- Email: [email protected]
Influence of DNA Binding on the Degradation of Oxidized Histones by the 20S Proteasome
Archives of Biochemistry and Biophysics Vol. 362, No. 2, February 15, pp. 211–216, 1999 Article ID abbi.1998.1031, available online at http://www.idealibrary.com on
Influence of DNA Binding on the Degradation of Oxidized Histones by the 20S Proteasome Oliver Ullrich,* Nicolle Sitte,* Olaf Sommerburg,* Volker Sandig,† Kelvin J. A. Davies,‡ and Tilman Grune* ,1 *Clinics of Physical Medicine and Rehabilitation, University Hospital Charite´, Humboldt-University, Berlin, Germany; †Max Delbrueck Center for Molecular Medicine, Berlin, Germany; and ‡Ethel Percy Andrus Gerontology Center, University of Southern California, Los Angeles, California 90089-0191
Received July 16, 1998, and in revised form October 6, 1998
The 20S proteasome is localized in the cytosol and nuclei of mammalian cells. Previous work has shown that the cytosolic 20S proteasome is largely responsible for the selective recognition and degradation of oxidatively damaged cytosolic proteins. Since nuclear proteins are also susceptible to oxidative damage (e.g., from metabolic free radical production, ionizing radiation, xenobiotics, chemotherapy) we investigated the degradation of oxidatively damaged histones, in the presence and in the absence of DNA, by the 20S proteasome. We find that both soluble histones and DNAbound histones are susceptible to selective proteolytic degradation by the 20S proteasome following mild oxidative damage. In contrast, more severe oxidative damage actually decreases the proteolytic susceptibility of histones. Soluble H1 showed the highest basal and maximal absolute proteolytic rates. Histone fraction H4 exhibited the greatest relative increase in proteolytic susceptibility following oxidation, almost 14fold, and this occurred at a peroxide exposure of 5 mM. At the other end of the spectrum, histone H2A exhibited a maximal proteolytic response to H 2O 2 of only 6-fold, and this required an H 2O 2 exposure of 15 mM. An oxidation of reconstituted linear DNA plasmid– histone complex makes up to 95% of the histones bound to DNA susceptible to degradation, whereas undamaged protein–DNA complexes are not substrates for the proteasome. Severe oxidation by high concentrations of H 2O 2 appears to decreases the proteolytic susceptibility of histones due to the formation of crosslinked histone–DNA aggregates which appear to inhibit the proteasome. We conclude that the degradation of nuclear proteins is highly selective and
1 To whom correspondence should be addressed: Clinics of Physical Medicine and Rehabilitation, University Hospital Charite´, HumboldtUniversity Berlin, Schumannstrasse 20/21, 10098 Berlin, Germany. Fax: 149 30 2093 7204. E-mail: [email protected].
0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
requires prior damage of the substrate protein, such as that caused by oxidation. © 1999 Academic Press Key Words: protein oxidation; histones; proteasome; DNA–protein aggregates.
Nuclear proteins are oxidized by reaction with oxygen radicals and other activated oxygen species, generated as by-products of cellular metabolism, or by exogenous sources, e.g., ionizing radiation or antitumor chemotherapy (1, 2). Little is known of the fate of such oxidatively modified nuclear proteins. In the cell cytoplasm most oxidatively damaged proteins undergo selective proteolysis, via an ATP- and ubiquitin-independent pathway (3). The 20S proteasome (4), a large 700-kDa multicatalytic, multisubunit proteinase, is largely responsible for the selective degradation of such oxidatively damaged cytoplasmic proteins (5, 6). Importantly, the proteasome is also found in high concentrations in the nucleus of mammalian cells (4), although its nuclear function is not known. Oxidative stress in the nucleus may contribute to histone modifications such as amino acid oxidation, protein–protein cross-linking, and DNA–protein crosslinking. We propose that selective proteolytic degradation of damaged and functionally altered long-lived histones may be essential for genomic integrity. Nucleosomal histones protect DNA from free radical mediated damage (7), and histone detachment and reattachment are closely connected with transcription and replication processes as with DNA repair and therefore requires functionally intact histones. Oxidatively damaged histones are able to cross-link with DNA (8, 9) and would impair the detachment-reassembly process. We, therefore, studied the ability of the 20S proteasome to degrade damaged histones as a function of the extent of their oxidation. Because histone modification 211
212
ULLRICH ET AL.
after oxidative stress may also change the DNA-binding properties of the histones, we also tested the influence of DNA on histone (protein) degradation by the proteasome. MATERIALS AND METHODS
Preparation of the Proteasome Proteasome was isolated according to Hough et al. (10). Cells were lysed in 1 mM dithiothreitol. After removal of membranes and nonlysed cells by centrifugation the proteasome was purified by DEAE chromatography, sucrose density gradient ultracentrifugation, and separation on a Mono Q column using an FPLC system. The purity of the preparation was controlled by both one-dimensional SDS– PAGE and nondenaturing polyacrylamide electrophoresis (10). Nondenaturing electrophoresis was followed by overlaying the gel with a 200 mM sLLVY-MCA solution for detection of peptidase activity. The proteasome preparation was diluted finally to a protein concentration of 0.35 mg/ml.
Radioactive Labeling of Isolated Histones Isolated histones were radiolabeled by reductive methylation with [ 3H]formaldehyde and sodium cyanoborohydride as described by Jentoft and Dearborn (11) and then dialyzed extensively.
Treatment of Histones with Hydrogen Peroxide Isolated histones or isolated histone fractions H1, H2A, H2B, H3, and H4 (2 mg/ml) were oxidatively damaged using various concentrations of H 2O 2 in 20 mM phosphate buffer (pH 7.4) for 2 h at 25°C. Aliquots (400 ml) were then dialyzed for 16 h at 4°C, against 2.0 liters of 10 mM phosphate buffer containing 10 mM KCl, with one exchange of dialysis fluid after 3 h.
of incubated substrate protein without protease and of incubated protease only). Degradation of 3H-labeled histones. Degradation of 3H-labeled histones by proteasome was measured in 50 mM Tris–HCl (pH 7.8), 20 mM KCl, 5 mM MgOAc, and 0.5 mM dithiothreitol. Percentage of histone degradation was determined by liquid scintillation in supernatants of trichloroacetic acid (TCA)-precipitated proteins and calculated (12) according to the following formula: (TCA-soluble radioactivity 2 background radioactivity)/(total radioactivity 2 background radioactivity) 3 100. Degradation of DNA-bound histones. Proteolysis experiments were performed by adding of 0.6 mg proteasome in 50 mM Tris–HCl (pH 7.8), 20 mM KCl, 5 mM MgOAc, and 0.5 mM dithiothreitol for 2 h at 37°C to 2 mg radiolabeled histones. Controls were incubated without proteasome. Degradation of suc-LLVY-MCA. Degradation of the fluorogenic peptide suc-LLVY-MCA was measured as follows: to 270 ml reaction mixture containing the proteasome, 30 ml of the fluorogenic peptide suc-LLVY-MCA (2 mM stock solution in DMSO) was added. After incubation at 37°C for 1 h the reaction was stopped by addition of an equal volume of ice-cold ethanol and 10 vol of 125 mM borate buffer, pH 9.0. Fluorescence was determined at 380 nm excitation/440 nm emission.
