Chromatin structure interferes with excision of abnormal bases from DNA

Biochimica et Biophysica Acta, 698 (1982) 15-21

15

Elsevier Biomedical Press BBA91094

C H R O M A T I N S T R U C T U R E I N T E R F E R E S W I T H EXCISION OF ABNORMAL BASES F R O M DNA KIICHI ISHIWATA * and ATSUSHI OIKAWA **

Department of Pharmacology, Research Institute for Tuberculosis and Cancer, Tohoku University, 4-1 Seiryomachi Sendai 980 (Japan) (Received February 5th, 1982)

Key words: DNA glycosylase; Chromatin structure; Uracil," Methy'ladenine-DNA; DNA repair; Base excision," (Human lymphoblast)

Cell-free extracts of human lymphoblastoid cells NL3 excised almost all uracil residues from free DNA with misincorporated dUMP, but only about half the uracil residues from nuclei, chromatin and reconstituted chromatin with dUMP-misincorporated DNA. This difference in susceptibility to uracii-DNA glycosylase of free and complexed DNAs was similar to the difference in susceptibility of free and complexed methylated DNAs to 3-methyladenine-DNA glycosylase. Methylated poly(dA-dT) was also protected by formation of complexes with calf thymus chromosomal proteins. It seems that the nucleosome structure prevents the action of DNA glycosylases. The very high sensitivity of PBS1 phage DNA, which contains uracil as a natural component, in complexes with calf thymus chromosomal proteins as well as in the free form [1] was confirmed. This high sensitivity seems ascribable to the high uracil content of PBS1 DNA. Methylated nucleosome monomers and dimers, and reconstituted nucleosome monomers containing methylated DNA of about 150 bp length, were considerably more resistant to 3-methyladenine-DNA glycosylase than chromatin reconstituted from methylated DNA of longer chain length. This may be due to the lower proportion of linker regions or free form stretches of the DNA chain in nucleosome oligomers.

Introduction

DNA glycosylases are, in some, if not all, cases, responsible for the first step of repair processing of DNA that carries abnormal bases misincorporated during replication or produced by modifications such as alkylation [2,3]. These enzymes seem ubiquitous in living things from bacteria to mammals. The action of these enzymes in eukaryotes may be complicated by the presence of nucleosomes, structural units of chromatin, which may interfere with the action of D N A glycosylases and also with that of nucleases. In fact it has been shown that after damaging * Present address: Cyclotron and Radioisotope Center, Tohoku University, Aoba, Sendai 980, Japan. **To whom correspondence should be addressed. 0167-4781/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press

by ultraviolet irradiation [4,5] or treatment with methyl methanesulfonate [6], repaired patches on D N A are sensitive to micrococcal nuclease for a short period, and then they gradually rearrange into normal, nuclease-resistant chromatin structure. Previously [1], we demonstrated that chromatin preparations reconstituted from methylated human D N A and calf thymus chromosomal proteins were only partially susceptible to human cell extracts with activity to release 3-methyladenine, whereas those reconstituted from PBS1 phage DNA, in which uracil replaces thymine as a natural component, and calf thymus chromosomal proteins were completely susceptible to activity for release of uracil. The apparent inconsistency between the actions of these two D N A glycosylases was re-examined using chromatin reconstituted with human D N A

