THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 252, No. 9, Issue of May 10, pp. 3003-3006, 1977

Printed LR U.S.A.

DNase-sensitive Sites in Nucleosomes THEIR RELATIVE SUSCEPTIBILITIES DEPEND ON NUCLEASE USED

(Received for publication, December 6, 1976)

JAMES P. WHITLOCK,JR.,GEORGE W. RUSHIZKY, AND ROBERT T. SIMPSON

From the Developmental Biochemistry Section, Laboratory ofNutrition and Endocrinology, National Institute of Arthritis, Metabolism, and Digestive Diseases, Bethesda, Maryland 20014

We have used three endonucleases having different cata- lytic and physicochemical properties to digest HeLa nucleo- somes (chromatin core particles) which had been labeled with :?lP at their 5’-DNA termini. Each endonuclease nicks nucleosome DNA at the identical sites, supporting the idea that the conformation of the DNA within a nucleosome is the major factor influencing its nuclease susceptibility and indicating that nucleases can indeed yield important infor- mation as to nucleoprotein structure. On the other hand, the relative susceptibility of a given site can differ for each nuclease, indicating that enzyme-substrate interactions unique for each enzyme influence the course of the reaction; this limits the structural information which can be obtained by using a single nuclease to study nucleoprotein structure.

Nucleases have been important tools for studying nucleopro- tein structure. For example, nuclease digestions of chromatin have provided much of the chemical evidence leading to the concept that chromatin is composed of repeating subunits (I-4) and nucleases have been used to investigate the internal architecture of the chromatin subunit (5-8). Other studies using nucleases suggest that regions of chromatin which are active in transcription have a different structure than tran- scriptionally inactive regions (g-11). However, experiments involving nuclease digestion of a nucleoprotein must be inter- preted with some caution, since (a) digestion destroys part of the nucleoprotein and thus may alter the properties of the remaining portion, and (b) the factors which influence the susceptibility of a given site to nucleolytic cleavage are not well understood. For example, the conformation of the DNA within the nucleosome is generally assumed to be the major factor influencing its nuclease susceptibility; however, it is unknown to what extent other factors, such as the catalytic

and physicochemical properties of the enzyme, affect the rate of cleavage at a given nuclease-susceptible site.

The availability of (a) methods for preparing a homogene- ous population of nucleoprotein particles (nucleosome cores) (121, and (b) techniques for determining both the location and the relative susceptibility of nuclease-sensitive sites within the particles (8) have allowed us to ask whether the nuclease susceptibility of a given site is a function only of the conforma- tion of the nucleoprotein, or, on the other hand, if susceptibil- ity also is influenced by nuclease-substrate interactions differ- ent for each enzyme. To study this question, we have used

three endonucleases with different catalytic, or physicochemi- cal properties, or both, to digest the same nucleoprotein sub- strate. The results show that each nuclease cleaves the DNA within the nucleosome at the identical sites. This finding indicates that the major factor influencing the nuclease sus- ceptibility of nucleosome DNA is, in fact, the conformation in which it is held by the nucleosomal proteins. On the other hand, the relative susceptibility of a given site along nucleo- some DNA is not the same for each enzyme. Our findings thus indicate that the conformation of the substrate is not the only factor influencing the nuclease susceptibility of the DNA within a nucleoprotein; the rate of cleavage of a given site also is a function of both the catalytic and physicochemical proper- ties of the particular nuclease used.

MATERIALS AND METHODS

Enzymes- DNase I (DPFF), DNase II (HDAC), and staphylococ- cal nuclease (NFCP) were from Worthington Biochemical Corp. The Aspergillus endonuclease was purified as described.’ Polynucleotide kinase was from P. L. Biochemicals, Inc.

Cell Culture- HeLa S3 cells were maintained in logarithmic growth as previously described (13).