Agarose Gel Electrophoresis DNA– histones were analyzed in a 0.8% agarose gel, 90 mM Tris (pH 8.0), 90 mM borate, and 2 mM EDTA, 640 Vh run at 4°C, and stained with 0.5 mg/ml ethidium bromide. The DNA band was cut and histone content was estimated by liquid scintillation analysis of 3 H-labeled histones after subtraction of the negative controls ( 3Hlabeled histone and DNA alone, respectively). Although differences in the electrophoretic mobility could be detected, counting of histones bound to DNA allowed a better quantification.
Detection of Covalent Cross-Links of DNA–Histones Incubation of Linear DNA with Isolated Histones Five-hundred micrograms of linear =pGEM (Promega) was digested with EcoRI (Boehringer Mannheim) to generate a defined linear DNA fragment, extracted with phenol, phenol/chloroform/ isoamyl alcohol (24/24/1), and precipitated. Next, 1 mg of linear plasmid DNA was incubated with 2 mg 3H-labeled histones for in vitro reassembly in 10 mM Tris (pH 7.4), 130 mM NaCl, and 1 mM dithiothreitol for 2 h at 4°C. The total reassembly reaction volume was 20 ml.
To detect covalent cross-links of DNA– histones an SDS– urea electrophoresis using nonreducing conditions was performed (13). Isolated radiolabeled 3H-histones were incubated with linear DNA and treated with H 2O 2 as described above. For the separation a 3% stacking gel, an upper and a lower running gel with 5 and 20% in 1.5 M Tris, 0.4% SDS, 8 M urea (pH 6.8) were used, respectively. After electrophoresis an ethidium bromide or a silver stain were performed. Gels were cut into pieces and the content of 3H was quantified by scintillation counting. The aim was to separate the DNA in the upper 5% running gel without loosing the relatively small histone molecules (molecular weight below 20 kDa).
Treatment of DNA–Histones with Hydrogen Peroxide DNA– histones were oxidatively damaged by incubation with 5, 15, or 30 mM H 2O 2 for 2 h at 25°C. Controls were incubated without H 2O 2. After incubation any remaining H 2O 2 was removed by addition of 0.1 mg catalase.
Proteolysis Measurements Fluorescamine assay. Degradation of histones was measured by incubation of 40 mg substrate protein with 0.6 mg proteasome in a buffer containing 50 mM Hepes (pH 7.8), 20 mM KCl, 5 mM MgOAc, and 0.5 mM dithiothreitol for 2 h at 37°C. The reaction was stopped by addition of an equal volume of 20% trichloroacetic acid. After centrifugation (15 min, 12,000g) the supernatants containing primary amines were neutralized with 1 M Hepes (pH 7.8). Fluorescamine (0.3 mg/ml in acetone) was added during rigorous shaking and the fluorescence was determined at 390 nm excitation/470 nm emission, using leucine as a standard. Proteolysis rates were calculated as the difference between sample values and blank values (sum
RESULTS
The proteolytic susceptibility of histones increased with mild oxidation and then decreased with more severe oxidative damage. The dependence of proteolytic susceptibility on H 2O 2 damage is reported for histone 2B in Fig. 1. After reaching an optimum degree of oxidative damage at 15 mM (7.8-fold increased proteolytic susceptibility), further damage decreased the proteolytic susceptibility of histone 2B (Fig. 1). When we compared the proteolytic susceptibilities of histone fractions H1, H2A, H2B, H3, and H4 and a histone mixture, all showed essentially the same pattern of response (as shown by histone 2B in Fig. 1) to oxidative modification. Histone H1 exhibited a basal rate of degradation 4- to 5-fold greater than any other histone
20S PROTEASOME IN THE DEGRADATION OF OXIDIZED PROTEINS
fraction, and its maximal degradation rate following oxidative damage was similarly high (Table I). Histone fraction H4 exhibited the greatest relative increase in proteolytic susceptibility following oxidation, almost 14-fold, and this occurred at a peroxide exposure of only 5 mM. At the other end of the spectrum histone H2A exhibited a maximal proteolytic response to H 2O 2 of only 6-fold, and this required an H 2O 2 exposure of 15 mM (Table I). Since only a small fraction of histones in the nuclei of mammalian cells is soluble, we decided to study the proteolytic degradation of oxidized histones bound to DNA. As a model system we chose plasmid DNA, linearized and loaded in vitro with histones. The loading of DNA with histones was followed by a gel mobility shift (Fig. 2). A clear band shift was seen in agarose gel electrophoresis at a histone:DNA ratio of 2:1. Higher concentrations of histones led to a loss of ethidium bromide staining, probably due to a band broadening. We tested the effect of H 2O 2 on the DNA binding of
FIG. 1. Susceptibility of oxidatively modified histone 2B to proteolytic degradation by proteasome. Isolated histone 2B was oxidatively damaged by exposure to various concentrations of H 2O 2 in 20 mM phosphate buffer (pH 7.4) for 2 h at 25°C and then dialyzed extensively. Proteolysis experiments were performed with the isolated proteasome from human erythrocytes (25 mg/ml), prepared according to Hough et al. (10), and the damaged or nondamaged histone (1.3 mg/ml). Proteolytic degradation was measured using the fluorescamine assay and proteolysis is expressed as nanomoles of free amines generated per minute per milligram of proteasome (nmol 3 min 21 3 mg 21). All values are the mean 6 SD of five independent experiments).
213
FIG. 2. Agarose gel electrophoresis of DNA– histone complexes. DNA was linearized with EcoRI and then incubated with 3H-labeled histones for in vitro reassembly in 10 mM Tris (pH 7.4), 130 mM NaCl, and 1 mM dithiothreitol for 2 h at 4°C. The complex was analyzed in a 0.8 % agarose gel, containing 90 mM Tris (pH 8.0), 90 mM borate, and 2 mM EDTA and 640 Vh run at 4°C. The staining was performed with ethidium bromide.
