Chromatin Structure and Gene Expression Citation Gottesfeld, Joel M. (1976) Chromatin Structure and Gene Expression. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/3ayy-cg55. https://resolver.caltech.edu/CaltechTHESIS:11142025-230228251 Abstract Rat-liver chromatin has been separated into nuclease-sensitive and resistant fractions after mild digestion with DNAase II. The nuclease-sensitive material is further fractionated into Mg ++ -soluble and insoluble chromatin fractions. The kinetics of production of these chromatin fractions have been investigated. After a brief enzyme treatment (5 min under standard conditions), 11% of the input chromatin DNA is found in the Mg ++ -soluble fraction. This DNA has a weight-average single strand length of about 400 nucleotides and, as determined by renaturation kinetics, comprises a subset of middle repetitive and nonrepetitive DNA sequences of the rat genome. Cross-reassociation experiments show that a fractionation of whole genomal DNA sequences has been achieved. Moreover, the Mg ++ -soluble fraction of liver chromatin is enriched in nonrepeated sequences coding for liver RNA but not for brain RNA. Fractionation does not depend on some general property of chromatin but is specific with regard to the template activity of the tissue from which the chromatin was obtained. The Mg ++ -soluble, template-active fraction is enriched five-fold in DNA sequences complementary to RNA. The Mg ++ -soluble fraction is enriched in nonhistone chromosomal proteins and depleted in histone protein. Histone I (fl) is absent from the Mg ++ -soluble active fraction. About half of the DNA of both Mg ++ -soluble and Mg ++ -insoluble fractions is resistant to prolonged digestion with DNAase II or staphylococcal nuclease. The nuclease- resistant structures of inactive (Mg ++ -insoluble) chromatin are DNAhistone complexes which sediment at 11-13S. Two nuclease-resistant species are present in active chromatin. These particles sediment at 15 and 20S, respectively, and contain DNA, RNA, histone and nonhistone proteins. Thermal denaturation studies suggest that the· DNA of active chromatin is complexed primarily with nonhistone proteins. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Biochemistry) Degree Grantor: California Institute of Technology Division: Biology Major Option: Biochemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Bonner, James Frederick Thesis Committee: Bonner, James Frederick (chair) Davidson, Norman R. Owen, Ray David Vinograd, Jerome Britten, Roy Defense Date: 23 June 1975 Record Number: CaltechTHESIS:11142025-230228251 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:11142025-230228251 DOI: 10.7907/3ayy-cg55 ORCID: Author ORCID Gottesfeld, Joel M. 0000-0002-4643-5777 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17762 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 19 Nov 2025 18:02 Last Modified: 19 Nov 2025 18:28 Thesis Files PDF - Final Version See Usage Policy. 38MB Repository Staff Only: item control page

CHROMATIN STRUCTURE AND GENE EXPRESSION thesis by Joel M. Gottesfeld In Partial Fulfillment of the Requirements For the Degre e of Doctor of Philosophy California Institute of Tec hnology Pasadena , California 1976 (Submitted June 23, 1975) -ii - ACKNOWLEDGEMENTS It has been said that"the work of the righteous is often done by others" (The Talmud, quoted by Koshland, 1971, unpublished). With no claim to righteousneness, this author would like to thank the others who often did the work reported herein. They are Ralph Wilson, Georgi Bagi, Stan Nakamoto and Beckie Berg . My sincere thanks and gratitude goes to my former teachers at Berkeley, especially Professors M. Calvin, R. D. Cole and D. E. Koshland, who sparked my interest in biochemistry. My thanks also goes to the members of my thesis committee, Drs. Roy Britten, Norman Davidson, Ray Owen and Jerry Vinograd, for the advice and counsel they have given me. Most of all, I want to express my grat itude to Professor James Bonner, for the opportunity to work in his laboratory and for his patience and constant encouragement. - iii- ABSTRACT Rat-liver chromatin has been separated into nuclease-sensitive and resistant fractions after mild digestion with DNAase II. nuclease-sensitive material is further fractionated into Mg and insoluble chromatin fractions. ++ The -soluble The kinetics of production of these chromatin fractions have been investigated. After a brief enzyme treatment (5 min under standard conditions), 11% of the input chromatin DNA is found in the Mg ++ -soluble fraction. This DNA has a weight-average single strand length of about 400 nucleotides and, as determined by renaturation kinetics, comprises a subset of middle repetitive and nonrepetitive DNA sequences of the rat genome. Cross- reassociation experiments show that a fractionation of whole genomal DNA sequences has been achieved. Moreover, the Mg ++ -soluble fraction of liver chromatin is enriched in nonrepeated sequences coding for liver RNA but not for brain RNA. Fractionation does not depend on some general property of chromatin but is specific with regard to the template activity of the tissue from which the chromatin was obtained. The Mg ++ -soluble, template-active fraction is enriched five-fold in DNA sequences complementary to RNA. The Mg ++ -soluble fraction is enriched in nonhistone chromosomal proteins and depleted in histone protein. Histone I (fl) is absent -iv- from the Mg Mg ++ ++ -soluble active fraction. -soluble and Mg ++ . . About half of the DNA of both . . -insoluble fractions is resistant to prolonged digestion with DNAase II or staphylococcal nuclease. The nuclease- . . . ( Mg ++ -insoluble . ) chromatin are DNAresistant structures of inactive histone complexes which sediment at 11-13S. species are present in active chromatin. Two nuclease-resistant These particles sediment at 15 and 20S, respectively, and contain DNA, RNA, histone and nonhistone proteins. Thermal denaturation studies suggest that the· DNA of active chromatin is complexed primarily with nonhistone proteins. -v- TABLE OF CONTENTS CHAPTER I II III TITLE PAGE ACKNOWLEDGEMENTS ii ABSTRACT iii INTRODUCTION 1 References 15 ISOLATION OF TEMPLATE ACTIVE AND INACTIVE REGIONS OF CHROMATIN 20 Publication 21 PARTIAL PURIFICATION OF THE TEMPLATE-ACTIVE FRACTION OF CHROMATIN: A PRELIMINARY REPORT Publication IV 33 STRUCTURE OF TRANSCRIPTIONALLY-ACTIVE CHROMATIN VI 28 SEQUENCE COMPOSITION OF THE TEMPLATE-ACTIVE FRACTION OF RAT-LIVER CHROMATIN V 27 100 BIOPHYSICAL STUDIES ON THE MECHANISM OF QUINACRINE STAINING OF CHROMOSOMES 130 Publication 131 Further Discussion 140 Reference 142 - 1- I. INTRODUCTI ON The phenomena of differentiation and cellular specialization in eukaryotes can be traced, in part, to differential gene activity. statement rests on three very important observations: This first, the various cell types of a given organism possess the same genetic complement (Gurdon, 1962); second, differentiated cells of metazoan creatures transcribe only a limitted fraction of their full genetic potential (Brown and Church, 1972; Grousse et al., 1972); and, third, the RNA populations of cells are, to a certain degree, tissue-specific (Grousse et al., 1972; Ryffle and McCarthy, 1975). When compared to the transcription of protein-free DNA, the in vitro transcription of isolated chromatin is highly restricted (Huang and Bonner, 1962). Even though this obser- vation was made over a decade ago, the specific chromosomal components responsible for the control of transcription have yet to be identified. Both RNA (Huang and Bonner, 1965; Britten and Davidson, 1969) and nonhistone proteins (Paul and Gilmour, 1968; Davidson and Britten, 1971, 1973; Barrett et al., 1974) have been suggested as potential modulators of gene activity. It appears that the regulatory components, whatever they might be, remain intact in isolated chromatin; in vitro transcription of both the- pea seed globulin gene and the globin genes has been f ound to be hi ghly tissue-specific (Bonner et al., 1963; Gilmour and Paul, 1973; Axel et al., 1973). One approach to the identification of the transcriptional control elements is to separate interphase chromatin into re gions which are trans criptionally active and into regions which are transcriptionally -3-. inert. We can then compare the constituents of these chromatin species. This is the approach ta.ken herein: a method is described for the isola- tion of the "template-active" fraction of chromatin and the properties of this fraction have been investigated in detail. This work was begun with the notion that the structure of transcriptionally active chromatin is markedly different from that of transcriptionally inactive regions of chromatin. There is a vast cytological and genetic literature which supports this contention. First of all, RNA synthesis appears to be restricted to the extended euchromatic regions of thymocyte nuclei; no transcription can be detected by E.M. autoradiography in the condensed heterochromatic regions of the nucleus (Littau et al., 1964). Second, during the pachytene stage of amphibian oogenesis, transcription takes place in the extended loops of the lampbrush chromosomes (Miller, 1965). Similarly, prior to transcription in polytene cells of Drosophila, the specific chromomeres to be transcribed undergo a marked structural alteration; the bands extend and become puffs (Berendes, 1973). This phenomenon of puffing is particularly well documented for the large Balbiani rings of Chironomus (Daneholt ahd Hosick, 1973; Grossbach, 1973). Griffith (1970) and many others have found that the basic fiber 0 0 diameter of interphase chromatin is 100 A. Some 250 A macrofibers are also seen; however, these macrofibers probably arise from the pairing 0 or aggregation of two 100 A fibers (Ris, 1969; Pooley et al., 1974 ). -4- Olins and 0lins (1974) and Woodcock (1973) have observed regular spacings of chromatin particles (termed v-bodies)in water-swollen nuclei centrio fuged onto electron microscope grids. These particles are 60-80 A in 0 diameter and are joined by thin filaments of DNA 15 A in diameter. At or near physiological ionic strength, the "particles-on-a-string" 0 structure condenses into the 100 A chromatin fiber (Griffith, 1975; Germond et al., 1975). The chromatin particles (the v-bodies) probably reflect the ordered packaging of DNA in the chromatin fiber (Griffith, 1975). Since the studies cited above were carried out with whole chromatin, unfractionated with respect to transcriptional activity, and since most of the chromatin in the cell s of higher eukaryotes is transcriptionally 0 inactive, the 100 A-fiber represents the structure of inactive regions . Transcriptionally active chromatin has been visualized in the electron microscope in many organisms and by many workers . Griffith (1970) has investigated the chromatin of the protozoan Tetrahymena. The genome of this organism is similar to that of typical higher eukaryotes; both repetitious and unique sequences are found and the DNA is complexed with both histone and nonhistone proteins. This organism is dissimilar from higher eukaryotes in that 40% of the genome appears to be transcriptionally active. Electron microscope studies of Tetrahymena nucle i 0 O lysed directly on grids show fibers 30 A in width. fibers were observed (Griffith, 1970). 0 No 100 A or 250 A- Similarly, Miller (1965) has -50 observed a fiber diameter of about 30 A in the active loops of lampbrush chromosomes. These loops are connected by inactive chromatin 0 fibers 200 A in diameter. Harnkalo (personal communication) has ob- served a structural transition in the chromatin of a Trypanosome. When this organism is transcriptionally dormant, the chromatin is seen in the "particles-on-a-str ing" or v-body configuration. Upon induction of RNA synthesis, the chromatin fibers appear more extended with a larger and more random spacing between particles. Other electron microscope studies of chromatin have identified several classes of chromatin fibers: Oudet et al. (1975) have fraction- ated chromatin into dense and light fractions in glycerol gradients. The dense fraction exhibits a regular close packing of particles (termed nucleosomes by these authors). In the li ght fraction, regions of clo s e~ packed particles are separated by long stretches (0.5 to>lOµm) of thin 0 fibers 15-25 A in diameter. This light fraction could correspond to ac tive chromatin, but this has not been proven. Langmore and Wooley (1975) have obtained qualitatively similar results with the high resolution scanning transmission electron microscope. Larger particle diameters, however, were observed by these authors when unstained and unfixed chromatin samples were examined. Griffith (1975) has demonstrated that SV 40 DNA complexed with host histones is compacted l/4th to l/7th its native contour length. SV 40 "chromatin" exhibits v-body or nucleosome structures (Griffith, 1975; Germond et al., 1975). It is interesting to note that ribosomal genes in the act of transcription -6- are the length of their transcription product (pre-rRNA) (Miller and Hamkalo , 1972). 0 Thus it is likely that the fully extended 15-30A chromatin fiber corresponds to genetically active chromatin while the 0 100 A or particulate fiber corresponds t o inactive regions of chr omatin . There is further genetic and cytological data which demonstrates that condensed chromatin is transcriptionally inert . Perhaps the most striking example of this sort is X-inactivation (Lyon, 1961) . Here the condensation of one X-chromosome leads to the total silencing of its genetic potential. Similarly, translocations of euchromatic seg- ments to chromosomal regions proximal to condensed heterochromatin can lead to the inactivation of the euchromatic segment. Position effect phenomena of this sort are documented in both the mouse and in Drosophila (Baker , 1968). The morphological differences between active and inactive regions of chromatin as revealed by both the electron microscope and the light microscope must have a biochemical basis. Separation of chromatin into active and inactive regions should allow us to determine the chromosomal elements responsible for both structure and genetic activity. The first attempt at chromatin fractionation was reported by Frenster, Allfrey and Mirsky (1963). Chromatin from calf thymus lymphocytes was sonicated and separated into euchromatin and heterochromatin by differential centrifugation. Based on the criteria of cofractionation of nascent RNA chains and microscopic appearance, the fractionation appeared successful. About 80% of the nuclear -7- DNA was found in the heterochromatin fraction, but this fraction contained only 14% of the newly synthesized RNA. The heterochromatin 0 appeared as dense masses of 100 A fibers; the euchromatin fraction was 0 composed, however, of 50 A fibrils. In a subsequent report (Frenster, 1963), the euchromatin fraction was found to have a decreased histone to DNA ratio and an increased content of nonhistone chromosomal proteins. The approach of differential centrifugation of sonicated or pressuresheared chromatin in sucrose gradients has been taken by Chalkley and Jensen (1968), Yun.is and Yasmineh (1971), Duerkson and McCarthy (1971), Nishiura (1972), Murphy et al.