Proc. Nat. Acad. Sci. USA Vol. 69, No. 8, pp. 2115-2119, August 1972 Blocking by Histones of Accessibility to in Chromatin (/RNA polymerase/ polymerase) ALFRED E. MIRSKY AND BERT SILVERMAN The Rockefeller University, New York, N.Y. 10021 Contributed by Alfred E. Mirsky, June 1, 1972 ABSTRACT The effect of histones on accessibility of to in chromatin of thymus nuclei has been studied by selective extraction of either lysine-rich or arginine-rich histones. It was found that all histones block accessibility but that, weight for weight, lysine-rich histones block much more effectively than do arginine-rich histones. We point to the contrast between accessibility of to and of to RNA polymerase, and to what may be the similarity between accessibility to and polymerase. of chromatin in the cell nucleus is to a large extent covered with protein. This is clearly shown by use of as a probe. This probe also shows that the combination between protein and is loose, for as the concentration of enzyme and the time of reaction are increased, more and more becomes accessible (1). We report here experiments concerning the role of histones in determining the accessibility of. Lysine-rich and arginine-rich histones are compared as blockers of accessibility to. Our procedure is to extract selectively either lysine-rich or argininerich histones from nuclei. Accessibility of to is measured before extraction and after extraction. In these experiments, we find that all histones block accessibility but that, weight for weight, lysine-rich histones block much more effectively than do arginine-rich histones. Lysine-rich histones are completely extracted from nuclei by 0.1 M citric acid; no more than traces of arginine-rich histones and other proteins are removed by this procedure. Arginine-rich histones are extracted stepwise by alcoholhydrochloric acid solutions. If HCl is held constant at 0.07 N, and the alcohol concentration is lowered from 86 to 75%, the amount of histone extracted increases; if the alcohol concentration is held constant at 80%/0, while the concentration of acid is raised from 0.03 to 0.14 N, the amount of histone extracted also increases. By either procedure, only about 7% of the extracted histone is of the lysine-rich type, and no more than a trace of nonhistone protein is removed. These selective extraction procedures have been used in this laboratory before (2, 3); those involving the use of alcohol- HCl were derived from a procedure of Johns et al. (12). The effect of histone removal on the activity of dependent RNA polymerases has been studied a number of times, in most cases on chromatin extracted from nuclei rather than, as in the present experiments, on chromatin in the nucleus (4-6). It is of interest to compare the accessibility of in chromatin as reported from experiments on RNA polymerase with accessibility of as revealed by the action of. 2115 MATERIALS AND METHODS Nuclei were prepared from calf thymuses which, chilled in ice, reached the laboratory within 1.5 hr after excision. Nuclei were isolated in 0.25 M sucrose-3 mm CaCI2 or in 0.01 M citric acid by methods that have been described (7, 8), and were stored at -60. The used was Kunitz's preparation (9) from pancreas, as supplied by Worthington. Lysine-Rich Histone Was Extracted from the citric acid nuclei with 0.10 M citric acid-0.12 M NaCl. Nuclei containing about 270 mg of were extracted twice by stirring in the cold for 30 min, each time with 45 ml. After centrifugation, the combined supernates were dialyzed against water to remove NaCl, and the histone was precipitated at room temperature with 10 volumes of acetone that contained 1 ml of HCl in 2 liters. The precipitated histone was washed in acetone, then in ether, dried in a vacuum dessicator, and weighed. The nuclear residue was repeatedly washed in the cold with 0.25 M sucrose-a mm MgCl2, and brought to a volume of 18 ml. Nuclei in this suspension were used for experiments on digestion by. An aliquot of the suspension was taken for determination, which was done by extraction in 0.5 M perchloric acid at 700 for 30 min, followed by measurement of the absorbairce at 265 nm. The extracted protein was 22.8% (varying only slightly in several preparations) of the mass of the in the nuclei. Polyacrylamide gel electrophoresis (Fig. 1, tube 2) and spectrophotometry on the stained gel showed that 90%O of the extracted protein was lysine-rich histone, the remainder of the material being mainly arginine-rich histone, so that lysinerich histone extracted