radiolabeled histones. DNA– histone complexes were treated with H 2O 2 and analyzed by gel mobility shift assay. The DNA band was then cut out of the gels and the histone-bound radioactivity was measured. As one can see in Fig. 3A a relatively modest amount of histones (35%) was detached from the DNA during treatment with 5 mM hydrogen peroxide. Treatment with 15 or 30 mM H 2O 2 caused even less histone detachment (probably due to the formation of DNA– histone aggregates, as discussed below). We next tested the degradation of oxidized DNA– histone complexes by the proteasome. As shown in Fig. 3A isolated proteasome was able to remove about 95% of the histones bound to linear DNA after oxidation with a modest amount of H 2O 2 (5 mM). More severely damaged histones bound to DNA became worse substrates for the proteasome (Fig. 3A), in agreement with the results on isolated histones (Fig. 1). In the electrophoretic mobility shift assay, H 2O 2-treated DNA– histone complexes showed the same electrophoretic mobility as did nontreated DNA– histones (Fig. 3B). The addition of proteasome to the DNA– histone complex oxidized with 5 mM H 2O 2 gave an electrophoretic mobility comparable to free DNA. These data are in accordance with the results obtained from the measurement of 3H-labeled histones bound to DNA and suggest an active proteolytic degradation of oxidized histones bound to DNA. Incubation of nondamaged DNA– histone complexes with the 20S proteasome caused no change in electrophoretic mobility, suggesting that only oxidized histones are substrates for the 20S proteasome. Since we found a decreased degradation of DNAbound histones at higher degrees of oxidation we addressed this question in the next series of experiments. It has been suggested that protein aggregates are able to inhibit the proteasome (14). Therefore, we measured proteasomal activity against histones (oxidized and non-oxidized) and against the fluorogenic peptide sucLLVY-MCA. In the presence of damaged DNA– histone complexes the activity of the proteasome decreased in a concentration-dependent manner (Table II). This result suggested that an inhibition of the proteasome by aggregates could be the reason for the decreased degradation of histones bound to DNA after excessive ox-
214
ULLRICH ET AL. TABLE I
Comparison of the Proteolytic Susceptibilities of Histone Fractions H1, H2A, H2B, H3, and H4
Histone fraction
Basal degradation (nmol 3 min 21 3 mg proteasome 21)
Maximal degradation (nmol 3 min 21 3 mg proteasome 21)
Ratio
H 2O 2 concentration at maximal proteolytic susceptibility (mM)
H1 H2A H2B H3 H4 Histone mixture
77.6 6 13.2 17.6 6 4.9 17.8 6 2.5 14.0 6 1.2 13.0 6 1.1 16.5 6 2.2
489 6 39 122 6 21 140 6 26 157 6 28 180 6 17 145 6 11
6.3 6.9 7.8 11.2 13.8 8.8
5 15 15 10 5 15
Note. Isolated histones or isolated histone fractions H1, H2A, H2B, H3, and H4 (2 mg/ml) were oxidatively damaged by exposure to various concentrations of H 2O 2 in 20 mM phosphate buffer (pH 7.4) for 2 h at 25°C and then dialyzed extensively. Proteolysis experiments were performed with the isolated proteasome from human erythrocytes (25 mg/ml), prepared according to Hough et al. (10), and the damaged or nondamaged histones (1.3 mg/ml). Proteolysis was measured using the fluorescamine assay. Basal and maximal proteolytic rates of histone degradation are expressed as nanomoles of free amines generated per minute per milligram of proteasome (nmol 3 min 21 3 mg proteasome 21). The H 2O 2 concentration required to elicit maximal proteolytic susceptibility is indicated in the far right column. All values are the mean 6 SD of three independent experiments.
idation. It is well known that oxidation can cause proteins to cross-link with DNA (8, 9) as well as with other protein molecules (14, 15). Therefore, we postulate that aggregates of histones and/or aggregates of histone– DNA may have inhibited the proteasome in our model system. To further explore this possibility we next measured the formation of aggregates using a nonreducing SDS– urea–PAGE. As shown in Fig. 4 we discovered that the amount of histones covalently bound
to DNA increased 7-fold after treatment with 30 mM H 2O 2. We were not able to find detectable amounts of protein–protein aggregates using silver staining or by counting of gel areas (Fig. 4). Therefore, we conclude that protein aggregates bound to DNA are responsible for inhibiting the proteasome in our system. As shown in Fig. 4C a concentration-dependent increase in the formation of histone–DNA cross-links was observed in our experiments.
FIG. 3. Removal of oxidatively damaged histones from DNA by the 20S proteasome. The legend on the bottom axis of A indicates the extent of histone oxidation expressed as H 2O 2 concentration during protein damage, and the ordinate indicates the 3H-histone-borne dpm bound to DNA after electrophoretic separation in an agarose gel. Values are given as means 6 SD of three independent experiments. Proteasome was isolated from human erythrocytes according to Hough et al. (10). DNA (1.0 mg ) was incubated with 2 mg 3H-labeled histones for in vitro reassembly in 10 mM Tris (pH 7.4), 130 mM NaCl, and 1 mM dithiothreitol for 2 h at 4°C. DNA– histones were damaged by incubation with 5, 15, or 30 mM H 2O 2 for 2 h at 25°C. Controls were incubated without hydrogen peroxide. After 2 h any remaining H 2O 2 was removed by 0.1 mg catalase. Proteolysis experiments were performed by addition of 0.6 mg proteasome in 50 mM Tris–HCl (pH 7.8), 20 mM KCl, 5 mM MgOAc, and 0.5 mM dithiothreitol for 2 h at 37°C. Controls were incubated without proteasome. DNA was analyzed finally in a 0.8% agarose gel, 90 mM Tris (pH 8.0), 90 mM borate, and 2 mM EDTA, 640 Vh run at 4°C, and stained with 0.5 mg/ml ethidium bromide. The DNA was cut out of each gel and histone content was estimated by liquid scintillation analysis of 3H-labeled histones. (B) The differences in the electrophoretic mobility of DNA– histone complexes after oxidation and incubation with the 20S proteasome.
215
20S PROTEASOME IN THE DEGRADATION OF OXIDIZED PROTEINS TABLE II
Activity of the Proteasome after Exposure to Hydrogen Peroxide-Treated DNA–Histone Complexes H 2O 2 suc-LLVY-MCA degradation (mM) (nmol MCA 3 min 21 3 mg proteasome 21) 0 5 15 30
Degradation of nondamaged histones (nmol 2NH 2 3 min 21 3 mg proteasome 21)
Degradation of damaged histones (nmol 2NH 2 3 min 21 3 mg proteasome 21)
1.4 6 0.2 1.3 6 0.1 1.2 6 0.2 0.8 6 0.1
5.6 6 0.7 5.2 6 0.4 4.9 6 0.7 2.9 6 0.3
25.8 6 1.8 26.4 6 3.7 24.5 6 3.4 5.7 6 0.9
Note. DNA/histone (nonlabeled) complexes were prepared as described under Material and Methods. Treatment of the complexes with the amounts of H 2O 2 indicated was performed for 2 h and the reactions were stopped by addition of catalase. Proteasome was then added to each reaction mixture for a 10-min incubation. After incubating for 10 min either damaged (15 mM H 2O 2) or nondamaged 3H-histones or the fluorogenic peptide suc-LLVY-MCA was added as substrates. Reactions were stopped after an incubation period of 2 h when radiolabeled histones were used as proteolytic substrates and after 30 min when the fluorogenic peptide substrate succ-LLVY-MCA was used. Degradation of the radiolabeled histone substrates and the fluorogenic peptide was measured as described under Materials and Methods. The data represent the mean 6 SD of three independent measurements.
DISCUSSION
Over the past several years numerous publications by various groups have demonstrated the key role of the 20S proteasome in the degradation of oxidized proteins (3, 5, 6, 12, 14 –19). These studies have been conducted with multiple cell types, including erythrocytes and reticulocytes, yeast, hepatocytes, and various rodent and human cell culture lines. All these studies focused on the degradation of cytosolic proteins by the cytosolic proteasome pool, yet nuclear proteins are also subject to oxidative damage and the nucleus has a high concentration of proteasome (4,30). To our knowledge, the present investigation is the first to suggest a role for the 20S proteasome in the degradation of oxidized nuclear proteins. In this study we were able to show that the susceptibility of histones to degradation by the 20S proteasome is strongly affected by (histone) protein oxidation.