(1973), Warnecke et al. (1973), McCarthy et al. (1973), Chesterton et al. (1974), and most recently by Doenecke and McCarthy (1975a,b). The major findings of these studies may be summarized as follows: the light chromatin fractions (euchromatin) exhibit a high template capacity for the support of RNA synthesis with exogenous polymerase. This result has now been obtained with both prokaryotic and eukaryotic RNA polymerases, Newly synthesized RNA molecules cosediment with the light chromatin fraction; further, different protein populations are found in the light and dense chromatin fractions. In general, light chromatin is enriched in nonhistone protein and depleted in histone protein. The heterochromatin fractions are enriched in non-transcribed satellite DNA. Doenecke and McCarthy (1975a) have concluded that the basis for chromatin fractionation is differential aggregation. Although the light chromatin fraction has many of the properties expected for transcriptionally-active chromatin (high tern- -8- plate activity, association of nascent RNA), the crucial hybridisation experiments have yet to be reported. In vivo template activity can only be established by demonstrating that the light fraction is enriched in DNA sequences which code for cellular RNA. Since the fraction of the genome which is transcribed in any given cell type is limited, the DNA of the active fraction must comprise a small subset of the genome. Furthermore, the size of this subset should correspond to the fraction of the genome which is transcriptionally active in the particular cell type under study. The validity of the sucrose gradient fractionation method can only be established once such hybridization experiments have been carried out. Mcconaughy and McCarthy (1972) have fractionated pressure-sheared chromatin by thermal chromatography on hydroxyapatite. The chromatin DNA which elutes at the lowest temperatures was found to be enriched in sequences complementary to RNA from homologous but not heterologous tissues. This procedure has the major drawback that only the active DNA can be recovered from hydroxyapatite; native protein-DNA interactions are destroyed by thermal denaturation. Janowski et al. (1972) have separated pressure-sheared chromatin by gel exclusion chromatography. The exluded fractions (extended chromatin) preferentially bind 3H-labeled E. coli RNA polymerase; it was also reported that the excluded fractions carry nascent RNA chains. Another chromatographic fractionation procedure which appears highly promising is anion-exchange chromatography on ECTHAM-cellulose using -9- a gradient of increasing pH (Ree ck et al., 1972; Simpson and Reeck, 1973). Although no differences were found in the overall protein to DNA ratios of the fractions obtained from ECTHAM-cellulose columns, the late-eluting (high pH) chromatin is enriched in nonhistone protein and depleted in histone protein. in the late-eluting fractions. Histone I (fl) is greatly diminished In agreement with the findings of Mcconaughy and McCarthy (197 2 ), the late-eluting, putatively active chromatin melt s at lower temperatures than the early- eluting material. The DNA of the late- eluting fraction is more available for binding RNA polymerase (Simpson, 1974) and more susceptible to nucleolytic attack (Simpson and Polacow, 1973) than the DNA of the early- eluting chromatin. Circular dichroism spectroscopy suggests that the late- eluting chromatin is in an e xtended DNA-like conformation (Polacow and Simpson, 1973). The ECTHAM-c ellulose fractionation method appears to produce a chromatin fraction with all the properties one would expect of transcriptionally active chromatin; however, the recent report of Howk et al. (1975) demonstrates that neither the late- eluting fraction from ECTHAM-cellulose nor the light fraction from glycerol gradients are enriched in specific transcribed sequences. No differences in the distribution of transcribed vj_ral sequences were found in the chromatin fractions obtained from either infected or noninfected cells. Further hybridiza- tion experiments are needed to determine whether the ECTHAM-cellulose method actually does separate active from inactive chromatin. -1 0- Turner and Hancock (1974) have separated sheared mouse chromatin by partition in a two-phase aqueous polymer system; the fractions obtained by this method were found to differ in their content of nonhistone proteins. Sheared chromatin l1as also be en fractionated by equili- brium centdf'ugation in gradients of CsSC\ (Wilt and Ekenberg, 1971), chloral hydrate (Hossainy et al., J973), renografin (Chan and Scheffler, 1975), and metrizamide (Rickwood et al., 1974; Monahan and Hall, 1974). Although sheared chromatin can be resolved into spec ies which differ in density, protein composition and template activity, Rickwood et al. (1974) have concluded that the metrizamide chromatin fractions do not correspond to active and inactive regions . No significant differ ences were found in the concentration of globin genes in the chromatin fractions obtained from either erythropoietic or nonerythropo :i. etic ceJ.ls . These r esults show that the criteria of chemical composition and in vitro template activity are insufficient to demonstrate that a particular chromatin fraction is truely transcriptionally active in vivo. This laboratory has adopted an altogether different approach to the problem of chromatin fractionation. Previous attempts at fraction- ation have b egun with sonicated or pressure-sheared chromatin . It is highly likely that these methods of shearing result in the rearrangement of' chromosomal proteins . Indeed, Noll et al. ( 1975) have found significant <lifferences in the subunit structure of native and mechani- -11- cally-sheared chromatin. 'ro allow for the subsequent fractionation into active and inactive regions, chromatin DNA must be sheared to a size smaller Umn the average unit of transcript ion. Marushige and Bonner (1971) int roduced limitted digestion of chromatin with the endonuclease DNAase II for this purpose. In our hands, nuclease digestion does not J ead to detectable levels of protein rearrangement ( Got tesfeld et al. , unpublished). In the nucle us, nontranscri bed het eroc hromatin is hi ghly aggregated while transcriptionally-active euchromatin is extended (Littau et al., 1964). It is well known that isolated chromatin is insoluble at or near physiological ionic strength. Therefore, Marushige ru1d Bonner (1971) postulated that inactive regions of chromatin s110uld be precipitated by salt wbile active chromatin, which is extended in vjvo, should remain soluble. It has been found that the bulk of chromatin DNA is precipitated with either saline+ citrate ([Na ]= 0.2 M) or MgC1 2 (2 mM) while a minor fraction of DNA remains soluble (Marushige and Bonner, 1971; Billing and Bonner, 1972). Moreover, the amount of DNA in the soluble fraction varies depending upon the source of the chromatin, but corresponds quite closely to the measured template activity of the particular chromatin under study (Billing and Bonner, 1972). This thesis extends the work of Marushige and Bonner ( 1971) and Billing and Bonner ( 1972). The Mg ++ -soluble fraction of rat-liver chromatin is enriched in sequences coding for the RNA of homologous but not heterologous tissue. Thus the Mg ++ -sol- uble fraction corresponds to transcriptionally-active chromatin in -12- vivo. This fraction also comprises a subset of nonrepeated DNA sequences and a subset of families of repeated sequences. The Mg ++ soluble active fraction contains a high proportion of nonhistone chromosomal proteins and a lower histone to DNA ratio than either unfractionated chromatin or the Mg ++ -insoluble fraction. In agree- ment with the findings of Reeck et al. (1972) and Simpson and Reeck (1973), histone I (fl) is absent from the active fraction. It appears that all the DNA in chromatin is organized into regions which are sensitive to prolonged nuclease treatment and into regions which are nuclease-resistant. The nuclease-resistant structures of inactive chromatin are DNA-histone complexes; these structures resemble the v-bodies observed by Olins and Olins ( 1974} ( Go,ttesfeld et al., 1975). The nuclease-resistant structures of active chroma- tin are more heterogeneous than those of inactive chromatin; they are complexes of DNA, RNA, histone and nonhistone protein. Thermal denaturation studies suggest that the DNA of active chromatin is complexed primarily with the nonhistone proteins. In agreement with the results of Mcconaughy and McCarthy (1972), active chromatin melts at lower temperatures than inactive chromatin. Circular di- chroism s tudies suggest that active chromatin is in a conformation more like that of protein-free DNA than that of inactive chromatin (Polacow and Simpson, 1973; Gottesfeld et al ., herein). -13- The discovery of single copy and repeated sequences in eukaryotic DNA (Britten and Kohne, 1968) suggested that repetitious sequences might be control elements in the modulation of transcription (Britten and Davidson, 1969; Davidson and Britten, 1973). Supporting evidence for this model comes from the finding of general interspersion of moderately repetitive and single copy sequences throughout the genome (Davidson and Britten, 1973; Davidson et al., 1973; Graham et al., 1974; Davidson et al., 1975a) and the finding that polysomal mRNA is transcribed mainly from the nonrepeated fraction of the genome (Davidson and Britten, 1973; Lewin, 1975). Furthermore, sequences which are transcribed into mRNA are found adjacent to repetitive sequences in the DNA of the sea urchin (Davidson et al., 1975b). It is tempting to speculate that the binding of activator molecules to middle re:retitive sequences promotes the transcription of neighboring single copy structural gene sequences. In its simplest form, the Britten- Davidson model predicts that if a given repetitive sequence is adjacent to a transcribed single copy sequence, then all the members of this repetitive family are also adjacent to transcribed sequences. Thus we would predict that only a limited fraction of the repetitive complexity of the genome should be found adjacent to transcribed single copy sequences. with this view. The results presented herein are consistant The active fraction of rat-liver chromatin (which is enriched five-fold in transcribed single copy sequences) contains about 11% of the total single copy complexity and 5-8% of the repeti- -14- tive complexity of the rat genome. This demonstrates that the distri- bution of repetitive sequences in rat DNA is decidedly nonrandom. What is more, preliminary results suggest that repetitive and single copy sequences are interspersed in the active fraction. These results offer insights into the functional organization of the eukaryotic chromosome; hopefully in the near future we will have a clear picture of the workings of the eukaryotic gene regulatory apparatus. REFERENCES _, Axel, R., Cedar, H., and Felsenfeld, G. (1973). Proc. Nat. Acad. Sci. USA 70, 2029-2032. Baker, W. K. _(1968) Adv. Genet. 14, 133-169. Barrett, T., Maryanka, D., Ha.mlyn, P. H., and Gould, H. J. (1974) Proc. Nat. Acad. Sci. USA 71, 5057-5061. Berendes, H. D. (1973) Int. Rev. Cytology 12.., 61-116. Billing, R. J., and Bonner, J. (1972) Biochim. Biophys. Acta 281, 453462. Bonner, J., Huang, R. C., and Gilden, R. V. (1963) Proc. Nat. Acad. Sci. USA .2.Q., 893-900. Britten, R. J.' and Davidson, E. H. (1969) Science 165, 349-357, Britten, R. . J.' and Davidson, E. H. (1971) Quart. Rev. Biol. 46, 111-138. Britten, R. J.' and Kohne, D. E. (1968) Science 161, 529-540. Brown, I. R., and Church, R. B. (1972) Develop. Biol. 29, 73-84Chalkley, R., and Jensen, R. (1968) Biochemistry 1, 4380-4388~ Chan, R. T., and Scheffler, I. E. (1975) Biochemistry, submitted. Chesterton, C. J., Coupar, B. E. H., and Butterworth, P. H. W. (1974) Biochem. J. 143, 73-81. Daneholt, B., and Hosick, H. (1973) Cold Spring Harbor Symp. Quant. Biol. 38 , 629-635. Davidson, E. H., Angerer, R., Ga±au, G. A., and Britten, R. J. (1975a) Chr omosoma, in press. -16- Davidson, E. H., and Britten, R. J. (1973) Quart. Rev. Biol. 48, 565613. Davidson, E. H., Hough, B. R., Amenson, C. S., and Britten, R. J. (1973) J. Mol. Biol. 77, 1-23. Davidson, E. H., Hough, B. R., Klein, W. H., and Britten, R. J. (1975b) Celli, 217- 238. Doenecke, D.' and McCarthy, B. J. (1975a) Biochemistry 14, 1366-1372. Doenecke, D.' and McCarthy, B. J. (1975b) Biochemistry 14, 1373-1377. Duerkson, J . G.' and McCarthy, B. J. (1971 ) Biochemistry 10, 1471- 1478. Frenster, J. H. (1963) J . Cell Biol. 19, 25a. Frenster, J. H.' Allfrey, V. G.' and Mirsky, A. E. (1963) Proc. Nat. Acad . Sc i. USA 22_, 1026-1032. Germond, J . E., Hirt, B., Oudet, P., Gross-Bellard, M., and Chambon, P. (1975) Proc . Nat. Acad. Sci. USA, in press . Gilmour, R. S., and Paul, J . (1973) Proc . Nat . Acad. Sci . USA 70, 34403442 . Gottesfeld, J.M., Kent, D., Ross, M. , and Bonner, J. (1975) In "Florida Colloquium on Molecular Biology" ed. Stein, G. (Academic Press, New York) in press. Griffith, J. D. (1970) Ph . D. thesis, Cal ifornia Institute of Technology, Pasadena . Griffith, J. D. (1975) Science 187 , 1202- 1203 . Graham, D. E. , Neu feld, B. R. , Davidson, E. H. , and Britten, R. J. (1974) Cell _!, 127- 137. -17- Grossbach, U. (1973) Cold Spring Harbor Symp. Quant. Biol. 38, 618-627. Grousse, L., Chilton, M. D., and McCarthy, B. J. (1972) Biochemistry 11, 798-805. Gurdon, J. (1962) Develop. Biol . .!