on DNA fragmentation, where inhibition of the caspase-activated DNase in the apoptotic cells, in conjunction with genetic ablation of lysosomal DNase II in the phagocytes, completely blocks p53 activation and consequently allows DNA replication of transferred DNA. We, therefore, suggest that there is a causal relationship between DNA degradation during apoptosis and p53 activation. In addition, we could further show that Chk2-/- cells were capable of replicating the hyg(R) gene taken up from engulfed apoptotic cells, suggesting involvement of the DNA damage response. These data show that the phagocytosing cell is sensing the degraded DNA within the apoptotic cell, hence preventing these genes from being replicated, probably through activation of the DNA damage response. We, therefore, hypothesize that DNase II together with the Chk2, p53, and p21 pathway form a genetic barrier blocking the replication of potentially harmful DNA introduced via apoptotic bodies, thereby preventing transformation and malignant development.     

Coordinated and sequential activation of neutral and acidic DNases during interdigital cell death in the embryonic limb

Interdigital tissue regression during embryonic development is one of the most representative model systems of morphogenetic cell death, but the degenerative cascade accounting for this process awaits clarification. Although the canonical apoptotic caspase pathway appears to be activated in the interdigital mesenchyme committed to die, neither genetic nor chemical blockage of caspases or their downstream effectors, is sufficient to prevent cell death. Hence, alternative and/or complementary dying pathways must also be responsible for this degenerative process. In this work we have chosen to study the endonucleases during the regression of the interdigital tissue of avian embryos to gain insights into the molecular mechanisms accounting for programmed cell death in this system. We show that caspase activated DNase, which is a neutral DNase associated with the caspase apoptotic pathway, appears to be the main endonuclease only at an initial phase of interdigit regression. However at peak stages of the degenerative process, the acidic DNases L-DNase II and lysosomal DNase IIB become predominant in the system and markers for cell autophagy become moderately up-regulated. Consistent with the activation of acidic endonucleases we observed that microenvironmental pH value in the interdigits decreased to levels only appropriate for acidic enzymes. Furthermore, we found that overexpression of lysosomal DNase IIB in embryonic limb mesoderm promoted cell death, which was also accompanied by up-regulation and activation of L-DNase II. Up-regulation of acidic DNases was maintained in interdigits explanted to culture dishes, where the participation of exogenous professional phagocytes of hematopoietic origin is avoided. Finally, and consistent with all our findings, up-regulation of acidic DNases was much reduced in the webbed interdigits of duck embryos, characterized by a rudimentary interdigital degenerative process. We conclude that the regression of the interdigital tissue involves a coordinated and sequential activation of the caspase and lysosomal degenerative molecular cascades.     

Biogenesis and Proteolytic Processing of Lysosomal DNase II

Deoxyribonuclease II (DNase II) is a key enzyme in the phagocytic digestion of DNA from apoptotic nuclei. To understand the molecular properties of DNase II, particularly the processing, we prepared a polyclonal antibody against carboxyl-terminal sequences of mouse DNase II. In the present study, partial purification of DNase II using Con A Sepharose enabled the detection of endogenous DNase II by Western blotting. It was interesting that two forms of endogenous DNase II were detected – a 30 kDa form and a 23 kDa form. Neither of those forms carried the expected molecular weight of 45 kDa. Subcellular fractionation showed that the 23 kDa and 30 kDa proteins were localized in lysosomes. The processing of DNase II in vivo was also greatly altered in the liver of mice lacking cathepsin L. DNase II that was extracellularly secreted from cells overexpressing DNase II was detected as a pro-form, which was activated under acidic conditions. These results indicate that DNase II is processed and activated in lysosomes, while cathepsin L is involved in the processing of the enzyme.     

DNase 2 is the main DNA-degrading enzyme of the stratum corneum

The cornified layer, the stratum corneum, of the epidermis is an efficient barrier to the passage of genetic material, i.e. nucleic acids. It contains enzymes that degrade RNA and DNA which originate from either the living part of the epidermis or from infectious agents of the environment. However, the molecular identities of these nucleases are only incompletely known at present. Here we performed biochemical and genetic experiments to determine the main DNase activity of the stratum corneum. DNA degradation assays and zymographic analyses identified the acid endonucleases L-DNase II, which is derived from serpinB1, and DNase 2 as candidate DNases of the cornified layer of the epidermis. siRNA-mediated knockdown of serpinB1 in human in vitro skin models and the investigation of mice deficient in serpinB1a demonstrated that serpinB1-derived L-DNase II is dispensable for epidermal DNase activity. By contrast, knockdown of DNase 2, also known as DNase 2a, reduced DNase activity in human in vitro skin models. Moreover, the genetic ablation of DNase 2a in the mouse was associated with the lack of acid DNase activity in the stratum corneum in vivo. The degradation of endogenous DNA in the course of cornification of keratinocytes was not impaired by the absence of DNase 2. Taken together, these data identify DNase 2 as the predominant DNase on the mammalian skin surface and indicate that its activity is primarily targeted to exogenous DNA.