Histones are at least in vitro substrates of the 20S proteasome and their proteolytic susceptibility is initially increased with the extent of their oxidation, whereas further damage leads to decreased proteolysis. This effect has already described for a number of cytosolic proteins by our group as well as by others (3, 5, 6, 14). All the histone fractions displayed this same general response to oxidation by H 2O 2 but each histone fraction exhibited unique concentration dependency and absolute proteolytic rates. The H 2O 2 concentrations required for maximal proteolytic susceptibility increased in the following order: H1 and H4 , H3 , H2A, and H2B. Both the basal and the maximal degradation rates for histone H1 surpassed those of all other histone types. Histone proteins are synthesized during a limited phase of the cell cycle and are coordinately dependent on DNA replication (20), but some are produced in a
FIG. 4. Formation of DNA– histone aggregates. DNA– histone complexes were prepared as described above and both peroxide-oxidized (30 mM H 2O 2) and untreated samples were analyzed by SDS– urea electrophoresis using nonreducing conditions (13). For the separation a 3% stacking gel, an upper and a lower running gel with 5 and 20% were used, respectively. The DNA– histone complexes were either stained with ethidium bromide to detect the DNA band (see arrow) or silver stained to detect the proteins (A). The gel was cut into seven equal parts and analyzed by b-scintillation counting. The dpm per fraction from a representative gel is reported in B. (C) The dependency of DNA– histone aggregate formation on H 2O 2 concentration. Values are given as the mean 6 SD of three independent experiments.
216
ULLRICH ET AL.
constitutive manner, even in nonproliferating cells (21, 22) and replace preexisting nucleosomes within the chromatin structure in the absence of DNA replication. Replacement of histones in nonproliferating cells is required for nucleosome assembly after transcription due to a loss of histones (22–25). A variant of histone H3 has been shown to be constitutively produced in animal cells (24) and histones H2A and H2B seem to exchange between nucleosomes during transcription and replication (26, 27), whereas histone H4 seems to be more stable (28, 29). Although it is acknowledged that histones have finite life spans, little is known about the protease(s) involved in the turnover of these proteins. In the present study we provide data suggesting that histones and histones bound to DNA are potential substrates for the 20S proteasome. Whereas oxidation alone was able to detach only a limited amount of histones from DNA, the 20S proteasome was able to remove a large quantity of oxidized histones from a reconstituted DNA– histone complex. Histones in the DNA– histone complexes were not degraded by the 20S proteasome unless they were first exposed to oxidative modification which (as demonstrated both by direct proteolysis measurements and by electrophoretic mobility shift assays). Therefore, oxidized histones bound to DNA are excellent substrates for the 20S proteasome. But as shown earlier excessive oxidative damage to proteins actually decreases their proteolytic susceptibility (3, 5, 6, 14). This effect has been interpreted both as stearic hindrance (3, 5, 6) and as an inhibition of proteasome by aggregated proteins (14). In our model system we found a decline in proteolytic degradation of histones bound to DNA. Since we were not able to detect protein–protein aggregates, but sufficient amounts of DNA–protein aggregates were found, we postulate an inhibitory effect of DNA– histone aggregates on the 20S proteasome. Since proteasome is present in the nuclei of mammalian cells (4, 30) and is able to selectively degrade oxidized histones, we further postulate that nuclear proteasome is responsible for degradation of oxidized nuclear proteins and that proteasome is able to remove oxidized histones from the chromatin. Each of these functions implies important roles for nuclear proteasome in the overall antioxidant defenses of the cell. ACKNOWLEDGMENTS We thank Ms. J. Gieche and Ms. K. Eckardt for her skillful technical assistance. The work of O.U. was supported by a fellowship from the Ernst-Schering-Research-Foundation. T.G. was supported by DFG.
REFERENCES 1. Ward, J. F. (1994) Radiat. Res. 138(1 Suppl.), S85–S88. 2. Halliwell, B., and Gutteridge, J. M. C. (1989) Free Radicals in Biology and Medicine, 2nd ed., Clarendon Press, Oxford.
3. Grune, T., Reinheckel, T., and Davies, K. J. A. (1997) FASEB J. 11, 526 –534. 4. Palmer, A., Rivett, A. J., Thomson, S., Hendil, K. B., Butcher, G. W., Fuertes, G., and Knecht, E. (1996) Biochem. J. 316, 401– 407 5. Grune, T., Reinheckel, T., Joshi, M., and Davies, K. J. A. (1995) J. Biol. Chem. 270, 2344 –2351. 6. Grune, T., Reinheckel, T., and Davies, K. J. A. (1996) J. Biol. Chem. 271, 15504 –15509. 7. Enright, H. U., Miller, W. J., and Hebbel. R. P. (1992) Nucleic Acids Res. 20, 3341–3346. 8. Jones, N. J., Ellard, S., Waters, R., and Parry, E. M. (1993) Carcinogenesis 14, 2487–2494. 9. Dizdaroglu, M. (1993) in DNA and Free Radicals (Halliwell, B., and Aruoma, O. I., Ed.), pp. 19 – 40, Ellis Horwood, Chichester, UK. 10. Hough, R., Pratt, G., and Rechsteiner, M. (1987) J. Biol. Chem. 262, 8303– 8313. 11. Jentoft, N., and Deanborn, D. G. (1979) J. Biol. Chem. 254, 4359 – 4365. 12. Pacifici, R. E., and Davies, K. J. A. (1990) Methods Enzymol. 186, 485–502. 13. Albert, P., Lacorre-Arescaldino, I., and Toublan, B. (1993) Anal. Biochem. 209, 224 –227. 14. Friguet, B., Stadtman, E. R., and Szweda, L. I. (1994) J. Biol. Chem. 269, 21639 –21643. 15. Pacifici, R. E., Kono, Y., and Davies, K. J. A. (1993) J. Biol. Chem. 268, 15405–15411. 16. Rivett, A. J. (1985) J. Biol. Chem. 260, 300 –305. 17. Rivett, A. J. (1985) J. Biol. Chem. 260, 12600 –12606. 18. Levine, R. L., Mosoni, L., Berlett, B. S., and Stadtman, E. R. (1996) Proc. Natl. Acad. Sci. USA 93, 15036 –15040. 19. Chao, C.-C., Ma, Y.-S., and Stadtman, E. R. (1997) Proc. Natl. Acad. Sci. USA 94, 2969 –2974. 20. Stein, G., Stein, J., Chrysogelos, S., Dell’Orco, R., Hickey, P., Kroeger, P., Morris, T., Nick, H., Pauli, U., van Wijnen, A., Weber, L., and Zambetti, G. (1989) in Histones and Other Basic Nuclear Proteins (Hnilica, L. S., Stein, G., and Stein, J., Eds.), pp. 269 –317, CRC Press, Boca Raton, FL. 21. Wu, R. S., Panusz, H. T., Hatch, C. L., and Bonner, W. M. (1986) CRC Crit. Rev. Biochem. 20, 201–263. 22. Marzluff, W. F. (1989) in Histones and Other Basic Nuclear Proteins (Hnilica, L. S., Stein, G., and Stein, J., Eds.), pp. 319 – 329, CRC Press, Boca Raton, FL. 23. Zweidler, A. (1984) in Histone Genes: Structure, Organization and Function (Stein, G., Stein, J., and Marzluff, W. F., Eds.), pp. 339 –371, Wiley, New York. 24. Jackson, V. (1990) Biochemisry 29, 719 –731. 25. Pfaffle, P., Gerlach, V., Bunzel, L., and Jackson, V. L. (1990) J. Biol. Chem. 265, 16830 –16840. 26. Louters, L., and Chalkey, R. (1985) Biochemistry 24, 3080 – 3085. 27. van Holde, K., Lohr, D. E., and Robert, C. (1982) J. Biol. Chem. 267, 2837–2840. 28. Grove, G. W., and Zweidler, A. (1984) Biochemistry 23, 4436 – 4443. 29. Tsvetkov, S., Ivanova, E., and Djondjurov, L. (1989) Exp. Cell Res. 180, 94 –105. 30. Coux, O., Tanaka, K., and Goldberg, A. L. (1996) Annu. Rev. Biochem. 65, 801– 847.