16

containing misincorporated dUMP, methylated nucleosome monomers and dimers, and reconstituted nucleosome monomers containing methylated DNA. The presented results confirm previous results and provide evidence that the core particle regions are resistant to uracil-DNA glycosylase and 3-methyladenine-DNA glycosylase. A possible explanation of the unusual sensitivity of PBS1 phage D N A is given. Materials and Methods Cells and cell culture. A human lymphoblastoid cell line NL3, which was established by transformation with Epstein-Barr virus of peripheral lymphocytes of a healthy female donor [7] was used. The cells were maintained in suspension culture in RPMI 1640 medium supplemented with 15% fetal bovine serum, 2 mM glutamine and 50 /~g/ml kanamycin. This medium was also used for large-scale cultures unless otherwise stated. Nuclei and cell extracts. NL3 cells were grown to confluency and then cultured in medium containing a reduced amount of fetal bovine serum (0.5%) for 2 days to arrest them in the G~ phase. Then they were cultured again in the standard medium described above for 20 h. Cells mainly in the S phase were harvested by centrifugation at 250 × g for 10 min, washed with Earle's balanced salt solution and suspended in 9 vol. buffer A (10 mM Tris-HC1/3 mM CAC12/6 mM 2mercaptoethanol, pH 8.0). Then they were gently homogenized in a hand-driven loosely-fitting Teflon-glass homogenizer, and the homogenate was centrifuged at 250 X g for 10 min. The precipitated nuclei were washed twice by gentle dispersion and centrifugation in 4.5 and 9 vol. buffer A. Finally, the nuclei were suspended in buffer B (10 mM T r i s - H C 1 / l m M E D T A / 4 m M MgC12/6mM 2mercaptoethanol, p H 8.0) containing 0.25 M sucrose at 0.5-1 • 107 nuclei/ml. The supernatant of the cell homogenate and the first washing fluid were combined and used as cell extracts. [3H]Uracil-labeled nuclei, chromatin and D N A .

The freshly prepared cell nuclei described above were incubated at 37°C for 60 min in solution containing final concentrations of 40 mM Tris-HC1 (pH 8.0)/5 mM MgCI2/0.3 mM E D T A / 2 mM 2-

mercaptoethanol/80 m M KC1/2.5 mM A T P / 3 0 /~M each of dATP, dCTP and d G T P / 2 . 5 /~M [uracil-3H]dUTP (2 C i / m m o l , RCC Amersham)/5 mM each of uracil and d U M P [8]. The reaction was stopped by cooling the mixture in an ice-water bath. Nuclei were washed three times with buffer B by centrifugation and suspended in a 1 : 1 mixture of buffer A and glycerol at 3-5 • 108 nuclei/ml. The suspension was stored at - 2 0 ° C . This nuclear preparation is designated as [3H]Ura-DNAn. The suspension of [3H]Ura-DNAn (0.85 mg D N A in 300/~1) was diluted with 0.5 vol. buffer A, centrifuged at 20000 rpm for 20 min at 4°C, resuspended in 200 /~1 buffer C (10 mM TrisHC1/0.7 mM EDTA, p H 7.6) supplemented with 5% sucrose and 0.6M NaC1 and gently stirred overnight at 0°C. Volumes of 50/xl of this suspension were loaded on 4 ml of a linear gradient of 5-20% sucrose in buffer C containing 0.6 M NaC1 layered on 0.4 ml 60% sucrose in the same buffer, and centrifuged at 50000 rpm for 6 h at 4°C in a Hitachi RPS-56T rotor. The bottom half of the gradient was collected, dialysed against buffer C, concentrated by ultrafiltration through an Amicon PM 10 membrane filter and used as chromatincontaining uracil-labeled DNA ([3H]Ura-DNAc). Then [3H]uracil-labeled free DNA was prepared from the [3H]Ura-DNAn by the method of Gross-Bellard et al. [9] and designated as [3H]UraDNA. Another [3H]uracil-containing DNA prepared from PBS1 phage as described previously [1] is designated as [3H]Ura-q~DNA. Nucleosome monomers and dimers. Freshly prepared NL3 cell nuclei, suspended at a concentration of 2- 108 nuclei/ml in a solution of 10 mM Tris-HC1 (pH 7.2)/0.75 mM CAC12/0.3 M sucrose, were digested with 100 units/ml micrococcal endonuclease (EC 3.1.31.1, Boehringer) for 10 min at 37°C. The reaction was stopped by adding onetenth vol. 0.1 M EDTA (pH 7.0) and the digest was gently dispersed in a hand-driven homogenizer in the cold and centrifuged at 10000 X g for 15 min. The supernatant, which contained 1.5-2.5 mg DNA, was loaded on a Sepharose 6B column (2.5 X 44 cm) equilibrated with buffer C and eluted with the same buffer [cf. 10]. Nucleosome monomer and dimer fractions were collected separately and concentrated by ultrafiltration