was equivalent to 20.6% of the in the nuclei. Practically all the lysine-rich histone was extracted by citric acid, for when arginine-rich histones were subsequently extracted in HCl, gel electrophoresis revealed no more than a trace of lysine-rich histone. Arginine-Rich Histone Was Extracted from sucrose nuclei. Before this was done a mass of nuclei was first thoroughly washed with 80%o alcohol to remove sucrose. Aliquots (containing 195 mg of ) of the nuclear suspension in 80% alcohol were taken for stepwise extraction of histones in alcohol-hcl solutions. Each nuclear sample was centrifuged, and the sediment was brought to 110 ml with alcohol-hcl, stirred rapidly at 40 for 30 min, and then centrifuged. Histone in the supernate was precipitated at room temperature by decantation into 500 ml of acetone, which contained 0.25 ml of HCl. The precipitate was washed in acetone, then in
2116 Biochemistry: Mirsky and Silverman FIG. 1. Histones were analyzed by polyacrylamide.gel electrophoresis in the system + described by Bonner et al. (10); 15% gels containing 6.25 M urea mw - _" were used. The gels were stained with amido black, then destained electrophoretically and 2 3 by diffusion. Histone bands were quantitated with a Zeiss PMQII spectrophotometer equipped with a Vicon gel transport. - Tube I shows 15 jig of calfthymus total histone resolved into three bands: (a) lysine-rich histone I, (b) an unresolved mixture of histones III and II, and (c) histone IV. [The nomenclature is that of Bonner et al. (10)]. Tube 2 has 10,g of lysinerich histone extracted from nuclei by the citric-acid procedure described in the text. Tube 3 represents 10 Mg of histone extracted by the alcohol-hol procedure. Lysine-rich histone is 7% of the total extracted histone, while the arginine-rich histones are extracted in the same proportion to each other as they occur in native total histone. There was no evidence that the arginine-rich histones were fractionated by the ethanol-hc1 extractions in the range of concentrations used in these experiments. Panyim and Chalkley's electrophoretic system (11), which resolves histones III, IIa, and IIb, was used to confirm this observation. Johns et al. (12) reported similar results, using the analysis of N-terminal amino acids as a criterion of the fractionation of arginine-rich histones by the range of ethanol-hcl concentrations used here. ether, dried in a vacuum dessicator, and weighed. The amounts of histone extracted in various alcohol-hcl solutions are given in Table 1. Electrophoresis (Fig. 1, tube 3) and spectrophotometry on the gel showed that there was no nonhistone protein in these preparations. Lysine-rich histone accounted for 7% of the total histone. The nuclear sediment remaining after extraction of histone was washed once with 80% alcohol, and then repeatedly at 40 with 0.25 M sucrose-4 mm CaCl2. This nuclear suspension was used for experiments on digestion by. Effect of on Selectively Dehietonized Nuclei. For each experiment a sample (volume from 0.6 to 0.9 ml) of the extracted nuclei containing 1.03 mg of phosphorus was suspended in 100 ml of 0.25M sucrose-3 mm MgC12-5 mm sodium phosphate buffer (ph 7.0). A 16-ml sample of the suspension was pipetted into each of six tubes, which were TABLE 1. Amounts of total histone and of arginine-rich histone extracted from nuclei in relation to concentrations of alcohol at a constant HCl concentration (0.07 N) Alcohol (%) 75 80 83 84.5 86 Weight of extracted histone (%) 45.6 36.4 29.0 19.1 8.72 Arginine-rich histone (%) 42.4 33.9 27.0 17.8 8.1 The amount of histone extracted from each preparation of nuclei is expressed as a percentage of the amount of present. TABLE 2. Proc. Nat. Acad. Sci. USA 69 (1972) Digestion by of nuclei from which no histone was extracted (Mg/ml) 15 30 45 60 90 120 (%) 4.76 0.690 0.870 1.00 1.02 1.10 1.16 71.5 2.44 0.580 0.740 0.890 0.980 1.05 1.05 64.6 1.23 0.435 0.595 0.698 0.780 0.830 0.910 56.2 0.62 0.360 0.472 0.620 0.700 0.752 0.790 48.8 0.31 0.190 0.296 0.380 0.448 0.535 0.610 37.7 None 0.040 0.040 0.044 0.042 0.045 0.070 0 Digestion follows the same course when nuclei are isolated in sucrose (as above) or in 0.01 M citric acid, or when sucrose nuclei are extracted with 80% alcohol (containing no HCl). In all cases digestion is in the same medium. The difference between this Table and Table 2 in a previous paper (1) is due to the use of different preparations of. The same preparation of enzyme was used in all the experiments reported in this paper. placed