Influence of DNA Binding on the Degradation of Oxidized Histones by the 20S Proteasome Oliver Ullrich,* Nicolle Sitte,* Olaf Sommerburg,* Volker Sandig,† Kelvin J. A. Davies,‡ and Tilman Grune* ,1 *Clinics of Physical Medicine and Rehabilitation, University Hospital Charite´, Humboldt-University, Berlin, Germany; †Max Delbrueck Center for Molecular Medicine, Berlin, Germany; and ‡Ethel Percy Andrus Gerontology Center, University of Southern California, Los Angeles, California 90089-0191
Received July 16, 1998, and in revised form October 6, 1998
The 20S proteasome is localized in the cytosol and nuclei of mammalian cells. Previous work has shown that the cytosolic 20S proteasome is largely responsible for the selective recognition and degradation of oxidatively damaged cytosolic proteins. Since nuclear proteins are also susceptible to oxidative damage (e.g., from metabolic free radical production, ionizing radiation, xenobiotics, chemotherapy) we investigated the degradation of oxidatively damaged histones, in the presence and in the absence of DNA, by the 20S proteasome. We find that both soluble histones and DNAbound histones are susceptible to selective proteolytic degradation by the 20S proteasome following mild oxidative damage. In contrast, more severe oxidative damage actually decreases the proteolytic susceptibility of histones. Soluble H1 showed the highest basal and maximal absolute proteolytic rates. Histone fraction H4 exhibited the greatest relative increase in proteolytic susceptibility following oxidation, almost 14fold, and this occurred at a peroxide exposure of 5 mM. At the other end of the spectrum, histone H2A exhibited a maximal proteolytic response to H 2O 2 of only 6-fold, and this required an H 2O 2 exposure of 15 mM. An oxidation of reconstituted linear DNA plasmid– histone complex makes up to 95% of the histones bound to DNA susceptible to degradation, whereas undamaged protein–DNA complexes are not substrates for the proteasome. Severe oxidation by high concentrations of H 2O 2 appears to decreases the proteolytic susceptibility of histones due to the formation of crosslinked histone–DNA aggregates which appear to inhibit the proteasome. We conclude that the degradation of nuclear proteins is highly selective and
1 To whom correspondence should be addressed: Clinics of Physical Medicine and Rehabilitation, University Hospital Charite´, HumboldtUniversity Berlin, Schumannstrasse 20/21, 10098 Berlin, Germany. Fax: 149 30 2093 7204. E-mail: [email protected].
0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
requires prior damage of the substrate protein, such as that caused by oxidation. © 1999 Academic Press Key Words: protein oxidation; histones; proteasome; DNA–protein aggregates.
Nuclear proteins are oxidized by reaction with oxygen radicals and other activated oxygen species, generated as by-products of cellular metabolism, or by exogenous sources, e.g., ionizing radiation or antitumor chemotherapy (1, 2). Little is known of the fate of such oxidatively modified nuclear proteins. In the cell cytoplasm most oxidatively damaged proteins undergo selective proteolysis, via an ATP- and ubiquitin-independent pathway (3). The 20S proteasome (4), a large 700-kDa multicatalytic, multisubunit proteinase, is largely responsible for the selective degradation of such oxidatively damaged cytoplasmic proteins (5, 6). Importantly, the proteasome is also found in high concentrations in the nucleus of mammalian cells (4), although its nuclear function is not known. Oxidative stress in the nucleus may contribute to histone modifications such as amino acid oxidation, protein–protein cross-linking, and DNA–protein crosslinking. We propose that selective proteolytic degradation of damaged and functionally altered long-lived histones may be essential for genomic integrity. Nucleosomal histones protect DNA from free radical mediated damage (7), and histone detachment and reattachment are closely connected with transcription and replication processes as with DNA repair and therefore requires functionally intact histones. Oxidatively damaged histones are able to cross-link with DNA (8, 9) and would impair the detachment-reassembly process. We, therefore, studied the ability of the 20S proteasome to degrade damaged histones as a function of the extent of their oxidation. Because histone modification 211
212
ULLRICH ET AL.
after oxidative stress may also change the DNA-binding properties of the histones, we also tested the influence of DNA on histone (protein) degradation by the proteasome. MATERIALS AND METHODS
Preparation of the Proteasome Proteasome was isolated according to Hough et al. (10). Cells were lysed in 1 mM dithiothreitol. After removal of membranes and nonlysed cells by centrifugation the proteasome was purified by DEAE chromatography, sucrose density gradient ultracentrifugation, and separation on a Mono Q column using an FPLC system. The purity of the preparation was controlled by both one-dimensional SDS– PAGE and nondenaturing polyacrylamide electrophoresis (10). Nondenaturing electrophoresis was followed by overlaying the gel with a 200 mM sLLVY-MCA solution for detection of peptidase activity. The proteasome preparation was diluted finally to a protein concentration of 0.35 mg/ml.
Radioactive Labeling of Isolated Histones Isolated histones were radiolabeled by reductive methylation with [ 3H]formaldehyde and sodium cyanoborohydride as described by Jentoft and Dearborn (11) and then dialyzed extensively.
Treatment of Histones with Hydrogen Peroxide Isolated histones or isolated histone fractions H1, H2A, H2B, H3, and H4 (2 mg/ml) were oxidatively damaged using various concentrations of H 2O 2 in 20 mM phosphate buffer (pH 7.4) for 2 h at 25°C. Aliquots (400 ml) were then dialyzed for 16 h at 4°C, against 2.0 liters of 10 mM phosphate buffer containing 10 mM KCl, with one exchange of dialysis fluid after 3 h.
of incubated substrate protein without protease and of incubated protease only). Degradation of 3H-labeled histones. Degradation of 3H-labeled histones by proteasome was measured in 50 mM Tris–HCl (pH 7.8), 20 mM KCl, 5 mM MgOAc, and 0.5 mM dithiothreitol. Percentage of histone degradation was determined by liquid scintillation in supernatants of trichloroacetic acid (TCA)-precipitated proteins and calculated (12) according to the following formula: (TCA-soluble radioactivity 2 background radioactivity)/(total radioactivity 2 background radioactivity) 3 100. Degradation of DNA-bound histones. Proteolysis experiments were performed by adding of 0.6 mg proteasome in 50 mM Tris–HCl (pH 7.8), 20 mM KCl, 5 mM MgOAc, and 0.5 mM dithiothreitol for 2 h at 37°C to 2 mg radiolabeled histones. Controls were incubated without proteasome. Degradation of suc-LLVY-MCA. Degradation of the fluorogenic peptide suc-LLVY-MCA was measured as follows: to 270 ml reaction mixture containing the proteasome, 30 ml of the fluorogenic peptide suc-LLVY-MCA (2 mM stock solution in DMSO) was added. After incubation at 37°C for 1 h the reaction was stopped by addition of an equal volume of ice-cold ethanol and 10 vol of 125 mM borate buffer, pH 9.0. Fluorescence was determined at 380 nm excitation/440 nm emission.
Agarose Gel Electrophoresis DNA– histones were analyzed in a 0.8% agarose gel, 90 mM Tris (pH 8.0), 90 mM borate, and 2 mM EDTA, 640 Vh run at 4°C, and stained with 0.5 mg/ml ethidium bromide. The DNA band was cut and histone content was estimated by liquid scintillation analysis of 3 H-labeled histones after subtraction of the negative controls ( 3Hlabeled histone and DNA alone, respectively). Although differences in the electrophoretic mobility could be detected, counting of histones bound to DNA allowed a better quantification.