Preparation of Nucleosomes - Nuclei were prepared by homogeni- zation of cells in the presence of 1% Triton X-100 (141, as previously described (13). Nucleosome cores were prepared from staphylococcal nuclease digests of nuclei as previously described (12). Nucleosome cores were concentrated by centrifugation, dialyzed to 0.25 mM EDTA, pH 7.0, and frozen until use. These nucleoprotein particles contain a 140 base pair length of DNA free of internal nicks, no histone Hl, two each of the four smaller histones, some non-histone proteins, and have a sedimentation coefficient of 11 S (12, 15).

Labeling ofNucleosome Cores with :j2P- The DNA of intact nucleo- some cores was labeled at its 5’ termini using polynucleotide kinase and [+y-3”PlATP (New England Nuclear) as previously described (8).

DNase Digestions-All digestions were performed at 37” at a nu- cleosome concentration of 10 AzGu units/ml. DNase II digestions were either in 0.5 rnM EDTA, 5 rnM NaOAc, pH 5.0, or in 0.5 rnM MgCl,, 5 rnM NaOAc, pH 6.5 (see text). Digestions with the Aspergillus endo- nuclease and DNase I were in 0.5 rnM MgCl,, 5 rnM NaOAc, pH 6.5. Digestions were stopped by adding 10% SDS* to a final concentration of 1%.

Analysis of DNA Fragments- DNA was purified by extraction of nucleosomes with phenol/SDS, as previously described (16). DNA fragments were analyzed in 12% polyacrylamide gels containing 7 M

urea (17) using Tris/borate/EDTA buffer system (18). Gels were stained with Stains-All, destained in running water, and scanned, as previously described (12). DNA fragment sizes were estimated by the

1 G. W. Rushizky, and J. P. Whitlock, Jr., manuscript in prepara- tion.

L The abbreviation used is: SDS, sodium dodecyl sulfate.

3003

by guest on March 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

3004 Nucleases and Nucleosome Structure

method of Maniatis et al. (17). Autoradiography was as previously described (18).

Analysis ofProteins- Samples were made 1% in SDS and 1% in p- mercaptoethanol and analyzed in discontinuous SDS-polyacrylamide gels using the buffers previously described (19). Gels were stained with Coomassie blue and destained as previously described (12).

Radioactivity Measurements - Aliquots of labeled nucleosomes were added to 100 pg of carrier calf thymus DNA and precipitated with ice-cold 10% trichloroacetic acid for at least 30 min. The precipi- tates were collected on glass fiber filters, washed with 5% trichloroa- cetic acid and ethanol, dried. and hvdrolvzed with NCS:H,O (9:l) for

I ”

at least 2 h. Radioactivity was measured by liquid scintillation counting in toluene containing 0.4% 2,5-diphenyloxazole (PPO) (w/v) and 0.1% acetic acid (v/v).

RESULTS

Our previous experiments demonstrated that nucleosome DNA contains a DNase I-susceptible site at each lo-nucleotide interval and that not all of these sites are equally susceptible to the nuclease. To recapitulate, nucleosomes which contain a 32P label at their 5’ termini are digested with the endonucle- ase, and the DNA is purified and analyzed by polyacrylamide gel electrophoresis and autoradiography. The pattern of bands on the autoradiogram reflects the distance(s) from the 5’ ter- mini at which nucleolytic cleavage occurred and thus reveals the distribution of nuclease-susceptible sites along nucleosome DNA. The intensity of a given band is proportional to the number of DNA fragments it contains and thus reflects the relative susceptibility of that particular site to nucleolytic cleavage (8).

We have used these same techniques to determine the distri- bution and relative susceptibility of sites along nucleosome DNA which are cleaved by DNase II and by an endonuclease purified from Aspergillus oryzae. These enzymes were selected because they differ from DNase I in their catalytic require- ments and physicochemical properties (Table I).