±_, 256-273. Hossainy, E., Zweidler, A., and Bloch, D. P. (1973) J. Mol. Biol. 74, 283-289. Howk, R. S., Anisowicz, A., Silverman, Y., Parks, W. P., and Scolnick, E. M. (1975) Cell.!±_, 321-328. Huang, R. C., and Bonner, J. (1962) Proc. Nat. Acad. Sci. USA 48, 12161222. Huang, R. C., and Bonner, J. (1965) Proc. Nat. Acad. Sci. USA _2.!±_, 960967. Janowski, M., Nasser. D. S., and McCarthy, B. J. (1972) In "Karolinska Symposia on Research Methods in Reproductive Endocrinology," ed. Diczqalusy, E. (Karolinska Institutet, Stockholm) 5th Symposium, pp. 112-129. Langmore, J. ,and Wooley, J. C. (1975) Proc. Nat. Acad. Sci. USA, in press. Lewin, B. (1975) Cell.!±._, 77-93, Littau, V. C., Allfrey, V. G., Frenster, J. H. , and Mirsky, A. E. (1964) Proc. Nat. Acad. Sci. USA~' 93-100. Lyon, M. F. (1961) Nature 190, 372-373, Marushige, K., and Bonner, J. (1971) Proc. Nat. Acad. Sci. USA 68, 29412944. -18- Mcconaughy, B., and McCarthy, B. J. (1972) Biochemistry 11, 998-1003. McCarthy, B. J., Nishiura, J. T., Doenecke, D., Nasser, D. S., and Johnson, C. B. (1973) Cold Spring Harbor Symp. Quant. Biol. 38, 763771. Miller, 0. L. (1965) Nat. Cancer Inst. Monographs 18, 79-89. Miller, 0. L., and Hamkalo, B. A. (1972) In "Molecular Genetics and Developmental Biology," ed. Sussman, M. (Prentis-Hall, New Jersey) pp. 183-199. Moaahan, J. J., and Hall, R. R. (1974) Nucleic Acid Res.!, 1359-1370. Murphy, E. C., Jr., Hall, S. H., Shepherd, J. H., and Weiser, R. S. (1973) Biochemistry 12, 3843-3853. Nishiura, J. T. (1972) Ph. D. thesis, University of Washington, Seattle. Noll., M., Thomas, J. 0., and Kornberg, R. (1975) Science 187, 1203-1206. Olins, D., and 0lins, A. (1974) Science 183, 330-333. Oudet, P., Gross-Bellard, M., and Chambon, P. (1975) Cell.!±_, 281-299. Paul, J., and Gilmour, R. S. (1968) J. Mol. Biol. 34, 305-316. Polacow, I., and Simpson, R. T. (1973) Biochem. Biophys. Res. Commun. ~. 202-207. Pooley, A. S., Pardon, J., and Richards, B. (1974) J. Mol. Biol. §2., 533-549. Reeck, G. R., Simpson, R. T., and Sober, H. A. (1972) Proc. Nat. Acad. Sci. USA 69, 2317-2321. Rickwood, D., Hell, A., Malcolm, S., Birnie, G.D., MacGillivray, A. J., and Paul, J. (1974) Biochim. Biophys. Acta 353, 353-361. -19- Ris, H. (1969) In "Handbook of Molecular Cytology," ed. Lima-de-Faria, A.(North Holland, Amsterdam) pp. 221-250. Ryffle, G. U., and McCarthy, B. J. (1975) Biochemistry 14, 1379-1384. Simpson, R. T. (1974) Proc. Nat. Acad. Sci. USA 71, 2740-27113. Simpson, R. T., and Polacow, I. (1973) Biochem. Biophys. Res. Commun. n.., 1078-1084. Simpson, R. T., and Reeck, G. R. (1973) Biochemistry 12, 3853-3859. Turner, G., and Hancock, R. (1974) Biochem. Biophys. Res. Commun . ..2§., 437-445. Yunis, J. J., and Yasmineh, W. G. (1971) Science 174, 1200-1209. Warnecke, P., Kruse, K., and Harbers, E. (1973) Biochim. Biophys. Acta 331, 295-304. Wilt, F. H., and Ekenberg, E. (1971) Biochem. Biophys. Res. Commun. 44, 831-836. Woodcock, C. L. F. (1973) J. Cell Biol. 22_, 368a. -20- II. ISOLATION OF TEJvlPLATE ACTIVE AND INACTIVE REGIONS OF CHROMATIN The material in this chapter is reprinted by permission of Academic Press from Methods in Enzymology, Volume 40, Part E, pp. 97102 (1975). -21- [8] Isolation of Template Active and Inactive Regions of Chromatin By JA'.\IES BoNNEH, JoEL GoTTESFELD, " ' 1LLIA'.\1 GARRARD, RONALD BILLING, and LYNDA rPHOL'SE In the differentiated eukaryotic cell, only a portion of the nuclear D:NA is transcribed into RNA while the majority of the genetic material is repres:::ed. Part of this transcriptional heterogeneity appears to be determined by the protein complement of chromo:::omes. To better understand the role of chromosomal proteins in chromatin function and structure, we and other innstigators have developed techniques to fractionate chromatin into template active and inactive regions. Such an approach encountns two main t.echnical problems. First, chromatin must be sheared to a size small enough to allow the subsequent separation of actin and inactin regions. Second, a means of selectivity fractionating the material is required. To minimize cross-contamination of active and inactive regions , chromatin should be sheared to less than the size of an awrage unit of transcription. :\lost inves tigators have used mechanical methods (including sonication 1 and French pressure cell passage 2 • 3 ) to shear chromatin . :;.\Iarushige and B01111er 4 have introduced the use of D~ ase II for this purpose. Xucleolytic c:lenrnge of chrcrnatin DNA is gentle and does not lead to detectable )enb of protein rea rrangement during t'hearing or fractionation . We h::n-e adopted this t ec hnique and desc ribe appropriate conditions on p . 98. Sheared chromatin has been separated into active and inactive regions 1 J . Frt> nsl er, Y. G. Allfrt>y, and A. E. Mir~ky, Proc. Nat . Acad. Sci. U.S . 50, 1026 (1963). 'B. L. :\lcConaughy and B. J. McCarthy, Biochemis/Jy 11,998 (1972). '?\I. Janowski , D. S. Nasser, and B. J. McCarthy, in "Karolinska Symposia on Research :\I et hods in Reproductive Endocrinology" (E. Diczfalusy, ed.), 5th Symposium . p. 112. Karolinska Institutet, Stockholm, 1972. • K. l\farushige and J. Bonner, Proc . Nat. Acarl. Sci. CS. 68, 2941 (1971). -22- 98 THE CELL XUCLEUS A~D CHRO:\IATIN PROTE!:-S:'; [8) by differential ccntrifugation, 1 •5 and by chromatography on hyclroxyapatite/ agaro:::e,J and anion-exchange resin. 6 l\Iarushige and Bonner' have den-lop ed a simple and rapid fractionation technique which inrnln's selectiw precipitation of the inactive region with standard saline-citrate. Thi:,: approach i;; based on the fact that chromatin is highly aggregated at physiological ionic strength. Under such conditions, the limited portions of chromatin a,·ailable to RNA polymerase would he predicted to be lE':oS aggregated and perhaps soluble in sheared chromatin preparations. Our method, based on the same princip'le, uses l\1gCl, to prec ipitate the inactiYe region. The minor portion of chromatin ,-o luble in 2.0 rnJf ::\IgCl, is the template actin region. Procedure for Fractionation Sucro::-e-purified chromatin· is homogenized at 4° in 10 volumes of 10 mJI Tris-Cl (pH 8) using a glass-Teflon homogeniz er (10 strokes by hand followed by 2 minutes of gentle motor homogenization). Unless otherwise stated, all subsequent operations arc performed at 4°. The result ing chromatin ~olut1on is dialyzed overnight again,-t 200 Yolumes of 25 mJI sodium acetate (pH 6.6) and pelleted by centrifugation at 25,000 g for 20 minut es. The resulting material is homogenized as above in 25 m.U sodium acetate lpH 6.6) to yield a chromatin rnlution hm·ing an A 2 r.o 1 cm of 10 (measured in 1 N ~aOH) . The solution is brought to 24° and D:-;-ase II l Worthington, HDAC) is added to a final con ce ntration of 100 units/ml 00 unit$ of enzyme per A 260 unit of chromatin) . After incubation at 2-P, the pH of the solution is adjusted to 7.5 by the addition of 0.1 Jf Tris-Cl (pH 11) nncl the mixture is cooled to 4'.) . Unsheared chromatin is remowd by centrifugation at 25,000 g for 20 minutes. The amount of chromatin DX.\ remaining in the supernatant is naturally dependent upon the duration of incubation with D1\:ase. After 5 minutes of enzyme treatment, 15~1c of the chromatin DNA is found in the supernatant, whilP after prolonged incubation (15 minutes or morel, 70 to 85% is found . The sheared chromatin is no\V fractionated into template actiYe and inactiYe portions by :\IgCL preci pitation. One ninety-ninth volume of 0.20 Jf :\IgCl~ is added dropwise with rapid stirring. After 30 minutes of further stirring, the turbid suspension is centrifuged at 25,000 g for • G. R. Chalkley and R. J ensen, Biochemi.~lry 7, 4380 (1968). • G. R. Reeck, R. T. Simpson, and H. A. Sober, Proc . Nat . .Acnd. Sci. U.S. 69, 231 i ( Hli2). 'J . Bonner, G. R. Chalkley. ~I. Dahmus, D . Fambrough, G. Fujimura, R.-C. Huang, J. Hubem1::m, R. Jen-;;,n, E:. ~fa rushige, H. Ohlenbusch, B. Olin' ra, and J. Widholm , thi..s ser:,-s, \·01 1213, ;•. 3. -23- [8] FHACTIO;"l;ATIO:s; OF DNASE-SHEAHEO CHR0\1.-\.TIN 100 ~~- c 9J 0 g ., 99 8') C 0. :, ~ 70 -~ <X 60 Q'. so z c., u 40 ~ 0 C .,I 30 ~ 20 ;t- 10 0 10 20 3,) 40 so &O 70 80 90 100 % of chromatin ONA in supernatant Fw. I. Localiz!ltion of 'H-lal,eled nusl'rnt H:\'A. She,1red chromatin was µreµ!lred from a:;cites cell:;, µul:,P labeled for 5 minutes with ['H]urid ine as dPscrihed by R. J. Billing and J . Bonner [Biochim. Bio phys . .-!ctn 28, 453 (1972)]. Aliquots of chromfltin were precipitated with rnriom conrc-nt.r,llions of l\IgCL. Soluble- D:\' .-\ nn<l rndioactiYity were detrrmined. Cirdrs ,don:,; eun·r l'f'J>l'('~ent suc-crssin,Jy decrP!\sing l\IgCL concentrations u,e, l for prrcipitation of tf'mplatr in,1di1·e portion~. 20 minut(•s. The resulting supernatant contains tlw tC'mplate active chromatin fraction, while' the pC'ilct contains the template inactive material. The l engt h of the 1).:-;A in th e template actin fraction is dependent upon the conditions of nuclease treatment. A 5 minute exposure produ ces fragments of 700--850 base pairs (500 nurleotides single stranded) .8 Longer time:; of inculmtion produce much shorter fragments.~ Evidence for Fractionation Evidence for the separation of the kmplatc actiYe and inaeti,·c regions by l\IgCI~ precipitation is based on the cofractionation of nascent RXA with the acti,·e region. 9 • 1° Chromatin ,ms pn'pared from l\'oYikoff ascites ce!ls 11 which were pulse labeled for 5 minutes with [:,H Juridine. Under such conditions, the majority of the label would be predicted to be a:Ssocintecl with the template nctin region as growing RNA chains. The chromatin was then sheared with DXa:3e II, and aliquots \Ycrc precipitated with decreasing concentrations of :.\lgCI,. Figure 1 shows the • The sizr of the sheared DX.-\ was estimated by neutrnl and fllkaline sedimentation rn·. F. Studier, J. Mal. Biul. 11, 3i3 (196-!)J. • R. J . Billing !Incl J. Bonner, Biochim . Bioph ys. A ctn 28, ·153 (19i2). "J. Bonnrr, W. T. Gflrrard, J. Gottt,;:fpJd, D.S. Holmes, J , S. ScY!lll, and M. Wilkes , Cold Spring If arbor Sy11171. Quant. Biol. 38,303 (1973). 11 M. E. Dahmus and D. J. 1IcConnell, Biochemistry 8, 1524 (1969). \ \ I -24100 THE CELL NUCLEUS AND CHHOM A TIN l'HOTEINS [8) CHJ·:MJCAL COMPOSITION OF THf: TEMl'LATI•: ACTIVE AND lNACTIVJ•: HEOIONS OF CHHOM .,TIN" Novikoff 1.~cit.esb Total DNA Acidex tract.ab! e protein: DNA Nonhistone: DNA Acidextract.able Non% Total protein: hist.one: DNA DNA DNA 80 ± 6 1. 14 ± 0.02 0 ,1;) ± 0.20 78 ± 3 1 . 12 0.36 12 ± 1 0.43 ± 0.19 0 8.'i ± 0.20 24 ± 2 0.60 0.46 . 24 ± 0.04 0 !iO ± 0. lS ii3 ± 4 1.43 0.29 % Sample Sheared chromatin Template active Template inactive !lat liver' 67 ± 6 • Chemical compos1t10n was determined as follows. DNA was estimated by absorbance at. 260 nm in 1 N N aOH (1.0 A 200 = 34 µg/ml). Chromatin fractions were extracted twice with 0.4 N H ,SO,. The nonhistone residue was dissolved in 1 N NaOH . Protein was estimated by a modification of the method of H . Kuno and H.K. Kihara [Nature (London) 216, 974 (1967)], using 2-~% TCA for precipitation. Purified hist.one and bovine serum albumin were used as · s tandards . b Average of five preparations plus or minus standard deviation. 'For DNA, average of three preparations plus or minus standard deviation. For protein, average of two preparations. Chromatin from both rat liver and ascites cells was incubated with DNase II for J.'i minutes prior to 1\IgCI, fractionation. amounts of chromatin DNA and nascent RNA remaining soluble at various MgC l2 concentrations. The bulk of labeled RNA remains in solution with a small portion of the DNA until the MgCl 1 concentration becomes sufficiently high to precipitate the active region. From the initial slope of the fractionation curve, it is estimated that the active region has been enriched 5.5-fold over unfractionated chromatin. (This value is estimated by assuming that all the labeled RNA is associated with the template active region. The enrichment of template active DNA would be greater if any redistribution of the labeled RNA had occurred.) More direct evidence for fractionation is based on DNA renaturation, RNADNA hybridization, and temp late activity measurements, and is reported elsewhere.12 P roperties of the Chromatin F ractions ) T he table shows the chemical compositions of chromatin fractions isolated from Novikoff ascites and rat liver. The template active regions " J .M. Gottesfeld, W. T . Garrard, G. Bagi, R. F. Wilson, and J. Bonner, Proc. Nat. Acad. Sci. U.S. 71 , 2193 (1974). -25- [8] FRACTIONATION OF DNASE- SHE.-\RED CIUW:\IATIN HI STONES Inactive Active ,-· i,,.c~ NONHISTONES Inacti ve Active >150,000MW - --n -j~ ,.. -~ Iii ~.;. Il b2 ___,./ II ~I___,./• .Jj -, • ....i - ! "\., 101 11 ~ ----- '2 - i j 11 - ... t~ ,, --r- ~ 17,000 MW~ ~ .,,~ FIG. 2. Disc elec trophoresis of chramasamnl protei n.-; from the template actire and inactive regions. Histone gels: Electrophoresis and staining was perform ed as deseribed by S. Panyim and G. R. Chalkley [Arch. Biochem. Biophys. 130, 337 ( Hl69) ]. Each gel was loaded with 30 µg of acid-extracted, ethanol-purifi ed protein . ?\ onhistone gels: Electrophoresis and staining was performed as des<'ribed by J. King and U. K. Lnemm li [J . Mal . Biol. 62, 465 (1972)). Acid-extracted chromat in was homogenized in and dialyzed against 2.5'/c sodium dode cyl sulfale, 65 m .\l Tris - Cl (pH 6.8). Prior to elec trophoresis, metastable aggfegateR we re disrupted by adding 5% tJ-m erca ptoethanol followed b)· boiling for I minut e. Gels were each lo:ided with 100 µg of protein. from both sources arc impoverished in acid-extractable protein, wh ile the t Pmplate inactive fractions are enriched in this material. Fraclionat.ion of chromatin abo yields an un equal partitioning of the nonhistone proteins. The actin rPgion contains about twice the amount of nonhistone protein per milligram of DNA as the inactive region . About 80% of the chromatin of both ascites and liYer is sheared by prolonged (15 minute) enzyme treatment. The amount of template actiYe D:\A found in the supernatant after l\lgCl, fractionation i-; strikingly different, however, in the two cases. The proportion of rat liver template active DNA is 2-fold higher than that of ascites, 24% versus 1~~ % . These values are in accordance with the preYious ly reported template activities of the two chromatins as measured \vith exogenous RN A polymerase. 