Detection of Covalent Cross-Links of DNA–Histones Incubation of Linear DNA with Isolated Histones Five-hundred micrograms of linear =pGEM (Promega) was digested with EcoRI (Boehringer Mannheim) to generate a defined linear DNA fragment, extracted with phenol, phenol/chloroform/ isoamyl alcohol (24/24/1), and precipitated. Next, 1 mg of linear plasmid DNA was incubated with 2 mg 3H-labeled histones for in vitro reassembly in 10 mM Tris (pH 7.4), 130 mM NaCl, and 1 mM dithiothreitol for 2 h at 4°C. The total reassembly reaction volume was 20 ml.
To detect covalent cross-links of DNA– histones an SDS– urea electrophoresis using nonreducing conditions was performed (13). Isolated radiolabeled 3H-histones were incubated with linear DNA and treated with H 2O 2 as described above. For the separation a 3% stacking gel, an upper and a lower running gel with 5 and 20% in 1.5 M Tris, 0.4% SDS, 8 M urea (pH 6.8) were used, respectively. After electrophoresis an ethidium bromide or a silver stain were performed. Gels were cut into pieces and the content of 3H was quantified by scintillation counting. The aim was to separate the DNA in the upper 5% running gel without loosing the relatively small histone molecules (molecular weight below 20 kDa).
Treatment of DNA–Histones with Hydrogen Peroxide DNA– histones were oxidatively damaged by incubation with 5, 15, or 30 mM H 2O 2 for 2 h at 25°C. Controls were incubated without H 2O 2. After incubation any remaining H 2O 2 was removed by addition of 0.1 mg catalase.
Proteolysis Measurements Fluorescamine assay. Degradation of histones was measured by incubation of 40 mg substrate protein with 0.6 mg proteasome in a buffer containing 50 mM Hepes (pH 7.8), 20 mM KCl, 5 mM MgOAc, and 0.5 mM dithiothreitol for 2 h at 37°C. The reaction was stopped by addition of an equal volume of 20% trichloroacetic acid. After centrifugation (15 min, 12,000g) the supernatants containing primary amines were neutralized with 1 M Hepes (pH 7.8). Fluorescamine (0.3 mg/ml in acetone) was added during rigorous shaking and the fluorescence was determined at 390 nm excitation/470 nm emission, using leucine as a standard. Proteolysis rates were calculated as the difference between sample values and blank values (sum
RESULTS
The proteolytic susceptibility of histones increased with mild oxidation and then decreased with more severe oxidative damage. The dependence of proteolytic susceptibility on H 2O 2 damage is reported for histone 2B in Fig. 1. After reaching an optimum degree of oxidative damage at 15 mM (7.8-fold increased proteolytic susceptibility), further damage decreased the proteolytic susceptibility of histone 2B (Fig. 1). When we compared the proteolytic susceptibilities of histone fractions H1, H2A, H2B, H3, and H4 and a histone mixture, all showed essentially the same pattern of response (as shown by histone 2B in Fig. 1) to oxidative modification. Histone H1 exhibited a basal rate of degradation 4- to 5-fold greater than any other histone
20S PROTEASOME IN THE DEGRADATION OF OXIDIZED PROTEINS
fraction, and its maximal degradation rate following oxidative damage was similarly high (Table I). Histone fraction H4 exhibited the greatest relative increase in proteolytic susceptibility following oxidation, almost 14-fold, and this occurred at a peroxide exposure of only 5 mM. At the other end of the spectrum histone H2A exhibited a maximal proteolytic response to H 2O 2 of only 6-fold, and this required an H 2O 2 exposure of 15 mM (Table I). Since only a small fraction of histones in the nuclei of mammalian cells is soluble, we decided to study the proteolytic degradation of oxidized histones bound to DNA. As a model system we chose plasmid DNA, linearized and loaded in vitro with histones. The loading of DNA with histones was followed by a gel mobility shift (Fig. 2). A clear band shift was seen in agarose gel electrophoresis at a histone:DNA ratio of 2:1. Higher concentrations of histones led to a loss of ethidium bromide staining, probably due to a band broadening. We tested the effect of H 2O 2 on the DNA binding of
FIG. 1. Susceptibility of oxidatively modified histone 2B to proteolytic degradation by proteasome. Isolated histone 2B was oxidatively damaged by exposure to various concentrations of H 2O 2 in 20 mM phosphate buffer (pH 7.4) for 2 h at 25°C and then dialyzed extensively. Proteolysis experiments were performed with the isolated proteasome from human erythrocytes (25 mg/ml), prepared according to Hough et al. (10), and the damaged or nondamaged histone (1.3 mg/ml). Proteolytic degradation was measured using the fluorescamine assay and proteolysis is expressed as nanomoles of free amines generated per minute per milligram of proteasome (nmol 3 min 21 3 mg 21). All values are the mean 6 SD of five independent experiments).
213
FIG. 2. Agarose gel electrophoresis of DNA– histone complexes. DNA was linearized with EcoRI and then incubated with 3H-labeled histones for in vitro reassembly in 10 mM Tris (pH 7.4), 130 mM NaCl, and 1 mM dithiothreitol for 2 h at 4°C. The complex was analyzed in a 0.8 % agarose gel, containing 90 mM Tris (pH 8.0), 90 mM borate, and 2 mM EDTA and 640 Vh run at 4°C. The staining was performed with ethidium bromide.