17 through Amicon PM 10 membrane filters. Each concentrate (0.3-0.5 mg DNA) was loaded on a linear gradient of 5-20% sucrose in buffer C (32 ml), centrifuged at 24000 rpm for 22 h at 4°C in a Hitachi RPS-25 rotor, and fractionated. This centrifugation was repeated once for the monomer preparation, and twice for the dimer preparation. Free DNA was prepared as described above from a portion of the nucleosome monomer preparation. Methylation of DNA and nucleosomes. D N A of NL3 cells, DNA of the nucleosome monomers described above, nucleosome monomers and dimers, and poly(dA-dT) (Boehringer) were methylated at a concentration of 0.5-1.0 mg D N A / m l with 17 m M [14C]dimethyl sulfate (60 mCi/mmol, RCC Amersham) in 0.25 M sodium cacodylate (titrated with HC104 to p H 7.0) containing 1 mM EDTA, for 6 h at room temperature [11]. These methylated preparations were purified on Sephadex G-50 columns (0.9 × 29 cm) by elution with 10 mM Tris-HC1 (pH 8.0)/0.25 mM EDTA. These labeled preparations are designated as [lac]MeD N A , [14C]Me-DNA150, [ 1 4 C ] M e - D N A m , [14C]Me-DNAa and [lac]Me-poly(dA-dT), respectively. Reconstitution of chromatins. Chromatins and nucleosome monomers were reconstituted from [3H]Ura-DNA, [3H]Ura-d~DNA, [14C]Me-DNA, [ 14 C]Me_DNA15 ° and [14C]Me-poly(dA-dT) as described previously [1]. Briefly, D N A and a freshly separated calf thymus chromosomal protein fraction were mixed in a weight ratio of 1:1.2 in a solution of 10 mM Tris-HC1 (pH 8.0)/0.2 mM dithiothreitol/0.2 mM phenylmethylsulfonyl fluoride, containing 2 M NaC1, and dialyzed against the same solution with a decreasing concentration of NaC1 to zero, and finally against 0.25 mM EDTA (pH 8.0). The dialyzed solution was centrifuged to remove insoluble material and concentrated by ultrafiltration through an Amicon PM 10 filter. These reconstituted chromatins are designated as [3H]Ura-DNArc, [3H]Ura-q, DNArc, [14C]Me-DNAr~ and [14C]Me-poly(dA-dT)r~, respectively. Enzymatic release of abnormal bases. All D N A preparations except [3H]Ura-DNAn were incubated with the cell extract (1.2 mg protein) in 250 /~1 100 mM Tris-maleate buffer (pH 7.5)/50

m M EDTA at 37°C for the indicated periods. The reaction was stopped by cooling the mixture in an ice bath, and released bases were promptly analyzed as described below. In the cases for [ 3H]Ura-DNAs, 5 % trichloroacetic acid-soluble radioactivity was determined. Enzymatic release of uracil from [3H]Ura-DNAn was examined in buffer A to protect the nuclear structure, but results were the same as in Tris-maleate-EDTA, in which the structure was rapidly destroyed. Base analysis. For determination of the base composition of DNA, uracil-containing D N A was hydrolyzed with 99% formic acid in sealed tubes at 176°C for 30 rain, and methylated D N A was hydrolyzed with 1 M HC1 at 100°C for 10 min. Bases in these hydrolyzates and the enzymatic reaction products were separated and identified by A G 50W (Bio-Rad Laboratory) column chromatography as described previously [1]. Radioactivities were determined with a Beckman liquid scintillation spectrometer LS 150. Polyacrylamide gel electrophoresis. Nucleotide chain lengths of nucleosome fractions were determined by electrophoresis on a 5% polyacrylamide gel slab as described by Loening [12], with HaelII restriction fragments of PM2 D N A as references [13]. Others. Protein was determined by Hartree's modification of Lowry's method [14] with bovine serum albumin as a standard. D N A was estimated spectrophotometrically assuming A260 = 20 c m - 1 . mg l. ml. Results