13 Disc electrophoretic patterns of the acid-extractable proteins from the active and inactiwi fractions of rat liver chromatin are shown in Fig. 2. The most pronounced difference in the bandi ng patterns is the complete ab$t:>ncc of histone I in the actin region. Furthermore, although over 951/o "J. Bonner, l\I. E. Dahmus, D. Fambrough , R.-C. Huang, K. Marushige, and D . Y. I-I. Tuan, Science 159, 47 (1968). -26- 102 THE CELL NUCLEUS AND CHl:W)I.-\..TI'.'," PROTEINS (9) of the acid-extractable protein from sheared chromatin and inactive region fractions are histone proteins, ,ve estimate that only 60% of the acid-extractable protein from the acfo·e region is histone (from the amount of protein that is applied to th e gels and the intensity of the histone staining bands). On this basis th e true histone: DNA ratio of the template active fraction is less than 0.36, while the nonhistone: DNA ratio of this fraction is greater th an 0.70. It is clear that the nonhistone proteins are concentrated in the template actiw fraction and impoverished in the template inactive fraction. Separation of nonhistone polypeptides on the basis of mol ecular size by sodium <lodecyl sulfate disc electrophoresis is also shown in Fig. 2. The inactive region has a high er proportion of high molecular weight bands. In addition, quantitatiYe differences in the concentration of individua l bands exist (marked with arrows) . Comment Sheared chromatin can be fractionated into template active and inactive portions on the basis of predicted differences in the physical properties between the two regions. If separation without gross cross contamination is to be achieved, the size of the fragments is critical. :i'l'!ild digestion with D:N ase II has the obvious advantage onr other reported techniques in that it yields substantially smaller fragment s. Fractionation of chromatin with MgCl 2 provides a rapid and reliable means of separating active and inactive regions. Acknowledgments R epo rt of work supported in part by U.S. Publie H ealth Sen·ice Grant G i\'l 13i62, by a U.S. Public HC':dth Sen-ice pos tdor tor:11 fellowship (L.U. ), by II Damon Runyon M emorial Fund for Cancer Research postdoctornl fellowship ('W.G. ), by U.S. Public Health Service Training Grant G'.\l 86 (J.G.) and by the California Institu te of Technology Gosney FC'llowship (R.B.). -27- III. PARTIAL PURIFICATION OF THE TEMPLATE-ACTIVE FRACTION OF CHROMATIN: A PRELIMINARY REPORT The material in this chapter is reprinted by permission of the National Academy of Sciences from the Proceedings of the Academy, Volume 71, Number 6, pp. 2193-2197, June 1974. Reprinted from Proc. Sr,/. Amrl . Sci. r.·sA Vol. 71, No. u, pp. ~I!J:l- 2l!J7, J1111c 1~74 -28- Partial Purification of the Template-Active Fraction of Chromatin: A Preliminary Ucport (chromatin fraclionntion/D"iasc 11/D:\ A· HN A hybridization) J. }I. GOTTESFELD, W. T . G.\HH.\H]) , G. Jl.\GJ, n. F. WILso::-,;, .\:'<I) J ,\:\rES HONNER Divi~i~n of Biology, California In ~titutc of Technology, Pasadena, Calif. 9l 109 Co11trib-uted by Jam es lJ01111cr, March 18, W74 AUSTHACT A fra ction of rat-liver rliromalin tl111t is tran~criptionally a(·livc in. l'i«·o Ian~ hccn purifit•·d 6- lo 7- fold over whole chroruatin. This ~a'.'- uccun1plish c ,I h~~clectively ~hcariuµ c hrornutin \.\ith U.'\u:--c II followrd hy frnctionatinp: tlu_ • rc ka:-.t•d portion 011 th e ha-.is of its :-.oluhility properties i11 2 1n,1 .\Jg{:1:i_ . Th e n •:-. ultin1,: soluhl<: ntatf'rial c-ompri-.cs l 1r~ of tlu• total chron1ali11 U'.'.A a nd is in1p0Ycri:-.;hcd in lth,ton e and t~nriclu·d i11 nonhi s ton c protein. Cornparcd with un !-- h eun·d chro111Hti11. thi~ 111inor fraction exhihits n1arkt.·d diffcrc1H' t.":-; in c liro1nosnn1ul protein :-;pccics. l);\A renaturation !--tudic:-- irulicule that thi s fraction is co1npost•d of u !-\ ()Cc ific suhset of whol e µ:c11on1ul D~ .\ sc<1ucn ce~. F11rlh c r111ort- , l)'.'.:\•H;\A hyhridization exp<-rin1cnt:,4 !4Ul,!l,!Csl that alrno!'4l 60'·; of the nnnrt•pctitious l}'.\'A ~e<111c11ccs of tlii!:I 1ni11or fraction could code for cellulur H"IA. Differentia ted euka rrntic cells transcribe a limited and tissuespecific portion of till'i 1· nnclrar D:\':\ ~cqu ences (1, 2-). It is now well established t hat transcription is restricted in isolated chromatin (3) . Furthermore, recent evi dence suggests that at least some of t he lll<'chanisms of genetic regulation rema in intact in isolated chromatin (-1, 5) ..\t prese nt, howev1•r , 110 conclu,ive data are availa ble to indi t'ate whi ch componen ts of chromatin ser•;e a~ specific regulator., of geneti c ncti,·it,·. A direct nppronch to t hi s question would hr to st ndy the co lllponcnts of the minor portion of chromntin thnt is trunscriptionally active. This, howcvrr, rcquin·s as a prrr<'qui site till' cleveloplllcnt of l' hromat in fractionation tPcl111iqul·.s. Two principal prohlrm., arr !'IICOllntcn'd in designing a strategy for thr fractionation of chromatin. Fi rs t, to avo id cro,s-contulllination, chromatin must hr shca rrd to a size less th an that of t he nverage 1111it of transc ription. Idcull,· , th is shoulrl be accomplished by a method that does rv, t lead to protein denat11 rntion, rearrnngcrnent, or dissoc iati on. i:ieco11d, a gentle method of physicallv srparnting: "11ctive" from " inan ive" 111ate ria l is required . Scvr rnl gronps han' intrnducc·d fraction ation tech11iqtw.-; based 011 mcchanic·al shearing follo\\·ed b_,. separation l"· difforcnti:il crntrifu_e;ation or rolumn chromatography (6- 10) . .\n a lternative nw t hncl suggested bv workers in our laboratorv utilizrs D:\'ase II for shcurin.z, a~d selecti ve precipitation !;_,· mono- or divalent cations for chromatin fraction at ion ( 1 l, 12) . Previous stud ies !,ave s hown that !):\'ase II preferentially attaeks a minor portion of chromatin DXA; the amount of this "DNasr-lnhil!'" fnu·tion vari,'s clc-pcncling m1 th<' l'hrornatin sourer hut corresponds direc tly to the ability of th e give n Abbrevial ion : C0 t I molar rn1u·ent rat ion or DNA 1111< •lf' ot idr~ 1111111 i pliC'rl b~· lime of irH'uhntio11. 4 chromatin to serve a, a template for e,oi,;cnous RN.\ polymerase (13). Furth<'nnore, fr11ctio11ation expcri mC'llts usin !-! chromati n prrpan•d from hcputorna cells pulsc-lalirb l with ['![ Jurirlinc reveal that over 60% of the label frnc·ti onate,; with 10% of the d1romati11 D:\".\ (13). Thr.-e observations prompted us to e,:1mi11c this tcch11iq uc of chromati n frnrtion:1tion in more dPtail. In the present cornmu11ic:ttio11 we rPport that rut-liver chro1 11 ati n ha.s been fractionate d into " predominantly "acti,·e" ~omponent t hat differs in at least fiv e ways from whole chro matin: che111ica l cornpo.,ition, chroniosornal protein populations, template activity fo r support of R N ..\ sy nthe.-is, n :,,;.\ sequence complexity, and DX.\ sequence homobgy with cellular R:\' .\. ME11l00 S Chromatin Fractionation. Hat-liver chromatin purifiPd by sucro.se gradic'nt C<'ntrifu:.(ation (:J) ·.vas washed once with JO m:\I Tris·HCI (pl! 8) a nd dialyzed overnight a t 4' ag:ain.-t 200 volumes of 25 m:\l sodi um ac·etatc buffer (pH u.6J. Th e volume of the dialys:i te was ad ju, ted to µ;iv c an A;,~~~m of 10 (measured in 0.9 N Z'\aOH). The .solu tio n wa;; broug ht t,o 2-1°, ar,d D;\'ase II (Worthington, !-!!>AC) wa.s added to lOO units/ml. The rP:1<·tio11 was termi nated uft.rr 5-min incubation by the addition of 50 m\l Tris· l·ICI (pH 11) to pll 7 ..5 a nd cooling on ire. F n.,hea red chroma.tin (Pl) "-a.~ remo\·c•c.l hy cc ntrif11gntio11 ut 27,.500 X (} for 20 min nt 4°. To the s11pN11ata nt one ninety- ninth volume of 0.2 :\I \J g:CI, wa.~ ad,bl dropwise with s tirrinp; a t 4°. Mter ao rni11 of additional stirrin!!. the susf)('nsion was c-entrifugc•d as aliove yielding a pellet (P2) and supernatant (S2) fraction. Chromatin Protein Analyses. Histon<" and nonhi,;tonc protein were determined a, desc ribed (3) . .-\ cid~xtructed prntein was purifil'd h.,· .,rthanol precipitation and a na lyzed liy di.-<' electrophoresis as reported elsewhe re ( 14) . The arid-i nso luble chro matin rr , iciue wn, homogeniz<'d in and dialyzed n,(!:iin~t 2.A% (w/ v) sod ium dodccy l s11l iatr-G.i rn.il Tris· II C: l (pH fl.8)-2% 2-rnercaptoethanol. S:unplr,s w••rc heated for l min before electrophoresis (15) . Gel,, were scanned at 600 nm with a Gilford 2000 sprc trophotomcter. DNA Reassociation Kinetics . DXA was purified from various chromatin fractions as described else where (JG ). IJ:\' .\ iso lat,•d from t he S2 chromatin fraC'tion had u , ingle-strn nd c:d length of ,SOO nurlcotid,•:;, as d r tc•rn1i11rd by s,~cli111l'11lation wlocitv centrifugation under alknlinr ronditions (17) . D:\'.\ isolatP;I from total chroma tin and from the Pl chromat in fraetion wa•; shl':tr<'d li,· two pa., ., rs through u Rilii-Sorvall ____ _ __ __,:..,,_>u,;L __ __ -- --- 219-1 llio..J1c111i:;t.ry: G11tl(•sfcld ct al. -29- l'ropati'cs of cl,ro111n 1i11 _(ructions 'L\11u: I Crn npo:-:it ion rl'lative to 11:\A (w/w) .:\'011- Chromatin ~~ C hromat in T cmpl:tl c Hi , tonc hi.--tonc sample J):\A• :u·tivi tyt proteiut protein Unfractiouatcd Fraction JOO !) I . Li O.ul - --- Pl ~2 P2 84 G ± 4 s IJ .:i ± J !) 4. I ± 2 ,;) - - - - - · -- - - - ----· o.n.-, '.!O .OG (j,) o.,,s I .fiO - § * :\[enn of 11 Uct ennin nl ions ± ~I) as e:,;timntrd hy ah:-:urhu11c·c at 2GO um. t Perrcnt template activity (24) a, compared to IJ:\'A i,ulnt ccl from thf' :--nme chrmnutin ,..:n mpl c. }:,-;chf'richia coli HXA polym- erase (fr:i"t ion I 1· ) wns prepnred accurdi11g lo :\l cConnell und Bonner (2.-,). t Historic protein delcrmine<l from urc:ts of dcn.., itornetcr :-.cn11s 0f polynn.domidc gels ( 1-1 ) loaded wit h ku own amouul.s of protein~. § :\ot detcrmiu(•d . pres.5ure cell a t 50,000 lbs/ iueh'. This proc,•clure yields doubkstramkJ fragmrnt:; -100 to 450 nuclco tid r:; in ll'ngt h, as nl('asured by elec tron microscopy (18). Kinetil's of I):\" _-\ reas:;oc intion were monitored by hyJroxyupat itc chromatography using stand ard techniques (19). Duta obtainecl at various DNA and sodium ion corll'entrntion:; ll'l're norrnrilized to· Cot values cqui,·alent to thOS<' obtained in 0. 12 :\I phosphate buffer (0.18 :\I :\"a +, ser, ref. 19). ComputN analysis wa:; performed acrorJing to Britten et al . (19). Isolation and labelinp of Nonrepetitfre DNA. D N.\ was incubated to nn r•1 ui vak nt Cot of 2.5 X 10' (:32 l)I\A) or 1.5 X 10' (Pl D:\'.\), a nd the single-stranded fraction was isolated hy hydroxynpa lite chroma tography. This materia l was dialyzed against di stilled water, concentrated by lyophilizntion, dissol\'ed in phosphate bufTer, nnd allowed to renature as before. Finally, the DJli:\ tha t remained sin p;lestrnncled after two cydes of purification wa5 incubated to a Cot of 10'. The rrsulting duplex material wa,: dialyzed against distilled water nnd thrn contl'ntrated. The puri{ic~d nomepeti tive DN .-\ was labeled with "'I by the Commerford method (20) as modified in our laboratory (2 1) . Specific activities of 1 X 10' rpm/ µg Wl're obtained. DNA -RNA lfybridizalion. Total cell RN.\ from rat liver was isola ted by a modifi ed hot phenol-sodium dodecy l sulfa te extraction procedure (2'.!) and sheared by two pa.s srs through the Rihi-Sorvall pre:;sun· eell at 30,000 lhs./itlC'ii'. The resultinr; Il::-1.\ hnd an :\\'erap;e lcn~th of 1000 11ud1·otidcs, as judged by ,;edimentati on \'clority ce ntri ft1gatiun under nondenaturinr; co11diti011s ('.Z:3). Hybridization r ., pt•r inwnts were µerformc ·d at 7-1 ° in 30 111\1 ,odium pl,o,phatr- buff(' r (µII 6.5)-0.675 \I :'.\a('l - 1 111 .\ I Ell'!'.\ at t111 H:\ .\ :"··1-1:ibl'IPcl DJ\' ,\ rn:i .,,; ratio of 2 to :u; X 10' :I and an H:i .\ c·otH'cntration of :.?O 111µ:/ 111!. Sample·, wr•rr· i11C"11hatcd in st•a l, ·d rnpillnry tub,,,. Il e:wt.ions wr• n · t.c•n11i11at<·d h,· a :!0-fold dilution in reaction hnfTl'r at ti0 ° [olloll'l 'd i,_1· nppliC":1tio11 to h,·d r<> ,_,·apa titr colu11111.,equilihrat,·d ll'ith ;l() m:\I pl,osphnt" h11ff,•r·-0.l :\I I'roc. Na! . . \cad. S ,i. c: S.-\ ,t (lm.n Na Cl at 60°. l'11d,•t· thr.0 r condition., l>o lh sin.td(•- and douhks tranded IIU<'k ic aeid, are absorlwd whil,· in•e "'•[ !':ts::,•, thrn11gh. :-ii11;de-:-tra1ulc•d D N.-\ and tli<' 11 ,ajorit,· of the H:\".\ were eluted 11·itl1 0.1~ :\[ phosphatl' Li11 ffer; j));'_\·H:\'.\ duple,l'S were el11trd with 0 ..JS .\! phospl1:1te buffpr. .\ho11t :.?ll':o of the shear(•d RK.\ a lso rlutcd in OAS \I pho,ph:1te h11tf,r. The• singlc-strnnded a nd hybrid fra .. t io n:; m• r1• p rrl' ipita tc-d with 10% triddoroaeetic acid aft,· r nddit.inn of .JO µg of bo,·ineserum al b11111i11 prr :-ri mpk• . Th<' a111ount of DX.\ in I));,\ . R:\" ,\ hyb rid s wa, dct('l'rnined by colleetinr; t.he n':-llltinµ: precipitate., 011 rncmbrnne fi lters, nnd count ing; dried filter, in ri tolue11e-h:1sc•d srinti llant . .-\t zero time of incubation, abvut 2% of thr " '·l -hbeled DNA eluted from hyJroxy,1pntitc in th e 0.4S .\[ pl, osph ritc l,uffer fraction; this backµ;round \'n lur wa:; subtml'ted from all t ime point., . DN.-\· DK.-\ reassociation, estimall'J by ine11batio11 of '"!-labe led D X.-\ with NaO H-hydrolyz,·d R:i' .\ , wns not detectable abo,·e zero-time 1,in,ling. RESULTS Fract io nation of Chromatin Co)//ponen/s. The experiments desrribeJ he rein h:t\'e utilized a stn ndnnl set of condition:; for chromatin frneti o!1t1tion (sec .1/ethods fo r detai ls). C'hrnmat in is firs t selet:ti,·cly shcaml by inc ubati on with D.i\asc I[ for 5 111in . Uns hea rl'd r hrorna tin is remond by cent r ifugat ion , yielding a pellet termed l'l. This fraction compri.,ps 85% of the input DX.\ (Ta hlf' !) . Tlw res,ilti11g supernata nt (S i ) is fract.ionated on thr bnsis of it.s solubility in 2 rn:\l :\l;(C'l, into a serond supernata nt frac tion (:3:.?J nnJ a 111i11or insnlul,lc fru etion (1'2J. Th e second ,upern a tant frac tion (S~) co111priscs I 1% of the input D:\'.