radiolabeled histones. DNA– histone complexes were treated with H 2O 2 and analyzed by gel mobility shift assay. The DNA band was then cut out of the gels and the histone-bound radioactivity was measured. As one can see in Fig. 3A a relatively modest amount of histones (35%) was detached from the DNA during treatment with 5 mM hydrogen peroxide. Treatment with 15 or 30 mM H 2O 2 caused even less histone detachment (probably due to the formation of DNA– histone aggregates, as discussed below). We next tested the degradation of oxidized DNA– histone complexes by the proteasome. As shown in Fig. 3A isolated proteasome was able to remove about 95% of the histones bound to linear DNA after oxidation with a modest amount of H 2O 2 (5 mM). More severely damaged histones bound to DNA became worse substrates for the proteasome (Fig. 3A), in agreement with the results on isolated histones (Fig. 1). In the electrophoretic mobility shift assay, H 2O 2-treated DNA– histone complexes showed the same electrophoretic mobility as did nontreated DNA– histones (Fig. 3B). The addition of proteasome to the DNA– histone complex oxidized with 5 mM H 2O 2 gave an electrophoretic mobility comparable to free DNA. These data are in accordance with the results obtained from the measurement of 3H-labeled histones bound to DNA and suggest an active proteolytic degradation of oxidized histones bound to DNA. Incubation of nondamaged DNA– histone complexes with the 20S proteasome caused no change in electrophoretic mobility, suggesting that only oxidized histones are substrates for the 20S proteasome. Since we found a decreased degradation of DNAbound histones at higher degrees of oxidation we addressed this question in the next series of experiments. It has been suggested that protein aggregates are able to inhibit the proteasome (14). Therefore, we measured proteasomal activity against histones (oxidized and non-oxidized) and against the fluorogenic peptide sucLLVY-MCA. In the presence of damaged DNA– histone complexes the activity of the proteasome decreased in a concentration-dependent manner (Table II). This result suggested that an inhibition of the proteasome by aggregates could be the reason for the decreased degradation of histones bound to DNA after excessive ox-
214
ULLRICH ET AL. TABLE I
Comparison of the Proteolytic Susceptibilities of Histone Fractions H1, H2A, H2B, H3, and H4
Histone fraction
Basal degradation (nmol 3 min 21 3 mg proteasome 21)
Maximal degradation (nmol 3 min 21 3 mg proteasome 21)
Ratio
H 2O 2 concentration at maximal proteolytic susceptibility (mM)
H1 H2A H2B H3 H4 Histone mixture
77.6 6 13.2 17.6 6 4.9 17.8 6 2.5 14.0 6 1.2 13.0 6 1.1 16.5 6 2.2
489 6 39 122 6 21 140 6 26 157 6 28 180 6 17 145 6 11
6.3 6.9 7.8 11.2 13.8 8.8
5 15 15 10 5 15
Note. Isolated histones or isolated histone fractions H1, H2A, H2B, H3, and H4 (2 mg/ml) were oxidatively damaged by exposure to various concentrations of H 2O 2 in 20 mM phosphate buffer (pH 7.4) for 2 h at 25°C and then dialyzed extensively. Proteolysis experiments were performed with the isolated proteasome from human erythrocytes (25 mg/ml), prepared according to Hough et al. (10), and the damaged or nondamaged histones (1.3 mg/ml). Proteolysis was measured using the fluorescamine assay. Basal and maximal proteolytic rates of histone degradation are expressed as nanomoles of free amines generated per minute per milligram of proteasome (nmol 3 min 21 3 mg proteasome 21). The H 2O 2 concentration required to elicit maximal proteolytic susceptibility is indicated in the far right column. All values are the mean 6 SD of three independent experiments.
idation. It is well known that oxidation can cause proteins to cross-link with DNA (8, 9) as well as with other protein molecules (14, 15). Therefore, we postulate that aggregates of histones and/or aggregates of histone– DNA may have inhibited the proteasome in our model system. To further explore this possibility we next measured the formation of aggregates using a nonreducing SDS– urea–PAGE. As shown in Fig. 4 we discovered that the amount of histones covalently bound
to DNA increased 7-fold after treatment with 30 mM H 2O 2. We were not able to find detectable amounts of protein–protein aggregates using silver staining or by counting of gel areas (Fig. 4). Therefore, we conclude that protein aggregates bound to DNA are responsible for inhibiting the proteasome in our system. As shown in Fig. 4C a concentration-dependent increase in the formation of histone–DNA cross-links was observed in our experiments.
FIG. 3. Removal of oxidatively damaged histones from DNA by the 20S proteasome. The legend on the bottom axis of A indicates the extent of histone oxidation expressed as H 2O 2 concentration during protein damage, and the ordinate indicates the 3H-histone-borne dpm bound to DNA after electrophoretic separation in an agarose gel. Values are given as means 6 SD of three independent experiments. Proteasome was isolated from human erythrocytes according to Hough et al. (10). DNA (1.0 mg ) was incubated with 2 mg 3H-labeled histones for in vitro reassembly in 10 mM Tris (pH 7.4), 130 mM NaCl, and 1 mM dithiothreitol for 2 h at 4°C. DNA– histones were damaged by incubation with 5, 15, or 30 mM H 2O 2 for 2 h at 25°C. Controls were incubated without hydrogen peroxide. After 2 h any remaining H 2O 2 was removed by 0.1 mg catalase. Proteolysis experiments were performed by addition of 0.6 mg proteasome in 50 mM Tris–HCl (pH 7.8), 20 mM KCl, 5 mM MgOAc, and 0.5 mM dithiothreitol for 2 h at 37°C. Controls were incubated without proteasome. DNA was analyzed finally in a 0.8% agarose gel, 90 mM Tris (pH 8.0), 90 mM borate, and 2 mM EDTA, 640 Vh run at 4°C, and stained with 0.5 mg/ml ethidium bromide. The DNA was cut out of each gel and histone content was estimated by liquid scintillation analysis of 3H-labeled histones. (B) The differences in the electrophoretic mobility of DNA– histone complexes after oxidation and incubation with the 20S proteasome.
215
20S PROTEASOME IN THE DEGRADATION OF OXIDIZED PROTEINS TABLE II
Activity of the Proteasome after Exposure to Hydrogen Peroxide-Treated DNA–Histone Complexes H 2O 2 suc-LLVY-MCA degradation (mM) (nmol MCA 3 min 21 3 mg proteasome 21) 0 5 15 30
Degradation of nondamaged histones (nmol 2NH 2 3 min 21 3 mg proteasome 21)
Degradation of damaged histones (nmol 2NH 2 3 min 21 3 mg proteasome 21)
1.4 6 0.2 1.3 6 0.1 1.2 6 0.2 0.8 6 0.1
5.6 6 0.7 5.2 6 0.4 4.9 6 0.7 2.9 6 0.3
25.8 6 1.8 26.4 6 3.7 24.5 6 3.4 5.7 6 0.9
Note. DNA/histone (nonlabeled) complexes were prepared as described under Material and Methods. Treatment of the complexes with the amounts of H 2O 2 indicated was performed for 2 h and the reactions were stopped by addition of catalase. Proteasome was then added to each reaction mixture for a 10-min incubation. After incubating for 10 min either damaged (15 mM H 2O 2) or nondamaged 3H-histones or the fluorogenic peptide suc-LLVY-MCA was added as substrates. Reactions were stopped after an incubation period of 2 h when radiolabeled histones were used as proteolytic substrates and after 30 min when the fluorogenic peptide substrate succ-LLVY-MCA was used. Degradation of the radiolabeled histone substrates and the fluorogenic peptide was measured as described under Materials and Methods. The data represent the mean 6 SD of three independent measurements.
DISCUSSION
Over the past several years numerous publications by various groups have demonstrated the key role of the 20S proteasome in the degradation of oxidized proteins (3, 5, 6, 12, 14 –19). These studies have been conducted with multiple cell types, including erythrocytes and reticulocytes, yeast, hepatocytes, and various rodent and human cell culture lines. All these studies focused on the degradation of cytosolic proteins by the cytosolic proteasome pool, yet nuclear proteins are also subject to oxidative damage and the nucleus has a high concentration of proteasome (4,30). To our knowledge, the present investigation is the first to suggest a role for the 20S proteasome in the degradation of oxidized nuclear proteins. In this study we were able to show that the susceptibility of histones to degradation by the 20S proteasome is strongly affected by (histone) protein oxidation.
Histones are at least in vitro substrates of the 20S proteasome and their proteolytic susceptibility is initially increased with the extent of their oxidation, whereas further damage leads to decreased proteolysis. This effect has already described for a number of cytosolic proteins by our group as well as by others (3, 5, 6, 14). All the histone fractions displayed this same general response to oxidation by H 2O 2 but each histone fraction exhibited unique concentration dependency and absolute proteolytic rates. The H 2O 2 concentrations required for maximal proteolytic susceptibility increased in the following order: H1 and H4 , H3 , H2A, and H2B. Both the basal and the maximal degradation rates for histone H1 surpassed those of all other histone types. Histone proteins are synthesized during a limited phase of the cell cycle and are coordinately dependent on DNA replication (20), but some are produced in a
FIG. 4. Formation of DNA– histone aggregates. DNA– histone complexes were prepared as described above and both peroxide-oxidized (30 mM H 2O 2) and untreated samples were analyzed by SDS– urea electrophoresis using nonreducing conditions (13). For the separation a 3% stacking gel, an upper and a lower running gel with 5 and 20% were used, respectively. The DNA– histone complexes were either stained with ethidium bromide to detect the DNA band (see arrow) or silver stained to detect the proteins (A). The gel was cut into seven equal parts and analyzed by b-scintillation counting. The dpm per fraction from a representative gel is reported in B. (C) The dependency of DNA– histone aggregate formation on H 2O 2 concentration. Values are given as the mean 6 SD of three independent experiments.