Properties of DNA substrates. The properties of [3H]uracil-containing DNAs and [laC]methylated DNAs are summarized in Table I. It should be noted that only Ura-~DNA contains about 70% bp each composed of a normal and an abnormal base, the latter being removable with uracil-DNA glycosylase. All other DNAs contain less than 0.1% excisable abnormal bases. Nucleosome monomers and dimers were separated from a partial digest of NL3 cell nuclei by Sepharose 6B column chromatography and purified by two and three runs, respectively, of sucrose density gradient centrifugation. The profiles in the last centrifugation of nucleosome monomer and

18 TABLE I CHARACTERISTICS OF LABELED D N A SUBSTRATES Base compositions were calculated from the specific radioactivities of [ 3H]dUTP or [ laC]dimethyl sulfate and the labeled products, except those for phage DNA, which were given by Takahashi and Marmur [15]. P r o t e i n / D N A weight ratios: (c), native chromatins (nucleosome oligomers) labeled with [14 C]dimethyl sulfate; (rc), reconstituted chromatins from [ 3H]uracil-containing or [14C]methylated DNAs; (rm), reconstituted nucleosome monomer from [14C]Me-DNA150. Specific radioactivity (cpm//~g DNA)

[ 3H]Ura-DNA [ 3H]Ura-q, D N A

Distribution of radioactivity (%)

89 2000 1-MeAde

[ 14C]Me-DNA [14C]Me-DNAls0 [14C]Me-DNAm [ 14 C]Me_DNA d [ 14 C]Me-poly(dA-dT)

210 340 590 a 490 a 54

14

Base composition (mol%)

Protein/ DNA (w/w)

Ura

Cyt

Ura

Cyt

100 67 3-MeAde 10 13 22 14 72

0 30 7-MeGua 80 79 69 79

0.0021 35.9 3-MeAde 0.011 0.023 0.067 0.035 0.020

14.7 7-MeGua 0.088 0.14 0.21 0.20

1-MeAde

0.004

0.84 (rc) 0.87 (rc) 1.11 0.93 1.12 1.31 1.17

(rc) (rm) (c) (c) (rc)

a Specific radioactivity after protein was removed.

dimer fractions are shown in Fig. 1. The respective peak fractions, a and b in the figure, were collected and concentrated. DNAs of the monomer and dimer preparations had chain lengths of 140-

!

I

I

I

E

TABLE II

t-

RELEASE O F URACIL

O

to 2.0 (kl

W

~ Z

m Q: 0 II1

160 and 280-320 bp, respectively, as judged by polyacrylamide gel electrophoresis (Fig. 2). In the same gel poly(dA-dT) migrated as a broad band but its mean length was clearly more than 1000 bp. Release of uracil. Human lymphoblastoid cell

The amounts of Ura-DNAs and Ura-q, DNAs in the assay mixture were 11 and 0.25 /tg, respectively. Acid soluble radioactivity was exclusively that of uracil as judged by column chromatography.

b 1.0

Substrate

Uracil content (pmol//t g DNA)

Acid soluble radioactivity (%) a Incubation time (min)

0

i

0

I0

20

FRACTION

30

40

NUMBER

Fig. 1. The sedimentation profiles of nucleosome monomer and dimer fractions. These two fractions separated with a Sepharose 6B column from nuclease digests of NL3 cell nuclei were centrifuged repeatedly in the sucrose density gradient in separate tubes. Monomer (©) and dimer (O) profiles shown are of the second and third runs, respectively. Sedimentation from fight to left. For details refer to 'Materials and Methods'.

60 [ 3 H]Ura-DNA

[3H]Ura-DNAn [ 3H]Ura-DNA¢ [3H]Ura-DNArc [ 3H]Ura-,~DNA [ 3H]Ura-,~DNArc

0.069 0.069 0.069 0.069 1 180 1 180

93 36 38 34 91 93

"Percent of total radioactivity of uracil residues in DNA.

120

44 41 36

19

M2MI

!

DI D2 i

!

!

I

lO0

0

5o

,

0

,

60

,

120

INCUBATION TIME (rain) Fig. 3. Release of acid soluble radioactivityfrom [3H]Ura-DNA (A) and [3H]Ura-DNAn(O) by cell extracts. Conditions are described under 'Materials and Methods'. Closed symbols are those for boiled extracts.