\. Frnction Pl ha~ :1 1'!1cm;cal compositi on not unlikl' thut of unfrnl'tionatr•d chromatin (Tab le 1). In contrast, fra ction S2 shows a i.;reat enrichment in nonhistone protein and a depiction in hi,to nes. F rnc tion 1'2 comprises only a trace of material undr·r the prese nt conditions of fra l' tio1uition; cnrl iPr s tudies using conditions resulting in more extensi ve shr:arinl{ hn\'e shown that fra ction !'2 co n.,is ts of D:\' :-\ complexed stoichiometrieally wi th histone (13). 'frmplate activity assays with exogenous polymerase rewa l that frnl'tion S2 chromatin is 3-[old supe rior to unfra ct ionat ed chromatin as template for H.K.-\ syn th esis (Table I) . In co ntrast, the t,•m plate actiYi t.y of fraction l'l is less than that of whole chroma tiu , Fig. I shows th e di sc clectrophorctiC' prnfiles of the chromosomal prntr ·i ns of frar,tions Pl and S2. Th~ popula tions of histone nnd nonhi:;t.om• prot eins of fraction l'l are ,;imi lar to tho.,p of unfractio11atrd rhromatin (data not ghc11'11), as one migh t predict, sinc e Ilic majority of chron,osmna l protein (n bout 85%J rema ins iii this irnction. In co11trn., t, thl· i>rotcin:; of fraction :'i2 arr quite d iffe rent. l!isto ne I i:; a li, ent a nd hi , t<JIH' I\' is pn•.se nt in a reduced proportion (Fi)!. 1.-l ). .\ prote in ba nd 1ni;srnli11g ,li.~h tl,· slower th a n histonc I appear.-: to be 1:mirhed in frneti on S~. The nnture of this co111po11r11t is 1111kno1rn, althouµ:h n hand at a simil ar position in whole ratliver hi stone prPp:1ration., hns been ,h,111·11 to have a turno,·rr rate at !vast 10--fold µ:reater tha n hi.st one I protein , 2G). :ionhi,;t<ine pol,·pc·ptidc.,; of fraction S2 ,h,,11· strikinµ: q11alitnti,·e and qua11tit:1ti\'C· difTc•n•n(·c•s us compared to tho,r of frnr:ti on l'l /Fig;. JR). It is of int erest thnt fraetion S2 is rid1 in two polypcptid,•:; in the molec: ular Wl'ight ranr;c• of 38,')00, the :1ppro ,i m:1t1 · sizl' o[ tlw , 1il11111its in,·oln•d in th,• p:11:ka.!!;ing of hl't <'rnµ:t• 11eo11s nudrar H:i ,\ (27J. Proc. ;Vat. :l ead. Sci. L'S .-! 71 (/!7:./.\ Fractionation of Chro11ut i11 -30- 2195 ,,' ,,,, /I I I I I ,, I I I \ ,--s2 I I ' I I m 11 I 11 I ~: II I 11 I ~ I : : I I I I I I I I I llr I I I I I I I I I I I nI , Pl I 1\ I P1 I, I I I I I l I I I ·, ,/ \ ,I\ I\ I I I rJ I llb 1 11 I I: I 1 I I I I I I I\ I I I I I I I I : I I / I I :\l~:'\J\(\: I I: \ I \ JI I . .) \, \/ ''1 \.I"', r,' S2 -✓ I I I I I \,' P1 Pl S2 S2 I I I I 140,000 73,000 55,000 28,500 I 17,000 MOLECULAR WEIGHT Fro. I. Disc elet! ropho rct ic pr<o fi lcs of the proteins of chromatin frnrti,,ns PI a nd ;:;:2. (A) Ili ,tonc protein. Ac-id-soluble protein (.-,O µg and 2,) µg} from c·hro1nat;.1 fraction:-- ~'.2 n1,J Pl, re . . . pe,·tivP!y, W(.)l'C :--eparatcd by ure.'.l.-d i.;r, ~el el<'"I rophe,r~:-;i:-- . (/1) Xonh i:-:to nc chrom~ soma! proteins. );onhi.; ton e pro1ei11 ( lt)(J µg) of each frartion w,c; sepa r:,ted by .sud inm dodeeyl ,uliate-<li.,c gel elcet rup horc.s:s . DNat Renaluralion I, inelics . The kinetics of reassocia tion of unfmctiona tcd chromat in D N".-\ and of frartion :-i2 I) );,\ arc preseuted in Fi g;. 2. Buth ~amplcs have ra :>idly, inten ncdiately, arvl slowly renaturir, .e: kin~ti<' rornponents, rcprb,•nting approximately 10%, 20%, and 7ll% of the inp,tt D?\.\, n·spec tively. For 11nfr:1dion:1tc,l D :\° ..\ the:;c components correspond to hi1;hly rPpr:titi,·r, mo<J,, rnle ly reµctitivc, und nonrepetit ive scqtH'1wc:;, and a;,rl'~ hot h in amou nt anu kinl!tic complexity with pulili.-,lwd ,latn (21). In eontrnst, ti,,. rutr of renaturation of fr:1et ion :'i2 I)\'_\ i:; ~tri kin!;IY Jifforent in t hat both the intcnnedi:itrlv a11<'. , low!_,· rl'annea li11 µ; ki1H'tic components h:c1·c :;ignifirnntly IPm·r C,,t .. , value:; t l1 t111 tho"! of whole gcnom:d D\' .-\ . If fr~c-t"m S2 !J:S,- .\ were cl"riv,·d from a random popula•.ion of ehro111 :1 ti n l JX .\ sequence;, its rr:i.-sociation c11n·e 1rn11ld l,c id cntira l to that of unfraet ior,atr•d DNA. The fact that both the i11tcrn1•:diutrly ancl , low ly rrannealing cornponrnt, vf lrar•t i,,n ::i~ D:\'.\ rr :ts,,;,·ia te fe.stc r than th0:;c of unfr:11' ti unated l 1:,,.\ ,-1t,:1rly indicut1·., that this material contains a .s pecifi c subset of the :;equen,·es of the rat genome. Th~ n•a:;.,o<'iation curl"<! for frac tion !' I IJX.\ is nearly identica l to th at for 11 nfrnc tio11a tecl D:\' .\ (da ta not shown); thi:; is n:uso nuhlr ,i ,H·,, f, w·tion l'l coruprisi's 85% of the ehrom11ti11 I >X .\. ltcn11 l~,,hlllrn hnth !,_,. ,·:1l,·1d:ili011 ,11 ,d !,_,· 1·x111•ri1111'11l11ti1J11 that the slowly n·:u1111':d111 .~ k11wl11· rn111po111'11l of fradiun :-i:.! DX.-\. co1w sponds to norm·petitive D:\'.·\ nnd not to some r~p~titi\·c comµoncnt. Fraction S2 contain,; 11.3% of the tota l ehrom:1ti 11 DN,\ (Table 1), a11d its slow kinetic cc:-nponcnt •-z w u 20 QC w n. 0 -2 -I 0 I 2 3 4 LOG EQUIVALENT C0 t Fi r;. 2. Hr.:1... :-a,ciation pro fi lc-s of unfraction:1.ted diromntin Dc\°A a nd fracti on :-i1 ll:'\A . ltcn:,t11rnti0n ui total t hn,matin D\" A ( A) and frnc: i1111 :--:! ll\' A ( □) wa., a.ssayed by· clm,m:i.togrnpliy on liydrux y,qntit,- ( l!I). Tiu: d1rcmin 1in I> :'\.\ po int s fall o n tl11• ro 11q ,1 1t1· r fit li, u· (.w, /i, / lim·) ,,f 1hr. d a 1a of f[ ,.l111c." nnd B111111 c r (2 1) Th,· li ,w tlir1111 g li I ii(' d :11 a for irnc·ti,m :--:·2 I >\'A Wl\."f oli!:1it1t'.d l1_y n :-: imilar L'l ll ll fHII C I' analy ...;i.-1. 2 l!)(j }'roe. XI/I .. lead. Sc-i. l'.':i.l 7' 1 11!1:~) -31- ---, - ·· ......_,. ' "·:,_, ;-:;-------.~1 0 ~ 90 z <I e: 80 <I) W7Q ~ '' vi 60 • t- '' (5 50 u ' '' a: ~40 30 '' . '' 0 - ~ - -~2 , - - - - - - 3 - -- - ' -- -- 20 ~. ~ ~ LOG EQUIVALENT C0 t F1G. 4. :. f_ labeled J)'.\A in the pre.,c11,·euf an ''"'e., .,of u11frnctio11:11,"l !J:s;A, Lauded 11011n,pcti1ive !J:s; .\ (.- et· .1/c-thr,d, ) of frnclim, :-;, ( ■ J or Fie:. :t Hca.ssocintion profilr-:- of i,olat<'rl nonrrprlitivl' 12 fract ~on Pl (-3.) wa.-; n1i xed wi1h unlah<'lcd lotnl d1roma1i11 l):\A in :lO m:\l pho., phat<• h11ffPr- 0.1;7,·, \ I ?\aCl ·· l m\l EIJTA , ~ W TIME (HOURS) 00 I ,oo ll yhrid i,.:1 tion of nonrcpetitive "'1-lahelc<l IJ:s; _.\ from frac·tion .--. Pl an<l ~:2 to :--hearcd liver HXA . The ma.-.:s r a ii, , of H\'.-\:1):s; _.\ for fr:1 c1i,,11 S2 IJXA ( ■) wns z.:, X 10':1, wh ile th:11 for fraf'tiun Pt u:-;_.\ ( ..i ) wa, 2.0 X 10': I. Tl,e upµ cr.scalr ,,n tht al,sci.~so, Hot, i .-; the produ<'l of tim (~ of in cubation ( in ;-.(•cond , J nn<l molar con,·enlrat ion of i::s;A (:,u m\l). dcun.tureU for :, 111 in at 100 °, and :illowcd to rcnal11rc at 74 ". The frnction of IJ '\,\ in dttpli,x wa, a." nyed hy hyc.lrn,yapa1i:e chromatograph_v a.-i de.--cri l>t:Li in .1/ rlhmls (scl'l ion 011 I~ :\ A· I):'\ A hybritli:rnt io11 ). Tl1c .-.; 1•,de tHI the <1/,.i:;,r 1·... ~·o., Pq11ival1~nt ( \ 1, n·f1-r...; 1n !he c-0 11 rc ntratior1 of 1111l:d1t·l('d I) \:\. ll:tla wcni olitni11r-d nt various cort<·e 11!r:llion s llf 1rnlaht: lcd l>;\A ; howc\'Pl' 1 the 111:i. ...;s ratio of tuthhell'd I)'\ ,\ lo lahl'led 11omt•pr.1 iI iv" I> '\,\ wa, maintained at :? lo.-, X IIJ'!: I. Cornpuler ntwly~is w :.i, 11,<·d to fil lines lo the dala ( l!l ). comprises 63% of this fral'tio11 (Fiµ;. 2). Since the rnt haploi<l µ;cnome contains 1.8 X 10" daltons of JJN .\ (28), tlw :u mlytical complexity of the slow compottcnt of fraction S2 UN.\ is (1.8 10 11) (U.ll3) (0. fj3J = 1.29 X 10 11 dalto11s. Thr ohscrvc, l kinetic com plexity of this c·ompone111. is 1.25 X 10 11 daltons (Fig. 2; relative to Hscfi,;ricliia coli, C'ul, , = -U). Thus, each seqlll't\!:C is n·1ircsc1,t,·d approxirnntdy 011ce i11 this kinetic portion of frac-tion S2 IJN.\. Thut thi s i, i11dee<l the case has hcc 11 dcmo 11.s t ra ll'd dir<'ctly !,.\' isolut in;.>; this c:0111ponent (sre .\Jcthuds ), la lll'linµ; it in vitro with " '•[, and rcanneali11;,; it i11 tlir pr0.,c•11,·1• of a \':tst cxc,•.ss of 1111frn,· tio11a tcd chroma ti11 llX.\ (Fiµ;.:;) . Th e C11t,_, o!tscn·ed for s11d1 lahc·led S2 DN.\ wa.s 1.3 X 10 3 . Similarl,,·, tl11: C' 0t, ., ol,-rn·Nl for the isolated nourcpeliti\'c· rompo 1H• 11t of frn!'li ott Pl J l);.\ was I .5 X 10' (Fi;,;. 3). The 111od1·rntcly r<'pet.iti\'e seq1t" 11 ccs of fraction S2 D N.\ also rq>rc'.,l'ttt a subset of tltc• n·pl'litivr seq ue1tc,•s of th(• ;,;c·11oml'. Thi, maltc·r will h,, disl'tl.ssc•d i11 dc·tai l clsewlwrr (l;lltl<'sfcld cf al., ;11 pr, ·paratio11) .. \1. pn·sr,11t it i 8 not dear wlll-tht•r till' fu :-; t tl':1:-:... <H'iati11µ; co111po11P11t of frnction S2 l)X .\ is mt:dog;vth lo lti;,;lily rcp,•liti ,·c· I J/'i ..\ ,,:- X quenccs, or whetlwr it n·pn•:--enb I>\".\ frag;ml'nts eu11t:i inin g; i11ternal 1·om1>h•n11·11tary .")t't!ll(' 11 ct•:-i. DNA -RNA llybr·i diwtion. Tiu, i.,olat,·d no11r<' p••titi\'(• "·1!labelccl DN.\ of Fig. ;J " ·a, a lso ltyl,ridizc·d to shcan·d, t ota l li ver IC\' .\ under c1>11ditio11, of vast H"i .\ ,•xc,·ss (Fig;. -l). Th e ~aturntion vr. lw·,, rst.imatl'd fro1n do11hlr-reriprn!'1t! plots of the data, we re 3.5% and 1-l.5% for fra ction l'l a nti ~2 11011n•petitivc DN.\ , resp,·,·ti,·l'ly. !11 a similar expcri11w11t, l:?.5% of frac :tion 82 11011n•1 wLiti\'l' l lN.\ 1,,·hridizc•d to u11sll('ared tota l li,·e1· H NA. The \'alul' of J.5% ohtaincd for frac-tio11 1'1 is in a,-i:ord \\'ith pu ldi.,ll!'d ,·:\ l<11•s for the ex tent of tra1t,niptio11 of 11onn·1u·titi\'t· :--1•q u1•nr<·s in mo11....;1•-lin' r ti.-..., w• t l , 2). Th,,,,, data s11;,;;,;l'.• l ti1al fr:H·tio11 S2 I)); .\ is ,•11 ri, ·!t,·d approx i- matcly -!-fold in ,;equcnces that rode for ce ll ula r l{N.\. ,re con,id0r this rnlue to he n n unclc-rc,stimatc, ., inec on ly 50~'o of th e· frnl'tiun ~2 D N.\ used in th,•,e e xperinll'nts was a b le to rcnnturc• to whol,: µ;rnoin:tl DN.\ at hi gh Cot ,·,tl11c •.,, ,d 1ile a bout 70% of frndion 1'1 ll.\".\ ,·01 dd form cluplt •x, •,; / Fiµ; . :1) . 'J'hC' r,•asOJh for thi:; arc' unk11ow11, hut mny It" cluP to bre a kdown clurinµ; hn11dling 1111d Ionµ; i1 1c1tbations. Thus, we consid e r the trn,· RX.\ hybridization :;aluration figure for fraction S2 11011r<•pe titive DX.\ to be almost 30% (or 60% if 01,e ass11mcs asymrnctric tra1tscriptio11). This would represen t a 6- to 7-fo ld enri,·l,m•mt i11 template aeti\'!· scq11c11ccs i11 frnetion 82 ]) :-;_\ over tlto,e in fraction Pl DX.\ . Furthermo re, the data oi Fi~. 2 s11~is<•,t. t.hat the nolirPpeliti\'e sequences of frn c-tion S2 have been enriclwcl h~- a similar factor [(C',.t, tota l si11~lc, copy DN.\J (C.,t , , S2 ,ini.,;lc copy D N ,\ )- 1] (15fJO) (225) -, = G.7 I, in agreement with tl1is predietion. DISCUSSION Based on several cri teria, the technique 11·e have a<lop!Pci for chro11, ati n frnl'tionation a ppea r:; to he succcs,;fu l. First of al l, the :;cqur nees of DN .\ found in fraction S2 con,i., t of a spec ifi c subset of the total genoma l sequences. To the bes t of our knowlc.,d.~c , this is the fir.s t clrmo11.,t.rntio11 ot sue h ,equ,•11ec fractionatio11 by renaturation kineties. Second , the nomc prtitivr scq u1,11('cs of fractio:1 .S2 I):\'.\ an· enric hed in thosr whi c· h ,·oclc fo ,· eellular l{N.\. This fi11 d i11;,; provides firn1 ,,,·idet\C'P for the parti al purifiealio11 of "te111plat.c ac·tivc" 1·hm11,uti1>, ori!!;i11ully s u;,;g;,·.,tc·d l,y 1•\prrim1·11ts. 011 the cofral'tio11ation ot' 1tusc!'1,t ]{);_\ ( t:J). Finally, chro111oso1nal proteins ha ve, hr,·n frntlionated lioth i11 a quantitati\'e and quulitati\'l' .,r11.,e. Oursu .. cr,s i1t tlH· isolatio11 of "t,·mplatc ucti,·,.•• · d1rrJ111ati11 can Ii•: <.•sti 111ut<·tl in at least two ways: from saturatio n ,·alu,.•., o!tt:1i1:,·cl by D::\'.\ · H:\' ..\ hyh ridi zatio11 experi,ne nt ., and frc,:n assay of templ:it<• aeti\'ity in t•itro. Both PtC'U,(lr('tn r 11ts yii,kl rnhw, of 50--t:i,5% purity of "l<' tn platc aeti\'e" chromnti n in frnc: ti on :32. ])::\'.\ reassol' iatio11 .,tudiP.s on fruc:tio11 S2 ,ho1,· a (i. to 7-fold emil'hmc11t of sinµ;l e-('()p,I' sequence:, o,·er tho,,. ,,f whole l'hromati11. ::iincc 50- 65% of these seque nces ar,, presumal,!y active in RN ,\ syuthcsis, isolation of pure· tc111plate acti,·c C' hrom a t i11 would rcquir1.· a 9- to 14-fold purifil'atir,n O\'l' l' wh0 l<~ chnimatin. Thi .-; su~_!!;P.,t-; that ,...;o nH' 7- 11 % ,.d whole l<'l\OJ\\a l I l::\' .\ is tra11,c riptio11all.,· activt•. Thi s ,•sti 1mHc • is i11 ac cord wi th publi.s hed HNA hyhridizuti,,11 dat.u (2) for -32- Proc. Nat .. lead . Sci. CS .I 71 U.'17 .',) the extent of f'.<·11..ti<: a!'ti,·i t_,· in li\"!'r if 011, • " " 11 11w, :isy111 mctric trn11s1·riptio11 (3.;i 5.,,';ti,J. It :;ho11ld IJI' 11,,t,·d ti1:1t 1111· co11tarni 11:ltio11 of fm,·ti"'' 82 l lX.\ ,•.-:th tr:111., .-ripti" ll:tily i1wrt seq 1.1t•11 ec•s i:=; not :t r:111dor11 prnt·P:-.s 1 i'or if it \\"I'/'('. tilt• n•as.,;orintio11 profile of fr:l('tiu11 :-'2 l>X.\ wo11ld IHll follow simple st•t·ond-onl1•r ki1wti ,· ... a... :--IJ1nn1 i11 Fig-.:!. Our fii1di11;,:s ,uµ;;,:rst wid,· npp,i rl1111iti f's fo r th,· npplir :itiun of this trch11iq11,• in the study of th,, control of µ;1·1H• Pxpn•s., i,,11. One is the enridn1wi1t t i,_,. almost one ord,•r of 1n:1e;11it1:d1·) from who le chro11i:1tin of th"·' " nomepctiti\'e I):,.; _\ .' l'qu r wt•., that arc cxpn•ssNI in that d1rom:1ti11 .. \ si•,·u!ld is the d in·,·t eom pnri .,on of I 1::-;-_ \ srqu,·111-.- 1·., p n·, ., ion in tlw 1·hrom:1tin nf different tis:;ues, orian., , or dewlopnH·11t11l states . .-\ third might be i11 the stud y of the fidelit_,. of n ·eo11.,titutio11 of chrnn1:1ti11 from it:; several ,·011stitur11,., . We thank llr. ..\ . l >«11v:1., and ,1 , ., r,. 11. Watsm , a11d .\ I. \\' ilkl's Fr::H'tio11 ati"11 of Ciirflmnti11 7. .,. ~im p--un, I: . T . & l:Pl'<·k 1 ( ; . I: . ( 1!17: >) Ui0d1rmi:;try 12, 10. \l ,·i.":1r1 hy, H.J ., \" i.- hiura, J. T. , ll11l'11 e,- k,. , I>.. :--i n, sc• r, ll. & ,l « h11,,.11, ('. 13. ( 1')7:l J Cc!./ Sµri11g llu r/u,r Sf1mp. !} «a11t. n:o /. 38, in pre;-;,-.:. .\ l:cn1,hi1,1·, K. & ll«111l!'r, J. ( 1!171 ) l'ro,:..\ 'a t. Anul. Sci. / 'S.-1 68, 2!111 - :!!IH. llil l; 11i,, I: . J. & 11,, ouer, .I. ( 197:! ) liio.·l. i111. Uio 71!111-,. Acla 28 1, -1.-,:;--11.i2. B 11 111u•r, J., (;an:1rd , \\' . T ., (;nltt'~ft>ld, J., l(ol111 t•:-- 1 I)_~-, i-',-,· ,,11, J. C: . ,~ \\'ilk,·.