The initial experiments with DNase II were performed at pH 5, in the absence of divalent cations. Fig. 1 shows that, under these conditions, the nucleosome contains a DNase II- susceptible site at each 10 nucleotide interval from the 5’ termini of the DNA. This distribution of nuclease-susceptible sites is identical with that observed for DNase I and demon- strates that both’ DNase II and DNase I do, in fact, cleave nucleosome DNA at the same sites. However, scans of the autoradiograms (Fig. 2) indicate that there are only small differences in the relative susceptibility of each site to cleav- age by DNase II; this contrasts with previous findings, which demonstrated substantial differences in the relative suscepti- bility of these same sites to cleavage by DNase I (8). In particular, sites 60 to 90 nucleotides from the 5’ termini were shown to be relatively resistant to DNase I; these same sites show no particular resistance to cleavage by DNase II. Both enzymes, however, cleave the sites 30 and 110 nucleotides from the 5’ termini relatively slowly.

TABLE I Properties of endonucleases used

Source Beef pancreas (DNase I)

Hog spleen (DNase II)

Aspergillus ory- zae

M, 31,000 38,000 52,000 Isoelectric point 4.7 10.2 9.2 pH optimum 7.0 4.8 8.2 Activators Mg*+ EDTA MgZ+ Product 5’-Phosphate 3’-Phosphate 5’-Phosphate

terminal terminal terminal Reference 20 21 Footnote 1

One possible explanation for these findings was that, under the conditions for digestion with DNase II (that is, pH 5 and 0.5 mM EDTA), the nucleosome had a different conformation than it had under the conditions previously used for digestion with DNase I (pH 8 and 10 mM MgCl,); in this hypothetical different conformation, the nuclease-susceptible sites might have been more equally sensitive. To test this possibility, we digested nucleosomes with DNase I or DNase II under identi- cal conditions (pH 6.5 and 0.5 mM MgCl,). The results (Fig. 3) indicate that differences in the relative susceptibility of cer- tain sites to either DNase I or DNase II persist even when the digestions are performed under identical conditions; DNase I cleaves the sites 60 to 100 nucleotides from the 5’ termini relatively slowly, whereas DNase II cleaves the same sites without difficulty. Thus, we conclude that the conformation of the nucleoprotein substrate cannot be the only factor influenc- ing the relative susceptibility of a given nuclease-sensitive site; the properties of the nuclease must also affect the rate at which a given site is cleaved.

Another possibility was that DNase II was contaminated with protease activity; during digestion with DNase II, prote- olysis might expose sites which previously had been relatively inaccessable to the nuclease. To test this hypothesis, we di- gested nucleosomes with DNase II until 50 to 60% of the DNA was rendered acid-soluble and then analyzed the nucleosomal

ABCDEAB

20-

FIG. 1. DNA fragments generated during DNase II digestion of HeLa nucleosomes containina a 32P label at the 5’ termini. Nucleo- somes (lOA,,, units/ml) in 0.5mM EDTA, 5 rnM NaOAc, pH 5.0, were digested with DNase II (1000 units/ml) at 37” for A, 0 min; B, 1 min (16% of the DNA was acid-soluble); C, 2 min (22% acid-soluble); D, 4 min (31% acid-soluble); E, 32 min (55% acid-soluble). Left, stained polyacrylamide gel; right, autoradiogram.

by guest on March 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Nucleuses and Nucleosome Structure 3005

FIG. 2. Scans of stained polyacrylamide gels (---) and autoradi- ograms (- -1 containing DNA fragments generated during DNase II digestion of HeLa nucleosomes, as described in the legend to Fig. 1. Upper panel, 1-min digest; lower panel, 32.min digest.

proteins by polyacrylamide gel electrophoresis (Fig. 4). The

gels reveal no evidence of protein degradation. In a second type of experiment, nucleosomes containing [‘Cllysine and [“Clarginine-labeled proteins were digested with DNase II until about 50% of the DNA became acid-soluble; none of the radioactivity was rendered acid-soluble under these conditions (data not shown). Analogous experiments with DNase I also failed to reveal detectable protease activity (data not shown). Thus, our findings are not due to artifacts resulting from contamination of either nuclease with protease activity.