Fig. 2. Polyacrylamidegel electrophoretic profiles of doublestranded DNAs from nucleosome monomer (Ml) and dimer (D~) preparations, collected from fractions a and b in Fig. 1, respectively. M E and D2 are for another pair of preparations. HaelII restriction fragments of PM2 DNA were co-electrophoresed as references of chain lengths, which are indicated as numbers of basepairs [13]. Ethidium bromide stain.

extracts released more than 90% of the radioactivity from [3H]Ura-DNA in a trichloroacetic acidsoluble form in 30 min, but less than 50% from [3H]Ura-DNAn in 2h, as shown in Fig. 3 and Table II. During the incubation of [3H]Ura-DNAn in buffer A, the shape of nuclei was kept apparently intact under a microscope. Ura-DNA~ and Ura-DNAr~ were also as resistant as Ura-

D N A n to hydrolysis by cell extracts. The radioactivity released was almost exclusively recovered in the position of free uracil from an A G 50 W column. Boiled extracts did not have any appreciable activity. These limited releases of uracil from Ura-DNAn, U r a - D N A c and Ura-DNArc are in striking contrast with that from Ura-~DNArc, which was confirmed to be highly sensitive to the activity of cell extracts to release uracil as reported previously (Table II) [1]. Release of 3-methyladenine. The time courses of release of 3-methyladenine by cell extracts from M e - D N A and Me-DNArc were similar to those of release of uracil from U r a - D N A and U r a - D N A re, respectively. Chromatin or nucleosomal D N A s are obviously more resistant than the respective D N A s of the free form, as shown in Table III. It was also noticed that M e - D N A m, Me-DNAr~n and MeD N A d were more resistant than Me-DNAr~ , and the same was true for the respective free-form DNAs, Me-DNAI5 0 and Me-DNA. Me-poly(dAdT)r c and Me-poly(dA-dT) were intermediate in resistance between these two extremes of complexed and free DNAs, respectively. 1-Methyladenine was as resistant to the enzymatic excision as 7-methylguanine [1].

20 TABLE III RELEASE OF 3-METHYLADENINE Substrate

3-methyladenine content (pmol/t~g DNA)

Amount of DNA/assay (/~g)

3-Methyladenine released (%) Incubation time (min) 60 120

[ 14C]Me-DNA [ 14C]Me-DNArc [ 14 C]Me-DNA15o [ 14C]Me-DNArm [ 14C]Me-DNAm [ 14 C]Me.DNAd [ 14C]Me-poly(dA-dT) [ 14C]Me_poly(dA_dT)rc

0.35 0.35 0.77 0.77 2.2 1.1 0.65 0.65

5.4 5.4 6.2 6.2 7.1 7.1 19 19

90 45 57 14 12 20 78 28

Discussion In this paper we showed that chromatin composed of dUMP-misincorporated D N A (UraD N A c ) was resistant to uracil-DNA glycosylase to a similar degree, as chromatin composed of methylated D N A (Me-DNArc) to 3-methyladenineD N A glycosylase (Tables II and III). This is in marked contrast to chromatin reconstituted with PBS1 phage DNA~ wtfich is highly susceptible to the enzyme. One possible explanation of this difference would be as follows. Because 36% of the bases in the phage D N A are uracil [15], more than two-thirds of the basepairs are A-U and because uracil-DNA glycosylase acts more readily on single-stranded than on double-stranded D N A [1], excision of uracil could spread easily to neighboring basepairs. This may start in internucleosomal regions, but could proceed into core particle regions as basepairing is loosened stepwise. In human cell nuclei or chromatin from these, on the other hand, the content of misincorporated uracil is very low (Table I), so it is unlikely to weaken the double strandedness of D N A for the glycosylase action. Thus, the nucleosome may be the structure responsible for protection from the enzyme action. The second possibility that U r a - ~ D N A does not form an exact chromatin structure is not ruled out by the fact that Ura-¢DNAr~ is slightly more susceptible to micrococcal nuclease than native chromatin [ 1]. The third possibility that Ura-DNA, but not Ura-e?DNA, is protected by some proteins intrinsic to chromatin structure and resistant to