- . .\I. ( !!17:l) /'old Sprio f1 lla rl,or S ,,mp. () 110,,t . /J iol. 38, in pn•... s. 1';1n,·i 111 , S. & ('h:tlklcy, IL ( JUG!J ) Arel,. Hiod1c111. /liophys. 130, :l:l7-:Hli . l-i:i1111;, J. & L:1cm111li, ll. K. ( l'l7l)J. :1fol. !ho/. 62,4G.-,--177 . (:011esfcld, J ., B«1111Pr, .I., Hadda, (; _ K. & Walker, I. 0. ( 1~)7-l } Uiochrmi.-; tr y, in pn•~s. C:tuJ ier, F. \\". ( 1%.i l J ..\ fol. /J iol. II, :;7:1--.l'ltl. I h\"i.,, IL\\' ., :-in1n11 , \1. & i)uvidson, :'.'<. ( l!li I) .l/d<01 /s in 11. 12. }:{. 1-J. J.i. lli. and U. Hou~h nod :\fr. \V. Pt·:t r.,11 11 iur man_,· hPlpfnl <li ... ru:--.-.;i11ns 17. and for c-rit i('al rv al 11: ll ir,n of I hl' 111:w11s1-rip 1. Thi ... work wns Iii . I. 2. ;3_ Rrown, I. IL & Cl111rcli , I: . I\. ( 1!!71 ) Dct·,·lop. /l iol. 20, 7:l-S-1. Grouse, L., C hil :,11,, \I. IJ . ,le ,r..Carthy, B. J. ( HJ7:2 J Uiochcm i~lry 11, 7qs- KO.-J . Bonn er, J., Chalkley,(;. n., lbh11111:--:, \I., F':111 1hr1111 g h, 1 )., Fujimur:1 F., Huang:, IL C ., ll11herm:rn, J., Jp11 :--: {' 1J. IL , \[arushi ~e, K., Oiil<.·11h11., d,, II. , ()!iv.,ra, B. & \\"idl111lrn, J. ( l:lO:•q in .\lclho,{s. i11 H11::_11111oloq y 1 ed .... < ;n,:,. .. 111:t:!, L. & ,rold ave, K. (.-\cauemi<' l're.- s, '.\cw York ), Vol. I:!, part B, pp. :!-6.-,. Axe l, IL, C<"dnr, II. & Fcl.,e11fcld, G. (197:) ) !'roe. .\'o/. A cail. Sci . USA 70, 20:!!J-:.'ll:::! . Gilmour, H. C:. & Paul, J . (1\)7:l) l'rr,c . .\'at. Ar.n,/. Sci. FSA 70, :l~-11J- :lH:!. Fre11ster, J. II. , Allfrev, 1·. (;_ & \Iir.ky, ,\_ E. ( l!Hi:l) /'roe. Sat. Awd. Sci. CS.-! SO, 10:!6- IO:l:Z. 1 4. ;j_ 6. \I,·< ·c111:111~li_,·, H. L. & \l1·Carthy, B. J. ( l ~l7:2 ) l hodH·1111·.-.:lry I ! I q'.},'-i llltJ:i . .\1 -, qd,y, I-:. l •...Jr .. llall, C:. II. , :-h,.ph,·rd, J. II. & \\'piscr, 1:. :--:. I Pl/: ; ) l-i!o, ·/1111 ·1,-.;/(!f 12 , :;;,q: :- ::s:,:i . D. for a-;...; i,..;t:1.n1·c with thi s pr11j1·1·t. \\' e al.,o thank l)r,...;_ I). lfo\1nc~ support ed by 1hp t :.c:. 1'11!,lie ll>'a llh C:ervi"e ((: r:1111 ., (;\t -~li a11d G;\[ t:j76:2 ). \V.T.( }. ,,·:1s :t rec-ipic.·111 of a l>a111on Huny or1 Fnnd for Cancer l\ esc:m·I, h •ll11w.,hip Ill I: F-7.-,.; ). 2197 H11zymnlogy, eds. ( ~r o:---.;man, L. ,'\'.. \l oldave, }(. ( Ar·a<lemic I\J. 20. 21. :!:!. 2:). 24. :1.:;. 26. 27. :!R. Press, :'sew York ), \' ol. 2 1, part. D, pp. 41:,---1:!S. Britten, I!. J. , c:raha1«, I)_£. & :s!cufelu, Il. H. ( 197~) in .\I ct hod:-; i11 E11zy111olr,qy, in prc:--s. C«mmnf11nl. C:. L. ( 1!171 ) Hioclu:mislry 10, 1!)!):;- 2001). Holmes, ll . "· & Bo n11er, J. (1 !17-1 ) lliochc111i.stn1, in p ress . \Iuriyam:t, Y ., ll11d11Ptt, J. L., Pre:--t ny ko, A. \ V. & llu:-:c- h, H. ( l!lliVJ J ..\lot . li iul. 39, :;:).i-:l~~:-piri11, .-\."· ( !!Ju:!) l'roy . .\'11c. A cid HC8. I, :JO J..:;,1.-,. Li11, I'., Wil.",n, J:. & ll11 n11 cr, J. ( 11)7:: J .\lot. Ccll liiochcm. I, 1!)7-207. , 1,•C"1111cll , I) _ J. & Bo1111er, J. (HJ7:1 ) Hiochcll!i-<try II, 4:J2!J- 4:::i.;_ ( :arranJ, \\' . T . & llo1111cr, J. ( 1974 )./. Ilia/. Chem., i11 press . ,!art i11, T . !•: . & :-wif1, I I. ( UJ7:l) Cold Spriog llarl,oc ~!i"'P· Q11n11f. 1/iol. 38, in pre:--."\. Britten , H.J. & D1,vid.'<rn, E . I-I. ( 1!)7l)Quar/. Heu . !ho/. 46, 111-rn-:. -33- IV. SEQUENCE CO:MPOSITION OF THE TE:MPLATE-ACTIVE FRACTION OF RAT-LIVER CHROMATIN To be submitted for publication. -34- Summary Rat liver chromatin has been separated into nuclease-sensitive and resistant fractions after mild digestion with DNAase II. The " ++ -solu-ole nuclease-sensitive material is further fractionated into Hg and insoluble chromatin fractions. The kinetics of production of these chromatin fractions have been investigated. After a brief enzyme treatment (5 min at 10 enzyme units per A unit of chromatin 260 at pH 6.6), 11% of the input chromatin DNA is found in the Mg++_ soluble fraction. This DNA has a weight-average single strand length of about l.:oo nucleotides and, as determined by renaturation kinetics, compri ses a subset of middle repetitive and nonrepetitive DNA sequences of th e :-2.:. genome . Cross-reassociation experiments between DNA from the dif:'erent chromatin fractions show that a fractionation of whole genomal JI;:, sequences has been achieved. Uoreover, the Mg ++ -soluble fracti on of liver chromatin is enriched in nonrepeated sequences coding for liver RNA but not for brain HNA. experir:ents confirm that the Hg ++ Cross-reassociation -soluble fraction of liver chromatin contains a low proportion of sequences in common with the Hg fracti on of b rain chromatin. ++ -soluble ~1 hus, fractionation does not depend merely on some general property of chromatin but is specific with regard to the template activity of the tissue from which the chromatin ·.-ras obt2..ined. -35Introduction Previous work from this laboratory has shown that a minor fraction of DNA in interphase chromatin is more rapidly attacked by the endonuclease D:riAase II than is the bulk of chromatin :QirA (i'-Iarushi ge and Bonner, 1971; Billing and Bonner, 1972; Gottesfeld et al., 1974a). The runoun.t of DNA in this fraction is variable depending upon the source of the chromatin but is approximately equal to the proportion of DNA available in a given chromatin preparation for transcription by exogenous RNA polymerase (Billing and Bonner, 1972). The nuclease- sensitive fraction can be separated from the major portion of chromatin by sim~le procedures bas ed on the solubility of this fraction in either st2.nd2.rd saline-sodium citrate (Marushige and Bonner, 1971) or diva:ent c~tions (Billing and Bonner, 1972; Bonner et al., 1973; Gottesfeld et al., 197l1a). This fraction differs from either unfrac- tionated chromatin or the nuclease-resistant fractions in many respects: namely, chenical composition, chromosomal protein populations (histone and nonhistone proteins), template activity for support of RNA synthesis with exogenous bacterial polymerase, DUA sequence complexity, and DNA sequence homology with cellular RNA (Gottesfeld et al., 1974a). The nuclease-sensitive fraction appears to have the properties expected for transcriptionally active chromatin: it is enriched in nonhistone chromosomal proteins and depleted in histone protein (Marushige and Bonner, 1971; Gottesfeld et al., 1974a). It has nascent RNA associated with it (Dilling and Bonner, 1972; Bonner et al., 1975), and it is -36in a conformation similar to extended, native DNA in solution (Gottesfeld et al., 1974b). Moreover, this fraction is enriched in nonrepetitive sequences complementary to cellular RNA (Gottesfeld et al., 1974a). In this report, we expand upon our earlier investigations. We find that the nuclease-sensitive fraction is enriched in sequences that code for the RNA of the tissue from which the chromatin was obtained, but not for the RNA of heterologous tissue. From cross- reassociation experiments between the DNAs of the different chromatin fractions, we conclude that fractionation is highly DNA sequencespecific. We find that the nuclease-sensitive fraction contains a subset of single copy sequences and a subset of families of repetitive sequences. Preliminary data also indicate that repetitive and single copy se~~ences are interspersed in the DNA of the template-active fraction. These findings are relevant to suggested models of gene regulation (Britten and Davidson, 1971). Results Fractionation of Chromatin Figure 1 illustrates the fractionation scheme used in this work. Sucrose-purified chromatin (Marushige and Bonner, 1966) is incubated with DNAase II at pH 6.6 for various lengths of time, and the reaction is stopped by raising the pH to 7.5 with 0.1 M-Tris-Cl, pH 11. With purified DNA as the substrate, DNAase II exhibits maximal activity at pH 4.8, and 6-8% of maximal activity at pH 6.6. At pH 7,5, less -37- Figure 1. Fractionation scheme. -38- Chromatin (370 µ.g/ml DNA, 25 mM Na Acetate, pH 6.6) DNase II ( 100 units/m I) Centrifuge (27,000g, 15 min) ~ I Pellet t Supernatant 0 0 ! 2 mM MgCl2 Centrifuge (27,000g, 15 min) I Pel let Supernatant @ 0 -39than 1% of the pH 4.8 activity is observed. Undigested chromatin is removed from solution by centrifugation yielding a pellet (Pl). The supernatant (Sl) is further fractionated by the addition of .MgC1 to 2 mM. 2 After stirring at 4°c for 20-30 min, insoluble material which forms is removed by centrifugation. 1~e second pellet is termed P2 and the final supernatant S2. After 1-2 min of enzyme treatment (inset, Figure 2), all the DHA libeyated into fraction Sl is soluble in MgC1 fraction S2. On longer times of digestion, Mg (P2) is found in fraction Sl. ++ 2 and is found in -insoluble material On very long times of digestion (ca. 30-90 ni::1), 80% of the chromatin DHA is solubilized and .found in fraction Sl. The nature of the enzyme resistant 20% has not been fully investigated (Billing and Bonner, 1972). After 30 min of enzyme treatnent 20-24% of rat liver chromatin DNA. is found in fraction S2. With rat ascites chromatin, 10% of the DNA is found in this fraction (Billing and Bonner, 1972). About 2-5% of duck reticulocyte chromatin DITA is found in fraction S2 (Axel and Felsenfeld, personal communication). Billing and Bonner (1972) have noted a correlation between the amount of DNA in this fraction and the template activity of the input chromatin. The chemical and physical properties of the chromatin fractions are described elsewhere (Bonner et al., 1973; Gottesfeld et al., 1974a; Gottesfeld et al., 197lrb; Gottesfeld et al., 1975, in preparation). Figure 2. Time course of fractionation with mIAase II. Chromatin was fractionated as described (Results) and the amount of nucleic acid in each fractionation determined by absorption at 260 run of an aliquot diluted in 0,9 N Na0H. -41- PERCENT OF CHROMATIN DNA 0 0) N 0 co 0 0 0 0 PERCENT OF DNA 0 ~ ~ srr, u, ..--.. 3 OJ ::J 0 --f -~ rri- - 0 3 ::J u, (>J 0 N 0 -42The size of the DNA in fraction S2 was determined after various times of incubation with DNAase II (Figure 3). After 30 sec of enzyme treatment, 1.5% of chromatin DNA was found in fraction S2. This DNA had a nu.mber-average length of 1800 nucleotide pairs. With longer times of incubation, more DNA is found in fraction S2 and the size of this DNA decreases. Long times of incubation with the nuclease (15 min or greater) reduce the size of S2 DNA to below the limits of resolution by electron microscopy. After 15 min incubation, the weight-average length of S2 DNA is approximately 50 base pairs as determined by sedimentation velocity. After 85 min of incubation, 60% of the S2 DNA becomes acid soluble while the remaining 40% is recovered as 120 nucleotide fragments (Gottesfeld et al., 1975, in preparation). Kinetic Components of Rat DIIA and DNA Isolated from Chromatin Fractions A reassociation curve 01:' rat ascites nuclear DNA has been presented by Holmes and Bonner (1974a). We have obtained similar results with rat DUA p:r-epared from liver chromatin (Figure 4a). In both cases DNA was sheared to a number-average length of 350 nucleotide pairs (as judged by electron microscopy) and reassociation was monitored by retention of duplex-containing DNA on hydroxyapatite. Sheared rat DNA consists of at least three frequency components: a rapidly reassociating component consisting of highly repetitive and foldback sequences (Wilkes and Bonner, in preparation), moderately repetitive Figure 3. Length of S2 DNA as a function of time of chromatin digestion with DNAase II. Chromatin from rat liver was prepared and fractionated as described. S2. DNA was isolated from fraction Double-strand lenghts were determined by electron microscopy. - 44 - LENGTH (Base pairs) (X) 0 0 0 0 0 or-_ _ _ _ _ _or-----7lii--:Jc::====-"\-l=-ior----, • ---i ~ fTl (J1 0 Tl z 0 C CD l> ---i 0 Zo -3 :::, (J1 Figure h. Reassociation profile of rat DNA and DNA from the chromatin fractions. Data were obtained at various DNA and PB concentrations and temperatures; however, all data are normalized to Ct values 0 expected for 0,12 M PB at 62°C (Britten et al ., 1974). The lines through the data were obtained by computer analysis (Britten et al., 1974). A. Rat DNA. Three sets of data are included: circles are the data of Holmes and Bonner (1974a) for rat ascites nuclear DNA; triangles and squares represent the data for two separate preparations of rat liver chromatin DUA.· B. Reassociation profile of fraction S2 DNA . isolated after 5 min of nuclease treatment. described. Chromatin was prepared and fractionated as DNA from fraction S2 was subjected to chromatography on Sephadex G-200 in 0.1 N Na0H. The excluded material was dialyzed versus 0.1 f,1 ammonium acetate, lyophilized and redissolved in PB. C. Reassociation kinetics of fraction Pl DNA. Chromatin was incubated with DNAase for 5 min and the DNA from fraction Pl was isolated. This DNA was mechanically sheared to 350 base pairs with the RibiSorvall pressure cell. D. Reassociation of fraction Pl DNA. Chromatin was incubated with DNAase for 15 min; DNA was isolated from fraction Pl and sheared prior to renaturat ion . O w 0 100~80f60 f- 20 ~ 40t ~ w ~ o -I 0 O t Z 5 100 o ~ , 1 - 1 (/) \.0 1 -=t ~ 80 w u 0.... 5 601 40f2 -2 --aJ:1t •I I - uu.....,_ •I ~ A A.RAT DNA •I 3 B. S2 DNA ~ -, 2 t . . •I -I . 0 0 O o 0 ~ ~ •I \ro •I 0 ~ D. Pl-15 MIN DNA 0 ~ 2 740 i 60 1 --ISO . 1100 4 -i40 -160 -180 •Ir 100 C.Pl-5MINDNA I 1 1 ~ 0 O •i ~ V ~ o . + 1 -2 ~ -}- t •lti 4 C0 t 3 :l. .. . . . -~ 1. . . . . . .r LOG E QU I VALENT -47sequences which reassociate between Ct values of 0.05 and approximately 0 100, and a component of nonrepetitive (single copy) sequences. Second order reaction curves have been fit to these latter components by the least squares method of Britten, Grahar.i., and Neufeld (1974). The para.met e rs describing the moderately repetitive and single copy components of rat DNA are given in Table 1. The intermediate or middle repetitive region of the reassociation curve of rat DNA can be described equally well with eithe r one or two components: the root mean square (r.m.s.) of a computer analysis with a single middle repetitive component is 3.12% while the r.m.s. of a two-coc?onent analysis is 3 . 