in a 260 bath. Graded amounts of were added to the tubes (the largest volume added being 0.8 ml), and they were stirred. At each time indicated in Tables 2-7, 2.0 ml were withdrawn and added to 1 ml of ice-cold 0.50 M perchloric acid. The absorbance at 265 nm was a measure of the quantity of in the nuclei by the. RESULTS AND DISCUSSION The results given in Tables 2-6 show how digestion by proceeds at various concentrations of enzyme. Data for digestion at a single time and for a single concentration of facilitate a comparison of the various nuclear preparations (Table 8). If one compares intact nuclei with nuclei from which all lysine-rich histone has been extracted and those from which about the same amount or twice the amount of arginine-rich histone has been extracted, it is clear that access to by the enzyme is increased by removal of either type of histone, but that, weight for weight, removal TABLE 3. Digestion by of nuclei from which all lysine-rich histone was extracted (20.5% of the amount of present) (.g/ml) 15 30 45 60 90 120 (%) 4.76 1.25 1.30 1.32 1.32 1.34 1.35 83.2 2.44 1.17 1.25 1.28 1.28 1.31 1.32 81.4 1.23 1.12 1.20 1.24 1.24 1.26 1.28 79.0 0.62 1.02 1.10 1.16 1.20 1.22 1.25 77.1 0.31 0.880 1.02 1.08 1.12 1.15 1.18 72.7 None 0.048 0.04 0.050 0.052 0.055 0.060 0
Proc. Nat. Acad. Sci. USA 69 (1972) TABLE 4. Digestion by of nuclei from which argininerich histone was extracted in 84.5% alcohol-0.07 N HCI (the arginine-rich histone extracted was 17.8% of the amount of present) (og/ml) 15 30 45 60 90 120 (%) 4.76 0.710 0.840 0.920 0.980 1.04 1.08 66.6 2.44 0.580 0.730 0.815 0.875 0.920 0.945 58.4 1.23 0.455 0.630 0.732 0.775 0.840 0.890 55.0 0.62 0.325 0.472 0.565 0.615 0.690 0.730 45.1 0.31 0.244 0.370 0.462 0.520 0.625 0.655 40.5 None 0.038 0.038 0.038 0.036 0.042 0.044 0 in Chromatin 2117 TABLE 5. Digestion by of nuclei from which argininerich histone was extracted in 83% alcohol-0.07 N HCl (the arginine-rich histone extracted was 27.0% of the amount of present) (jug/ml) 15 30 45 60 90 120 (%) 4.76 0.860 0.980 1.10 1.18 1.22 1.26 77.6 2.44 0.785 0.890 0.985 1.02 1.05 1.08 66.6 1.23 0.650 0.778 0.860 0.910 0.960 0.970 60.0 0.62 0.510 0.660 0.740 0.782 0.840 0.865 53.4 0.31 0.418 0.560 0.655 0.715 0.770 0.830 51.2 None 0.048 0.052 0.048 0.045 0.041 0.045 0 of lysine-rich histone is decidedly more effective in making accessible to. Experiments, to be reported in another paper, on addition of histones to intact nuclei, to nuclei from which lysine-rich histone has been extracted, or to nuclei from which argininerich histones have been extracted show that addition of any type of histone restricts accessibility to by. As in the experiments reported here, weight for weight, lysine-rich histone when added blocks accessibility much more effectively than does arginine-rich histone; and, furthermore, results with several fractions of arginine-rich histone differ from each other. The data presented here on the differences between lysinerich and arginine-rich histones in blocking accessibility to recall the observations on the fine structure of chromatin made in this laboratory (2, 3) concerning the loosening of chromatin structure by selective extraction of lysine-rich or arginine-rich histones. As in the present experiments, it was found that removal of lysine-rich histone was far more effective than the removal of the same amount of argininerich histone in decondensing chromatin, and also that addition of lysine-rich histone was more effective in condensing chromatin when added to nuclei from which this histone had been selectively extracted. Despite a certain resemblance between the effects of selective histone extraction on. accessibility of to and on the decondensation of chromatin, we do not know whether these two events are really linked. * The difference in the effects of lysine-rich and argininerich histones on the fine structure of chromatin has been "questioned" by Smart and Bonner (13), who point out that "so far as the properties of chromatin which we studied are concerned, all classes of histones appear to contribute nearly equally." Even if this statement holds for the properties studied by them, there is no reason to "question" the differences observed in the structure of chromatin as shown by electron microscopy when lysine-rich and arginine-rich histones are selectively extracted; if the striking differences * Reference should be made to several recent studies concerning the effect of selective histone extraction on decondensation