54 17 17 29 39

separation from D N A during hypertonic treatment, can not be excluded, although it is not very likely because 'free' U r a - D N A was as sensitive to the D N A glycosylase as Ura-~bDNA. Chromatins reconstituted from methylated D N A or poly(dA-dT), in which the 3-methyladenine content was less than 0.1 mol%, were more resistant to excision of 3-methyladenine than the respective free forms of polydeoxynucleotides as in the case of dUMP-misincorporated D N A (Ura-DNAn, U r a - D N A c and Ura-DNArc vs. UraDNA). Because 3-methyladenine should be randomly distributed in Me-DNA, the resistant fraction of 3-methyladenine residues in Me-DNArc should indicate the proportion of regions in D N A protected by the chromatin structure. The low values for release of 3-methyladenine from Me-DNAm, Me-DNArm and M e - D N A d may be ascribed to the presence of few if any naked D N A regions in these structures. On the other hand, reconstituted chromatin from methylated D N A of high molecular weight possibly contains internucleosomal regions, resulting in more release of 3-methyladenine (Table III). This is consistent with results on Me-poly(dA-dT)rc. Among free DNAs, those with shorter chains seemed more resistant to 3-methyladenine-DNA glycosylase. It is possible that 3-methyladenine residues near the end of the D N A chain are resistant to the glycosylase because oligonucleotides and single strand D N A s are more resistant to the enzyme [1], resulting in lower activity on methylated D N A of nucleosome monomer size.

21

Thus, it may be concluded that uracil- and 3-methyladenine-DNA glycosylases excise the respective abnormal bases mainly from internucleosomal regions, but little if at all from core particle regions when abnormal bases are sparse in the DNA chain. These results suggest that the first step of base excision repair requires redistribution of nucleosomes along the DNA chain for its completion [16-18], although the possibility of very slow processing on nucleosome core particles after fast processing in internucleosomal regions is not ruled out from the present results. The protection of DNA against DNA glycosylases by chromatin structure should be further analyzed in detail using DNA of defined structure and its reconstituted chromatin.

Acknowledgment This study was supported by Grants-in-Aid for Cancer Research (401508, 501508 and 501070) from the Ministry of Education, Science and Culture of Japan.

References 1 Ishiwata, K. and Oikawa, A. (1979) Biochim. Biophys. Acta 563, 375-384

2 Lindahl, T. (1979) Prog. Nucleic Acid Res. Mol. Biol. 22, 135-192 3 Haseltine, W.A., Gordon, L.K., Lindan, C.P., Grafstrom, R.H., Shaper, N.L. and Grossman, L. (1980) Nature 285, 634-641 4 Cleaver, J.E. (1977) Nature 270, 451-453 5 Smerdon, M.J., Tlsty, T.D. and Lieberman, M.W. (1978) Biochemistry 17, 2377-2386 6 Bodell, W.J. (1977) Nucleic Acid Res. 4, 2619-2628 7 Tohda, H., Oikawa, A., Katsuki, T., Hinuma, Y. and Seiji, M. (1978) Cancer Res. 38, 253-256 8 Ishiwata, K. and Oikawa, A. (1981) Nucleic Acid Res., S10, 149-152 9 Gross-Bellard, M., Oudet, P. and Chambon, P. (1973) Eur. J. Biochem. 36, 32-38 10 Davie, J.R. and Candido, E.P.M. (1977) J. Biol. Chem. 252, 5962-5966 11 Uhlenhopp, E.L. and Krasna, A.I. (1971) Biochemistry 10, 3290-3295 12 Loening, U.E. (1967) Biochem. J. 102, 251-257 13 Kovacic, R.T. and Van Holde, K.E. (1977) Biochemistry 16, 1490-1498 14 Hartree, E.F. (1972) Anal. Biochem. 48, 422-427 15 Takahashi, I. and Marmur, J. (1963) Nature 197, 794-795 16 Smerdon, M.J. and Lieberman, M.W. (1978) Proc. Natl. Acad. Sci. U.S.A. 75, 4238-4241 17 Smerdon, M.J. and Lieberman, M.W. (1980) Biochemistry 19, 2992-3000 18 Bodell, W.J. and Cleaver, J.E. (1981) Nucleic Acid Res. 9, 203-213