17%. To decide between the se two interyrete.tions of the data, we have isolated the middle repetitive sequence s of 3H-labeled rat ascites DNA and measured the kinetics of reassoci at ion of this DNJ\ in the presence of an excess of unlab eled rat DNA (Fi gure 5). The tracer was isolated by standard techniques of reass8ci~ti on and hydroxyapatite chromatography (Britten and Kohne, 1968). Hi ghly repetitive DNA was removed after incubation to Ct 0 0.1; singl.e-stranded DNA was then allowed to reassociate to Ct 50 . 0 The resulting duplex-containing material was used as the middle repetitive tracer. Reassociation takes place over 3-4 log units of Ct 0 (Fi gure 5), indicating heterogeneity in the population of middle repetitive sequences. A two-component least-squares analysis of the data has been performed (Pigure 5) and the Ct 1/2 values of 0 the components agree quite closely with the two-component analysis of the middle repetitive region of whole rat DNA reassociation (Table -48- Tnb1.C' J.. - - - - -- --··- -r::r~ ----·----- - - - - :,1. :! L ii ;._ ; ro lated fr''Jrn t. h·: Chrv!.: u'vin Fract.jons !". i , tr. ~~ - Ct:::-. p,: 1 ::':'J':~- cf i<;..:.t ·- -·--·-- - - - - - - - - - - - - - - ~ 1':·TA of 'rv ~a.1 Chromatin Ki~?t.ic Sc.inplf> t.!\ / . F'ruct:'... 0 :1 c1 f }J. ;.-i. l n Ct 1/2 0 O:.i!:•::,~veila. Cc,~.1-.:..H.:· :,'... Es~ imnt ed. Cct 12 Average K:ir. -2ti c fo1· i'1 re. C0! -:p.;.1 .t:,":t. c Ccmpl.ex1ty Pela tivc to E. c ol ic Repetition Fr~quenc/1 - - - - - - - - - - - - - ~ (_J'._o._s_e_P~T_r_s~- - - - - -- - - - o.o", !00 Chro!:1,i.t in r ::;, .::.r.1.1 :rr.1ed i::1.'te Fus t O.l~ C 0 . i:;3 0,095 1.0 X '.,.01 3,3 X 10 60 l, 4 X lO'J 1.0 1259 10? 6 3200 3.17.!g Fre.ction ~: ;, r,:,t. 0.109 (5 r.:in nu~ !cn:· ~ 0 . .1 'j 0.025 2,'( 0.C',5 2. 76 0.l' 1.7 X 10 5 90 1.6 X 108 1.2 0 . 628 Vci·y f,~ct () rnin nucJ ,, .~ :; e l r.!.e:·::,d iht c exvcsur.,) 8l .. E!4 }'raction rJ p ::,\ exposure ) .ee ?na•, 1,B C ; , -) ) R.?•LS . Fraction l'! ~:,~ C X 10 1721) 1.ey~~ O.C4 '{ L 0, 13:.. 0,51; 0.0'(l 8,0 X lO 0.175 33 5,8 3,3 X 1c6 50 0,(;::,t, n,16 1037 l. 2 x 10 9 l.O !< . ?-: . s . L , 3;,if 3160 0. ,j;O 4 C.11 0.018 2.C x 10 1L .b 1.06 2.1 X 1c6 202 lL~l.S. ( Footnotes C() 1'tinu<'J) L 0. Lf')", 3 , ;15;,:C 2.2 X 10 8 5400 40 1 Table 1 (continued) aCalculated from Figures 4a-4d. b Ct 1/2 for pure component= Ct 1/2 observed x fraction of DNA in 0 0 component. cThe complexity of!· coli is 4.2 x 10 6 Ct 1/2 observed for E. coli DNA is 4.1. o - -- base pairs (C~irns, 1963). The . The difference in base composi- ti on of rat (41% GC) and E. coli DNA ( 50% GC) slows the rate of renaturation of rat DNA by a factor of 0.83 (Wetmur and Davidson, 1968) relative to E. coli. This assumes that the components of each DNA studied are of the same average base composition as unfractionated rat DNA, ¾epetition frequency is calculated by dividing the chemical or analytic complexity of each component by the average kinetic complexity observed for that component. Chemical complexity is the total number of nucleotide pairs in a given component, and is obtained from the following expression: CC = G(fC )(fD) Where CC is the chemical complexity, G is the genome size of the rat (2.7 x 10 9 base pairs; Sober, 1968), fC is the fraction of total chromatin DNA in a given chromatin sample, fD is the fraction of DNA in the kinetic components. For example, the chemical complexity of the slow component of fraction S2 DNA is (2.7 x 10 9 ) (0.113) (0.628) base pairs. eData of Gottesfeld et al. f Data of Gottesfeld et al. (1974a) (1974b). gRoot mean square (r.m.s.) deviation of computer analysis (Britten et al., 1974). -50- Figure 5. Reassociation profile of isolated 3H-labeled middle repeti- tive rat DNA in the presence of an excess of unlabeled rat DNA. The niddle repetitive tracer was isolated from 3H-labeled rat ascites DNA as described (Experimental) and mixed with a 200-fold excess of 350 nucleotide-long rat liver DNA in 0.12 M PB (0) or 0.48 M PB ( D). The abscissa refers to Ct values of driver rat DNA. The 0 low Ct points (~) were obtained in the absence of driver DNA·, tracer 0 Ct values were multiplied by a factor of 5 to correct for the fraction 0 of middle repetitive DNA in whole rat DNA (20%). 'l'he line through the data describes a two-component least squares fit. values of the two components are 1.0 and 27.5. fit is 3.8%. The Ct 1/2 0 The r.m.s. of the ~I I w 0 z 0 ' I -2 Size I -I 0 cot 1/2 I 0 . EQUIVALENT 1 37% 1.0 2 36% 28 RMS = 4.07 % ~ Component I I00L <! 80 0:: Cl) r- 40 w 60 _J z<._9 - Cl) I- z W 20 u 0:: Q_ w 0 -3 LOG I L I I') I C0 t 3 -521). We conclude that rat middle repetitive DNA is best described by two components. These components consist of families of sequences which are repeated, on average, 60-70 and 2000-3000 times, respectively, per haploid genome. Chromatin from rat liver was treated with DNAase II for 5 min and then fractionated as described above (Figure 1). The DNA isolated from fraction S2 was determined to have a number-average double stranded length of 700 nucleotide pairs, as estimated by electron microscopy; however, alkaline sedimentation velocity studies revealed a weightaverage single strand length of 200-600 nucleotides, with a mean of 38 0 nucleotides (four determinations). S2 DNA was purified by chromatoe;raphy on Sephadex G-200 prior to use in reassociation experiments. The data of Figure 4b provide a reassociation curve for fraction S2 DNA of rat liver chromatin. 5 min of enzyme treatment. Fractionation was performed after A second order fit of the data of Figure 4b has been performed (Britten et al., 1974) and the parameters describing the kinetic components are presented in Table 1. The DHA of this fraction contains four kinetic components in nearly the s ame proportion as found in Dl'IA of unfractionated chromatin. We do not know whether the rapidly reassociating component of this DNA corresponds to highly reiterated simple sequence DNA or to fragments containing inverted complementary sequences (fold backs). The inter- mediately and slowly renaturing components of fraction S2 DNA reanneal -53at lower Ct values than their respective counterparts in whole rat 0 DNA. This increased rate of reassociation is not due to differences in fragment length or reaction conditions. Since a random population of DNA sequences would renature with the sane kinetics as whole rat DUA, we conclude that fraction S2 represents a specific subset of DNA sequences of the rat genome. This proposit ion will be substantiated below. DHA from fraction Pl obtained after 5 min of nuclease treatment was found to have a weight-average single strand length of 3300 nucleotides (aD:aline sedimentation velocity). 7h i s DNA was sheared to a nw~ber-average length of 350 nucleotide pairs and reassociation kinetics were determined. Fraction Pl DNA and unfractionated chromatin DNA reanneal with similar kinetics (Figure 4c and Table 1). This is reasonable since fraction Pl contains 80~85% of the input DNA. After 15 min of nuclease treatment, fraction Pl contains 24% of whole ~hromatin DNA. This amount does not decrease after up to 1,5 hr of incubation with DNAase. DNA prepared from fraction Pl (15 min nuclease exposure) was found to have a number-average double strand length of 800 nucleotide pairs. This DNA was sheared to 350 base pairs and reassociation kinetics were determined (Figure 4d and Table 1). About 30% of the DNA failed to reassociate, presumably due to degradation caused by the extensive DNAase treatment. Under standard reassociation conditions (0.18 M l!a+, 62°), 33% of whole rat DNA fragments (350 base pairs in length) contain repeated sequences . On the other hand, 50% of the hybridizable Pl DNA contains repeated -54sequences. Thus, Pl DNA appears to be enriched in repeated sequences. * The middle repetitive and nonrepetitive components of Pl DNA reassociate 2 to 5 times faster than their respective counterparts in whole rat DHA. This suggests that fraction Pl obt a ined after 15 min of nuclease treatment contains a subset of whole genomal DI/A sequences. Cross-P.eassociation Experiments To test the degree of homology in DNA sequence between the chromatin fractions, we have performed cross-reassociation experiments. Figure 6 depicts the reassociation of labeled Pl DNA in the presence of an excess of either unfractionated chromatin DNA or fraction S2 DNA. When driven by whole chromatin DNA, the labeled Pl DNA tracer reassociates with kinetics essentially identical to the kinetics of the driver. On the ot:1er hand, when fraction S2 DNA is used as the driver (upper curve, Figure 6) reassociation is limited to 28% at a driver Ct 0 of 2000 . :Sy this C t, 90~~ of S2 DNA has reassociated (Figure 4b). O As much as one-third of the observed Pl tracer reassociation could be due to self-reaction since an S2 driver Ct of 2000 corresponds 0 This conclusion rests on the assunption that the nonhybridizable Pl DNA is a random population of sequences and not specifically single copy sequenc es. From our knowledge of chror.,atin structure, it is hard to envisage a mechanism for the specific degradation of single copy DNA. -55- Figure 6. Cross-reassociation of S2 and Pl DNAs. Chromatin was incubated with DNAase II for 5 min, fractionated and DNA prepared from the fractions as described. Fraction Pl DNA was labeled in vitro with 125 1 (Commerford, 1971; Holmes and Bonner, 1974a). Fast reassociating material was stripped from the labeled DNA (Experimental). In the lower curve 125 1-Pl DNA (0 ) was mixed with total rat liver chromatin DNA (b) at a driver to tracer ratio of 12,800:1. 125 upper curve In the I-Pl DNA (0) was mixed with fraction S2 DNA (5 min nuclease treatment) at a driver to tracer ratio of 6250:1. The abscissa refers to Ct values normalized to standard conditions (0.12 M PB, 0 62°C) for the concentration of driver DNAs. The line through the data of the lower curve is the computer fit of the reassociation curve for rat DNA (Figure 4a). The upper curve represents a least squares fit of one second-order component. The root mean square of the fit is 3.4% and the Ct 1/2 value observed is 600; the reaction 0 goes to 28% with 23% in this single component. -56- PERCENT SINGLE STRANDED 0 N 0 0) (X) 0 0 0 0 I N r 0 0 C) fT'1 0 0 < l> C r fT'1 z -I (") - 0 N -57to a tracer Ct of 0.32 (Ct 2000 divided by the driver to tracer 0 mass ratio). 0 AC t 1/2 of 600 is observed for a single ·transition 0 in the Pl DNA tracer reassociation. This should be compared to a driver Ct 1/2 of 225 for the nonrepeti-t;ive portion of S2 DNA reaso sociation. L'1.ese data suggest that most (60-70%) of the DNA sequences present in fraction Pl are absent from fraction S2. It is possible, however, that the reassociation of Pl tracer has not terminated by C t 2000 of the S2 driver DNA. 0 If f'urtl1er reassociation was observed at higher Ct values, this would indicate that Pl sequences are found 0 in low abundance in S2 DNA. With this reservation in mind, we may still conclude that fractionation according to DNA seq_uence has been accomplished oy the methods employed (Figure 1). ~'1en fraction S2 DNA was used as a tracer and fraction Pl DNA was used as a driver, very different results were obtained. The S2 DNA tracer reassociated to the extent of 70% when unfractionated 4 chromatin DNA (Ct= 10 ) or fraction S2 DNA (Ct= 5 x 10 3 ) was 0 used as~ driver. 0 When fraction Pl DNA was the driver, 53% reassociation of the S2 tracer was observed at a similar driver Ct. 0 The Ct 1/2 0 values for the slow components of both tracer and driver DNAs were comparable (1.07 x 10 3 and 0.86 x 10 3 ). These data indicate that after 5 min of nuclease treatment, most S2 DNA sequences (>75%) are still found in fraction Pl. In agreement with this finding, Figure 2 shows that not all S2 chromatin is separated from Pl chromatin by fractionation after a 5 min exposure to DHAase. -58To determine the degree of tissue specificity in the DNA sequences of fraction S2, we isolated fraction S2 from both liver and brain chromatin. The chromatin was treated with DNAase II for 5 min, frac- tionated, and DNA was prepared. in vitro with 125 S2 DNA of brain chromatin was labeled 1 (Commerford, 1971; Holmes and Bonner , 1974). The labeled brain S2 DNA reassociated to 82% with whole rat DNA as a On the other hand, when liver S2 DNA was the driver, the driver. reassociation kinetics of Figure 7 were obtained. sociated to only 31% at a driver Ct of 2000 . 0 Brain S2 DNA reas- At this Ct, the liver 0 S2 driver DNA reassociated to 90% (lower curve , Figure 7). A computer analysis of the data suggests that brain S2 DNA might reassociate 4 to 40% at higher Ct values (>10 ); however, we have no data to support 0 this. Furthermore, control experiments show that 10% out of the observed 30% reassociation at Ct 2000 could be due to tracer selfo reaction. We take these data to mean that .chromatin fractionation is decidecly tissue-specific; we estimate that no more than 25 to 40% of the sequences of brain S2 DNA are present in the major frequency components of liver S2 DNA. Further Studies on the Kinetic Components of Fraction S2 DNA We indicated earlier that the slow component of fraction S2 DNA (Figure 4b) corresponds to nonrepetitive sequences. This can be shown directly by isolating this component by hydroxyapatite chromatography, labeling the DNA in vitro and reannealing in the presence of an excess of unlabeled whole rat DNA (Figure 8). The labeled -59- Figure 7. Cross-reassociation of brain and liver fraction S2 DNAs . Chromatin was prepared from brain and liver, fractionated, and DNA isolated as described. Incubation with DNAase II was for 5 min. Brain S2 DNA was labeled with 125 I, and the labeled DNA was mixed with unlabe.led liver S2 DNA at a liver DNA to brain DNA ratio of about 2000:1. Both DNAs were chromatographed on Sephadex G-200 (0 . 1 N HaOH) and the excluded material was utilized in these experiments . Incubations were in o.48 M PB (74°c). The abscissa refers to Ct 0 values of driver liver S2 DNA, corrected to 0.12 M PB (62°c). The lower curve is a reassociation profile of liver S2 DNA (Figure 4b); the upper curve is a least squares fit of the data (root mean square = 0.8%). The computer fit describes two components, one (7.4%) with a Ct 1/2 of 50 and another (23.9%) with a Ct 1/2 of 1500. 0 extrapolated final reaction is !~1%. 