of chromatin (14, 15). in chemical composition between these types of histones are considered, it would indeed be surprising if they were "to contribute nearly equally" to all properties of chromatin. Surely the observations reported in this paper show that lysine-rich and arginine-rich histone are definitely unequal in blocking access of to in chromatin. These observations will now be discussed with reference to reports by other investigators on differences between these two classes of histones in blocking access of another enzyme, RNA polymerase, to. There have been three studies of selective extraction of histones from dissolved thymus chromatin, and subsequent reaction of the extracted chromatin with -dependent RNA polymerase (4-6). These experiments deal with the relative effects of removal of lysine-rich and arginine-rich histones on the template activity of. The reports of the three studies do not agree and we can now consider only one of them. The results reported by Spelsberg and Hnilica (5) indicate that removal of lysine-rich histone did not significantly change "the templating properties of", but that removal of arginine-rich histone brought about a considerable change; in the terms used in the present paper, TABLE 6. Digestion by of nuclei from which argininerich histone was extracted in 80% alcohol-0.07 N HCl (the arginine-rich histone extracted was 33.9% of the amount of present) (Ag/ml) 15 30 45 60 90 120 (%) 4.76 1.08 1.20 1.28 1.34 1.33 1.34 82.6 2.44 0.830 1.05 1.11 1.15 1.17 1.21 74.5 1.23 0.775 0.950 1.01 1.12 1.17 1.17 72.1 0.62 0.620 0.810 0.920 0.960 1.02 1.07 65.9 0.31 0.505 0.752 0.880 0.895 0.970 1.02 62.9 None 0.055 0.050 0.050 0.055 0.056 0.055 0
2118 Biochemistry: Mirsky and Silverman TABLE 7. Digestion by of nuclei from which argininerich histone was extracted in 765 alcohol-o.07 N HCl (the arginine-rich histone extracted was 42.4% of the amount of present) (,g/ml) 15 30 45 60 90 120 (%) 4.76 1.12 1.22 1.28 1.35 1.35 1.35 83.2 2.44 1.05 1.18 1.25 1.31 1.35 1.35 83.2 1.23 0.920 1.05 1.12 1.16 1.18 1.20 74.0 0.62 0.790 0.930 1.05 1.05 1.08 1.15 70.8 0.31 0.665 0.860 0.955 0.970 1.02 1.04 64.2 None 0.040 0.040 0.038 0.038 0.042 0.044 0 removal of lysine-rich histone did not increase the accessibility of to RNA polymerase, but removal of argininerich histone did. Comparing the results of selective histone extraction on the accessibility to of and polymerase it is clear, first, that the difference between the two types of histone is much more marked for polymerase than for in blocking access to, and, secondly, that the relative effects of the two types of histones are reversed, blocking by lysine-rich histone being more effective for and blocking by arginine-rich histone more effective for polymerase. Another difference between RNA polymerase and is in the striking difference between the amounts of in chromatin accessible to these two enzymes. digests at least 71.5% of the in chromatin of an isolated nucleus, and even more after removal of lysine-rich histone. RNA polymerase, on the other hand, has access to less than 10% of the in chromatin that still contains all its histone, and removal of lysine-rich histones makes no significant difference. Even in a living cell, only the same small part of the serves as a template for RNA synthesis. Nearly all the in chromatin is accessible to because the combination between histones, other proteins, and is loose. Consequently, in the course of time, gains access to. The looseness of combination between proteins and in chromatin is also shown by TABLE 8. Digestion of in nuclei by, in relation to histone previously extracted from nuclei Type nucleus Lysine-rich Arginine-rich histone extracted histone extracted Intact nuclei (20.6%) 19.1 29.0 36.4 45.6 0.595 1.20 0.630 0.778 0.950 1.05 The absorbance at 265 nm of nucleotide released after digestion for 30 min by 1.23 Ag/ml of. These figures are taken from Tables 2-7. The amount of histone extracted from each preparation of nuclei is expressed as a percentage of the amount of present. Proc. Nat. Acad. Sci. USA 69 (1972) some instructive experiments by Itzhaki and Cooper (23), in which a significant quantity of polylysine of molecular weight 14,000 penetrated to the in chromatin, combining with over 40% of its phosphoric acid groups. "Looseness" in the structure of chromatin is not apparent in its reactions with RNA polymerase, for only a very restricted part of the serves as a template for RNA synthesis either in vitro or in vivo. If, however, we