L I I40 80 so

NUCLEOTIDES 20

DNase I and DNase II differ in both their catalytic proper- ties (pH optimum, divalent cation requirement) and physico- chemical properties (isoelectric point). The endonuclease from A. oryzae has catalytic properties similar to those of DNase I and an isoelectric point near that of DNase II (Table I). We have used the Aspergillus endonuclease to determine whether the catalytic or the physicochemical properties of the nuclease were more important in influencing the relative susceptibility of nuclease-sensitive sites within the nucleosome. The diges- tions were performed under the same conditions (pH 6.5 and 0.5 mM MgCl,) used for DNase I and DNase II. The results (Fig. 5) indicate that the Aspergillus endonuclease nicks nu- cleosome DNA at the same sites as does DNase I and DNase II, again indicating that the nucleosome has a limited number of well defined sites susceptible to nucleolytic cleavage. The pattern of susceptibility of these sites to the Aspergillus endo- nuclease has similarities to the patterns of susceptibility to both DNase I and DNase II. For example, the site 80 nucleo- tides from the 5’ end of nucleosome DNA is quite resistant to cleavage by the Aspergillus endonuclease, similar to the re-

sult with DNase I. On the other hand, the remaining sites are more uniformly susceptible to the Aspergillus enzyme, a find- ing similar to that for DNase II. Thus, these results suggest that both the catalytic and the physicochemical properties of a nuclease can influence the rate of cleavage of a nuclease- susceptibile site within a nucleoprotein.

FIG. 3. DNA fragments generated during digestion of HeLa nu- cleosomes with either DNase I or DNase II under the same condi- tions of pH and ionic strength. Upper panel, DNase II (0.5 mM EDTA, 5 ITIM NaOAc, pH 5.01, 46% of the DNA was acid-soluble; middle panel, DNase II (0.5 rnM MgCl,, 5 rnM NaOAc, pH 6.51, 49% acid-soluble; lower panel, DNase I (0.5 mM MgCl,, 5 mM NaOAc, pH 6.51, 29% acid-soluble. -, scan of the stained gel; - - -, scan of the autoradiogram.

DISCUSSION

These studies provide strong evidence that nucleosomes (and, potentially, other nucleoprotein molecules) have a lim- ited number of well defined sites which are susceptible to nucleolytic cleavage, since each of three endonucleases with different properties cleaves the nucleosome at the identical sites. This finding implies that the major factor influencing the overall nuclease susceptibility of nucleosome DNA is, in fact, the conformation in which it is held by the nucleosomal proteins, and supports the concept that nuclease digestion of a nucleoprotein can yield valuable information regarding nu- cleoprotein structure.

On the other hand, even though each nuclease nicks nucleo- some DNA at the same sites, the relative susceptibility of these sites differs for each enzyme. Our findings show that DNase I is much more effective in detecting regional differ- ences in nuclease sensitivity along nucleosome DNA than is

by guest on March 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

3006 Nucleases and Nucleosome Structure

,H3 -H2B ‘H2A ‘H4

FIG. 4. Absence of protease activity in DNase II. HeLa nucleo- somes were digested with DNase II until 55% of the DNA was acid- soluble. The dieest was made 1% in SDS. 1% in S-mercantoethanol. 0.25 M in sucrose, 10 rnM in phosphate buffer, pH 7.2, heated to 100”; and analyzed in a discontinuous SDS-18% oolvacrvlamide gel. Mi- gration was from top to bottom. Left, undigested nucleosomei; right, digested nucleosomes.

DNase II. The use of either nuclease alone to digest nucleo- somes would lead to different predictions as to nucleosome structure. Therefore, our findings also reveal a limitation on the value of nucleases in studying nucleoprotein structure, namely, that the pattern of DNA fragments which is gener- ated reflects not only the conformation of the nucleoprotein but also enzyme-substrate interactions which are unique for the individual nuclease. Thus, the use of several nucleases to digest a nucleoprotein may yield more complete, detailed, and unbiased structural information than can be obtained if only one is used.

Both DNase I and the Aspergillus endonuclease, which require divalent cations for activity, cleave the site 80 nucleo- tides from the 5’ termini relatively slowly; DNase II, on the other hand, which does not require divalent cations, cleaves this same site without difficulty. This finding may mean that at this site magnesium ion cannot interact with the DNA to form the correct complex required for catalysis. The sites 30 and 110 nucleotides from the 5’ termini are somewhat resist- ant to cleavage by all three nucleases; this may reflect steric hindrance by nucleosomal proteins.