216
ULLRICH ET AL.
constitutive manner, even in nonproliferating cells (21, 22) and replace preexisting nucleosomes within the chromatin structure in the absence of DNA replication. Replacement of histones in nonproliferating cells is required for nucleosome assembly after transcription due to a loss of histones (22–25). A variant of histone H3 has been shown to be constitutively produced in animal cells (24) and histones H2A and H2B seem to exchange between nucleosomes during transcription and replication (26, 27), whereas histone H4 seems to be more stable (28, 29). Although it is acknowledged that histones have finite life spans, little is known about the protease(s) involved in the turnover of these proteins. In the present study we provide data suggesting that histones and histones bound to DNA are potential substrates for the 20S proteasome. Whereas oxidation alone was able to detach only a limited amount of histones from DNA, the 20S proteasome was able to remove a large quantity of oxidized histones from a reconstituted DNA– histone complex. Histones in the DNA– histone complexes were not degraded by the 20S proteasome unless they were first exposed to oxidative modification which (as demonstrated both by direct proteolysis measurements and by electrophoretic mobility shift assays). Therefore, oxidized histones bound to DNA are excellent substrates for the 20S proteasome. But as shown earlier excessive oxidative damage to proteins actually decreases their proteolytic susceptibility (3, 5, 6, 14). This effect has been interpreted both as stearic hindrance (3, 5, 6) and as an inhibition of proteasome by aggregated proteins (14). In our model system we found a decline in proteolytic degradation of histones bound to DNA. Since we were not able to detect protein–protein aggregates, but sufficient amounts of DNA–protein aggregates were found, we postulate an inhibitory effect of DNA– histone aggregates on the 20S proteasome. Since proteasome is present in the nuclei of mammalian cells (4, 30) and is able to selectively degrade oxidized histones, we further postulate that nuclear proteasome is responsible for degradation of oxidized nuclear proteins and that proteasome is able to remove oxidized histones from the chromatin. Each of these functions implies important roles for nuclear proteasome in the overall antioxidant defenses of the cell. ACKNOWLEDGMENTS We thank Ms. J. Gieche and Ms. K. Eckardt for her skillful technical assistance. The work of O.U. was supported by a fellowship from the Ernst-Schering-Research-Foundation. T.G. was supported by DFG.
REFERENCES 1. Ward, J. F. (1994) Radiat. Res. 138(1 Suppl.), S85–S88. 2. Halliwell, B., and Gutteridge, J. M. C. (1989) Free Radicals in Biology and Medicine, 2nd ed., Clarendon Press, Oxford.
3. Grune, T., Reinheckel, T., and Davies, K. J. A. (1997) FASEB J. 11, 526 –534. 4. Palmer, A., Rivett, A. J., Thomson, S., Hendil, K. B., Butcher, G. W., Fuertes, G., and Knecht, E. (1996) Biochem. J. 316, 401– 407 5. Grune, T., Reinheckel, T., Joshi, M., and Davies, K. J. A. (1995) J. Biol. Chem. 270, 2344 –2351. 6. Grune, T., Reinheckel, T., and Davies, K. J. A. (1996) J. Biol. Chem. 271, 15504 –15509. 7. Enright, H. U., Miller, W. J., and Hebbel. R. P. (1992) Nucleic Acids Res. 20, 3341–3346. 8. Jones, N. J., Ellard, S., Waters, R., and Parry, E. M. (1993) Carcinogenesis 14, 2487–2494. 9. Dizdaroglu, M. (1993) in DNA and Free Radicals (Halliwell, B., and Aruoma, O. I., Ed.), pp. 19 – 40, Ellis Horwood, Chichester, UK. 10. Hough, R., Pratt, G., and Rechsteiner, M. (1987) J. Biol. Chem. 262, 8303– 8313. 11. Jentoft, N., and Deanborn, D. G. (1979) J. Biol. Chem. 254, 4359 – 4365. 12. Pacifici, R. E., and Davies, K. J. A. (1990) Methods Enzymol. 186, 485–502. 13. Albert, P., Lacorre-Arescaldino, I., and Toublan, B. (1993) Anal. Biochem. 209, 224 –227. 14. Friguet, B., Stadtman, E. R., and Szweda, L. I. (1994) J. Biol. Chem. 269, 21639 –21643. 15. Pacifici, R. E., Kono, Y., and Davies, K. J. A. (1993) J. Biol. Chem. 268, 15405–15411. 16. Rivett, A. J. (1985) J. Biol. Chem. 260, 300 –305. 17. Rivett, A. J. (1985) J. Biol. Chem. 260, 12600 –12606. 18. Levine, R. L., Mosoni, L., Berlett, B. S., and Stadtman, E. R. (1996) Proc. Natl. Acad. Sci. USA 93, 15036 –15040. 19. Chao, C.-C., Ma, Y.-S., and Stadtman, E. R. (1997) Proc. Natl. Acad. Sci. USA 94, 2969 –2974. 20. Stein, G., Stein, J., Chrysogelos, S., Dell’Orco, R., Hickey, P., Kroeger, P., Morris, T., Nick, H., Pauli, U., van Wijnen, A., Weber, L., and Zambetti, G. (1989) in Histones and Other Basic Nuclear Proteins (Hnilica, L. S., Stein, G., and Stein, J., Eds.), pp. 269 –317, CRC Press, Boca Raton, FL. 21. Wu, R. S., Panusz, H. T., Hatch, C. L., and Bonner, W. M. (1986) CRC Crit. Rev. Biochem. 20, 201–263. 22. Marzluff, W. F. (1989) in Histones and Other Basic Nuclear Proteins (Hnilica, L. S., Stein, G., and Stein, J., Eds.), pp. 319 – 329, CRC Press, Boca Raton, FL. 23. Zweidler, A. (1984) in Histone Genes: Structure, Organization and Function (Stein, G., Stein, J., and Marzluff, W. F., Eds.), pp. 339 –371, Wiley, New York. 24. Jackson, V. (1990) Biochemisry 29, 719 –731. 25. Pfaffle, P., Gerlach, V., Bunzel, L., and Jackson, V. L. (1990) J. Biol. Chem. 265, 16830 –16840. 26. Louters, L., and Chalkey, R. (1985) Biochemistry 24, 3080 – 3085. 27. van Holde, K., Lohr, D. E., and Robert, C. (1982) J. Biol. Chem. 267, 2837–2840. 28. Grove, G. W., and Zweidler, A. (1984) Biochemistry 23, 4436 – 4443. 29. Tsvetkov, S., Ivanova, E., and Djondjurov, L. (1989) Exp. Cell Res. 180, 94 –105. 30. Coux, O., Tanaka, K., and Goldberg, A. L. (1996) Annu. Rev. Biochem. 65, 801– 847.