consider as a template for synthesis when is replicated in mitosis, all of the must serve as a template, and a close approximation to this occurs even in an in vitro system. In a notable paper, Schwimmer and Bonner (16) reported experiments on the reaction of nucleohistone with the Kornberg polymerase in which "nucleohistone is nearly as effective in support of synthesis as is deproteinized," although the same nucleohistone "is totally ineffective in the support of -dependent RNA synthesis." in nucleohistone seems to be about as accessible to polymerase as is the of chromatin to. Experiments are now under way to see whether the same factors that control accessibility to in chromatin for also do so for polymerase. In this report the experiments on accessibility to have been concerned only with histones, other proteins not being mentioned. There are, of course, other proteins in chromatin and they can be distinguished sharply from histones, for the nonhistone proteins contain tryptophan, whereas histones do not (22). Nonhistone proteins, as well as histones, are combined with (17, 18) and, furthermore, the amount combined varies considerably in nuclei of different tissues. Large amounts of nonhistone protein are present in cells with an abundant and active cytoplasm (19-21), "an indication that the cytoplasm may influence the composition of chromosomes and presumably their behavior" (22). The amounts of the various histones, on the contrary, in relation to quantity of are about the same in many cell types. A question that arises here is whether the nonhistone proteins block access of to. The evidence presently available is that they do not, for the course of action of on calf liver nuclei (whether prepared in sucrose or in 0.01 N citric acid) is just the same as on thymus nuclei, although liver nuclei contain far more nonhistone protein; the figures in Table 2 could have been for liver nuclei. Itzhaki (personal communication) has made much the same point: in experiments on the accessibility of to polylysine, she found "that chromatin extracted from very diverse sources is structurally similar, despite large differences in nonhistone protein content." The apparent ineffectiveness of nonhistone protein in blocking access of to points to a difference between and RNA polymerase in their reactions with of chromatin; many recent reports show that the access of RNA polymerase to in chromatin is determined in a specific way by nonhistone proteins of chromatin. B. S. is a postdoctoral fellow of the U.S. Public Health Service (GM 35130). We thank Miss Ana Yam for her assistance. 1. Mirsky, A. E. (1971) Proc. Nat. Acad. Sci. USA 68, 2945-2948. 2. Littau, V. C., Buidick, C. J., Allfrey, V. G. & Mirsky, A. E. (1965) Proc. Nat. Acad. Sci. USA 54, 1204-1212. 3. Mirsky, A. E., Burdick, C. J., Davidson, E. H. & Littau, V. C. (1968) Proc. Nat. Acad. Sci. USA 61, 592-597.
Proc. Nat. Acad. Sci. USA 69 (1972) 4. Georgiev, G. P., Ananieva, L. N., & Kozlov, J. V. (1966) J. Mol. Biol. 22, 365-371. 5. Spelsberg, T. C. & Hnilica, L. S. (1971) Biochim. Biophys. Ada 228, 212-222. 6. Smart, J. E. & Bonner, J. (1971) J. Mol. Biol. 58, 675-684. 7. Daly, M. M. & Mirsky, A. E. (1955) J. Gen. Physiol. 38, 405-413. 8. Alifrey, V. G., Mirsky, A. E. & Osawa, S. (1957) J. Gen. Physiol. 40, 451-490. 9. Kunitz, M. (1950) J. Gen. Physiol. 33, 349-362. 10. Bonner, J., Chalkley, R., Dahmus, M., Famnbrough, D., Fujimura, F., Huang, R. C., Huberman, J., Jensen, R., Marushige, K., Ohlenbusch, H., Olivera, B., & Widholm, J. (1968) Methods Enzymol. 12B, 32-37. 11. Panyim, S. & Chalkley, R. (1969) Arch. Biochem. Biophys. 130, 337-346. 12. Johns, E. W., Phillips, D. M. P., Simson, P. & Butler, J. A. V. (1960) Biochem. J. 77, 631-636. in Chromatin 2119 13. Smart, J. E. & Bonner, J. (1971) J. Mol. Bido. 58, 661-674. 14. Brasch, K., Seligny, V. L. & Setterfield, G. (1971) Exp. Cell Res. 65, 61-72. 15. Berlowitz, L., Kitchin, R. & Pallota, D. (1972) Biochim. Biophys. Acta 262, 160-168. 16. Schwimmer, S. & Bonner, J. (1965) Biochim. Biophys. Acta 108, 67-72. 17. Mirsky, A. E. & Ris, H. (1947) J. Gen. Physiol. 31, 7-18. 18. Wang, T. Y. (1967) J. Biol. Chem. 242, 1220-1226. 19. Mirsky, A. E. & Ris, H. (1951) J. Gen. Physiol. 34, 475-492. 20. Dingman, C. W. & Sporn, M. B. (1964) J. Biol. Chem. 239, 3483-3492. 21. Spelsberg, T. C. & Hnilica, L. S. (1971) Biochim. Biophys. Acta 228, 202-2i1. 22. Mirsky, A. E. (1950-1951) Harvey Lect. Ser. 46, 98-115. 23. Itzhaki, R. F. & Cooper, H. K. (1971) Biochem. J. 125, 101 P.