These studies also illustrate the value of using a homogene- ous, well characterized nucleoprotein population to study var- ious nucleases and the factors which affect their activity. Nucleases, polymerases, and regulatory proteins must inter- act with chromatin in vivo; knowledge of how their physico- chemical, or catalytic properties, or both, influence their func- tional interaction with chromatin may lead to a better under-

FIG. 5. DNA fragments generated during digestion of HeLa nu- cleosomes with an endonuclease from Aspergillus oryzae. The diges- tion was performed in 0.5 mM MgCl*, 5 mM NaOAc, pH 6.5. Eighteen per cent of the DNA was acid-soluble. -, scan of the stained gel; --- , scan of the autoradiogram.

standing of how these proteins are involved in the specific expression of genetic information.

Acknowledgments-We thank Ms. Linda Propst for photo- graphic assistance and Ms. Carol Sartain for secretarial assist- ance.

REFERENCES

1. Hewish, D. R., and Burgoyne, L. A. (1973) Biochem. Biophys. Res. Commun. 52, 504-510

2. Noll, M. (1974) Nature 251, 249-251 3. Axel, R. (1975) Biochemisty 14, 2921-2925 4. Sollner-Webb. B.. and Felsenfeld. G. (1975) Biochemistry 14,

2915-2920 5. Noll, M. (1974) Nucleic Acids Res. 1. 1573-1578 6. Lacy, E., and Axel, R. (1975) Proc. k&l. Acad. Sci. U. S. A. 72,

3978-3982 7. Shaw, B. R., Herman, T. M., Kovacic, R. T., Beaudreau, G. S.,

and Van Holde, K. E. (1976) Proc. N&l. Acad. Sci. U. S. A. 73, 505-509

8. Simpson, R. T., and Whitlock, J. P., Jr. (1976) Cell 9, 347-353 9. Gottesfeld, J. M.. Muruhv, R. F., and Bonner, J. (1975) Proc.

Natl. A&d. Sci’. U. s‘. A: 72, 4464-4408 10. Gottesfeld, J. M., Bagi, G., Berg, B., and Bonner, J. (1976)

Biochemistrv 15. 2472-2483 11. Weintraub, H:, and Groudine, M. (1976) Science 193, 848-856 12. Whitlock, J. P., Jr., and Simnson, R. T. (1976) Nucleic Acids

Res. 3,.2255-i266 _

13. Whitlock, J. P., Jr., and Simpson, R. T. (1976) Biochemistry 15, 3307-3313

14. Hymer, W. C., and Kuff, E. L. (1964) J. Histochem. Cytochem. 12. 359-363

15. Simpson, R. T., and Bustin, M. (1976)Biochemistry 15,4305-4312 16. Simpson, R. T., and Whitlock. J. P., Jr. (1976) Nucleic Acids

R&. 3; 117-127 17. Maniatis, T., Jeffrey, A., and van desande, H. (1975) Biochemis-

try 14, 3787-3794 18. Peacock, A. C., and Dingman, C. W. (1967) Biochemistry 6,1818-

1827 19. Le Stourgeon, W. M., and Rusch, H. P. (1973) Arch. Biochem.

Biophys. 155, 144-158 20. Laskowski. M.. Sr. (1971) in The Enzvmes (Bover. P. D.. ed) 3rd

Ed, Vol.‘4, pp. 289-311, Academic Press, New York. 21. Bernardi, G. (1971) in The Enzymes (Boyer, P. D., ed) 3rd Ed,

Vol. 4, pp. 271-287, Academic Press, New York

by guest on March 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

J P Whitlock, Jr, G W Rushizky and R T Simpsonnuclease used.

DNase-sensitive sites in nucleosomes. Their relative suspectibilities depend on

1977, 252:3003-3006.J. Biol. Chem. 

  http://www.jbc.org/content/252/9/3003Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/252/9/3003.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from