0 The -60- PERCENT SINGLE STRANDED ,o I\) (j) (X) 0 0 0 0 0 I\) , - - - - - , - - - - - - T " - - r - - - - r - - - - - - - - r - - - . - - - - r - - - - - - - - r - - - , - , , - - - - - , 0 r 0 G) rn 0 0 C ~ r rn z -, 0 0 - I\) I I I I -61- Figure 8. Reassociation profiles of isolated nonrepetitive 125 1- labeled DNA from. fraction S2 in the presence of an excess of unfractionated DNA. Nonrepetitive DNA from fraction S2 was prepared and 2 labeled as described (Experimental) and mixed with a 2-3 x 10 -fold excess of 350 nucleotide-long unlabeled rat DNA (0). Labeled DNA was also isolated from DNA-RNA hybrids either with ( 0) or without (6) ribonuclease treatment (see Experimental) . The labeled DNAs were mixed with sheared , unlabeled whole rat DNA at a driver to tracer ratio of ei tber 970: 1 ( □) or 31,000: 1 ( 6) . to Ct values of the driver DNA. 0 The abscissa refers The line through the data is a . least squares second order fit of the experimental data for the initial tracer D~JA ( 0) ; the C t 1/2 is 1. 3 x 10 0 of the fit is 3.1%. 3 and the root mean square -62- PERCENT SINGLE STRANDED 0 (.;J (Jl 0) 0 0 0 -.J 0 (X) 0 0 <.O 0 0 r - - - - . , . . . - - - , - - - - - r - - - ~ - - ~ - - - r - - - - , ---, 0 0 r 0 G) N C> 0 rn 0 C j§ r-1 rn ~ 0J 0 0 (Jl L-_ _,L.__....J._.J___ __.___ _...L-_ _-1..-._ _ _ _ _ ___. -63DNA reassociated with only the nonrepetitive sequences of the driver DNA; the C t 1/2 observed for the tracer DrJA was 1. 3 x 10 3 while 0 the Ct 1/2 observed for the slow component of driver rat DNA was 0 2.0 x 10 3 (Figure 4a). The Ct 1/2 value of labeled S2 DNA could 0 be in error since terminal dat a was not obt a ined (Figure 8) ; further reassociation would increase the value of Ct 1/2 for the reaction . 0 However, this would not alter the conclusion that the S2 slow component is derived from single copy sequences of whole rat DNA. The labeled DNA reassociated to only 50% completion, presumably due to breakdown during the several cycles of denaturation and incubation necessary to purify this material (see Experimental). From the reassociation profile of :fraction S2 DNA (Figure 4b), it can be calculated that if >fraction S2 represents a specific subset of the rat genome (11.3%) , then the slow component reassociates at the'rate predicted for nonrepeated sequences (Table 1). We may therefore conclude that since this slow component is derived from the nonrepeated sequences of whole rat DNA, and since it reassociates 8.5 times more rapidly than whole rat DNA, this fraction must represent 11-12% of the complexity of whole rat no:::i.repetitive DNA. We next turn attention to the moderately repetitious DNA of fraction S2 , The computer analysis of the reassociation curve of S2 DNA (Figure 4b, Table 1) shows that the intermediate frequency components comprise a limited fraction (5-27%) of the complexity of whole rat middle repetitive DNA. To obtain a more precise estimate of this kinetic enrichment, we have isolated the middle repetitive -64DNA of fractions S2 and Pl and determined the kinetics of reassociation of these DNAs (Figure 9). The middle repetitive DNAs were isolated by standard techniques of reassociation and hydroxyapatite chromatography; Ct values were chosen to minimize contamination with highly 0 repetitive and nonrepetitive sequences (see Experimental). The DNAs were labeled in vitro by treatment with E. coli DNA polymerase I, 3 • H-TTP, and unlabeled deoxynucleoside triphosphates. The labeled DNAs were then mixed with an excess of unlabeled S2 or Pl middle repetitive DNA, and reassociation kinetics were determined (Figure 9). The reassociation of S2 and Pl middle repetitive DNA takes place over four to five log units of Ct (Figures 9a, 9c) . 0 These data confirm the result of Figure 5; middle repetitive DNA of the rat consists of a spectrum of frequency components. Computer analysis suggests that the reassociation curves are best approximated by two second-order kinetic components. The size of these components and Ct 1/2 values are given in Figure 9. 0 The data demonstrate that chromatin fraction S2 contains a smaller proportion of the complexity of whole rat middle repetitive DNA than chromatin fraction Pl. From the data of Figures 9a and 9c, we calculate that the middle repetitive complexity of fraction S2 is 5-7% that of fraction Pl . The kinetic complexities of the components of Pl and S2 middle repetitive DNAs are listed in Table 2. The complexity of the S2 fast reassociating component borders on the complexity of simple sequence DNA. This complexity is 6-8% that of the fast component of Pl middle repetitive -65- Figure 9. Reassociation curves of 3 tt-labeled middle repetitive DNAs from fractions Pl and S2 in the presence of excess unlabeled Pl and S2 middle repetitive DNAs. A. Middle repetitive DNA of fraction S2 was labeled by nick translation (Experimental) and mixed with unlabeled S2 middle repetitive DNA at driver to tracer ratios of 1200:1 (0) and 5:1 ( □). B. Pl middle repetitive DNA. The driver to tracer ratios were 2600:1 (0) and 450 :1 (0). C. S2 tracer in the presence of unlabeled Pl middle repetitive DNA, labeled by nick translation, was mixed with unlabeled Pl middle repetitive DNA at driver to tracer ratios of 5700: 1 ( □) and 1000: 1 ( 0). in the presence of S2 middle repetitive D~A. ratios were 4500:l (0) and 900:1 ( □). Pl tracer The driver to tracer All reactions were carried out in 0.12 M PB in sealed capillary tubes . the concentration of driver DNAs. D. Ct values refer to 0 The labeled DNAs exhibit 7-8% zero time binding to hydroxyapatite in 0.12 M PB. The data have been corrected for zero time binding according to the equation of Davidson et al. (1973). The lines through the data are the best two-component least-squares analysis (Britten et al., 1973). component sizes and Ct 1/2 values are given for each curve. 0 The I \.D \.D I w 0 0 z <t 0::: Iif) ~ 0 I 0 0.001 0 0 C t 1/2 22% 0. 18 Size 0 1 48% 1 Component 26% Size 1.3 0 .023 cot 1/2 RMS = 7.3% -2 -I 5 1% -3 2 ~ RMS = 5.4% 2 Component 100;,-.....__ 801-60 40 0 20 ,_J (9 100 w z -if) 80 60~ I- w z u w 0::: 0.. 40 20~ o· -4 2 0 0 0 0 ~- S2 ~R/S2 ,MR 0 2 0 I 0 2 B. S2 MR/PI MR 0 t + - 3 I cot 1/2 • Size 0 .013 I 25% 3.3 - 1 57% Component 2 RMS = 4.9% Component 33% Size 0.034 cot 1/2 ~ 1 60% 83 RMS = 4 .4% -2 -I 2 -3 LOG EQUIVALENT C0 t I I I C. Pl MR/Pl MR '\. -"" -j40 -l60 -l 80 • -1100 0 3 0 20 --140 ---l60 -J 80 100 0 ~-l20 2 ~ I 0 D. PI MR1/S2 MR 0 -67Table 2. Complexity of Middle Repetitive DNA DNA of Chromatin Fraction Complexity of Middle Repetitive Components a (base pairs) Fast Component (1) Slow Component (2) S2 270 107,000 Pl 4000 2,000,000 aKinetic complexity calculated relative to E. coli (see Table 1, footnotes). -68DNA. The complexity of the S2 slow repetitive component is 5.2% that of the Pl slow repetitive component. The slow repetitive component of fraction S2 could accommodate 350 families of sequences each 300 nucleotides in length while the slow repetitive component of fraction Pl could accommodate nearly 7000 different sequences of this length . The complexity calculations of Table 2 are more likely to be accurate than those of Table 1 since the present calculations are based on the reassociation curves of isolated components. The calcula- tions of Table 1 are based on computer analyses of small regions (<25%) of the reassociation curves (Figure 4). Furthermore, scatter of the eX})erimental points (especially Figure 4c) makes calculations based on these data tenuous. The reassociation experiments of Figures 9b and 9d were performed to estimate the degree of cross-contamination of middle repetitive sequences between the chromatin fractions. When labeled S2 repetitive DNA was allo~ed to reassociate in the presence of an excess of unlabeled Pl middle repetitive DNA, the data of Figure 9b were obtained. This reassociation curve is essentially identical to the reassociation curve of Pl middle repetitive DNA (Figure 9c). Since the S2 tracer reassociates to the same extent with either S2 or Pl middle repetitive DNA as the driver, we conclude that all sequences of S2 middle repetitive DNA are present in fraction Pl. However, the S2 sequences com- prise a low percentage of the total complexity of Pl middle repetitive DNA. The reassociation kinetics of Figures 9a and 9b allow us to calculate that S2 repetitive sequences make up between 4.4 and 13.8% of the complexity of Pl middle repetitive DrlA. This calculation was made by comparing the Ct 1/2 values and sizes of the kinetic 0 components. If we compare the overall Ct values of the curves of • 0 Figures 9a and 9b, we calculate that S2 sequences comprise 11% of Pl middle repetitive complexity. The data of Figure 9d allow us to estimate the contamination of Pl repetitive sequences in S2 middle repetitive DNA. In this experiment labeled Pl DNA was used as tracer and unlabeled S2 DNA was used as driver. Since the tracer reassociates to about the same extent with either Pl or S2 DNA as driver (Figures 9c and 9d), fraction S2 must contain all Pl sequences. We note, however, that during the last phase of reassociation (at Ct values >50) tracer selfo reaction could contribute significantly to the observed reassociation. Reassociation of the tracer has been divided into two components: one component (33%, Ct 1/2 = 0.034) reassociates over the same range 0 of Ct values as the major portion of S2 driver reassociation (Ct 0 0 The second component (60%, Ct 1/2 = 83) reassociates 0.005 to Ct 1.0). 0 0 nearly 500 tines slower than the slow component of the S2 driver (Ct 1/2 = 0.18). 0 This suggests that about 60% of Pl sequences are_ present in the S2 middle repetitive DNA preparation in extremely low concentration ( 0 8 8~ = 0.2%). This could be an underestimate since Pl sequences in S2 DNA might have been partially removed by the Ct 2 fractionation step used in the preparation of S2 middle 0 repetitive DNA (see Experimental). The data of Figure 9d suggest that about one-third of the mass of Pl repetitive DNA has complements -70in S2 DNA at significant concentrations. The experiment depicted in Figure 9b showed that about 11% of the complexity of fraction Pl repetitive DNA is due to the presence of S2 repetitive sequences , Table 3 summarizes our knowledge of the distribution of repetitive sequences in fractions Pl and S2. We have indic ated our estimates of the degree to which each kinetic component may contribute to the middle repetitive DNA of fractions Pl and S2 . To obtain an estimate of the f'raction of total middle repetitive DNA complementary to S2 DNA we performed the experiment of Figure 10. In different experiments 3 H-labeled and 125 I-labeled middle repetit :..ve DNA from unfractionated chrorr.atin was mixed with an excess of S2 DNA (unfractionated with respect to kinetic components). Figure 5 provides a control for this experiment, the reassociation of 3Hlabelec middle repetitive DNA in the presen~e of an excess of DNA from unfractionated chromatin. above. This experiment is described in det ail W-nen the middle repetitive tracer was used with unlabeled S2 DNA as a driver, the results of Figure 10 were obtained. Data for three separate experiments (three different tracer preparations) are included, and data are normalized to 100% reassociation of the tracer when driven by whole DNA. The reassociation of the repetitive tracer takes place over at least four logs of driver Ct. 0 A two component lea st squares fit of the data is shown in Figure 10. I M t-I Table 3, Chromatin Fractionb S2 Pl 0 0.18 Slow Repetitive 3,3 83 Ct 1/2 of Repetitive Component Fast Repetitive 0.034 0.013 51% (9a) 0.2-5% d Composition of Middle Repetitive DNA of Chromatin Fractions Pl and 82a 0.001 Percent of ReEetitive DNA in ComEonent (bi mass) 48% ( 9b) possible (9a) 26% (9a) 22% ( 9b) 60% 10-20% (9a)c 57% (9d) possible (9c,d) (9c) 33% (9d) 5-10% 25% (9c) <2-4% (9c,d) aFrom data of Figures 9a-d, bChromatin fractions were obtained after 5 min of nuclease treatment. cindicates that this estimate was obtained from Figure 9a, etc. dThis is an underestimate since some of this component, if present in S2 DNA, would 0 have been lost from isolated S2 middle repetitive DNA by Ct 2 fractionation (see text). -72- Figure 10. Reassociation profile of labeled middle repetitive rat DNA in the presence of an excess of unlabeled fraction S2 DNA. Middle repetiti7e DNA was prepared from either 3H-labeled ascites nuclear DNA( □) 125 or from unlabeled rat liver DNA and labeled in vitro with r (O,ti). The ratio of unlabeled S2 DNA to tracer repetitive DNA was 2000:1 (0), 1400:l (ti), 3000:1 ( □). S2 DUA driver Ct values. 0 The abscissa refers to The line through the data is a two component least squares fit (root mean square = 4. 5%). Component 1 ( 37. 5;; ) has an observed Ct 1/2 of 0.33, and component 2 (58%) has an observed 0 -73- PERCENT SINGLE STRANDED m ~ N ro o , 0,-----.-------.o--r----ro-r--_--r--_...,..o,--_~_--,o_ _-.--r--,o (>J I N r Oo G) 0 rr, 0 0 C <l> 0 r rr, z -I ON 0 D - J u,L-_..l,.__ __,L__ _J..._ ____,__ _ _ ._ _. . _ _ _ ~ - ~ - - - - -74On first inspection these data indicate that all sequences of middle repetitive DNA are present in fraction S2; however, since the reaction proceeds over several decades of Ct, the families of 0 middle repetitive sequences are present at various levels of abundance in S2 DNA. From the fit of the data (Figure 10), 37% of the hybridiz- able tracer reassociates over the same range of Ct values as the 0 intermediate components of fraction S2 DNA (Ct 0.02 to Ct 10). 0 The second component (58%) exhibited a Ct 1/2 of 380. 0 0 This value is nearly the Ct 1/2 value for nonrepetitive DNA in the S2 driver 0 DNA ( C t 1/2 = 225, Table 2) . 0 We interpret these data to mean that 37% of the mass of rat middle repetitive DNA has complements in S2 DNA in high abundance (i.e., 100-2000 copies per haploid genome), while the rr.ajor fraction of repetitive sequences (58%) is represented only once or a few times in fraction S2 DNA .. DNA-RNA Hybridization We have shown previously (Gottesfeld et al., 1974a) that fraction S2 is enriched by a factor of 4 to 5 over fraction Pl in nonrepetitive sequences which code for cellular RNA. 125 This was shown by hybridiz~.ng r-labeled nonrepetitive DNA from liver chromatin fractions to an excess of total cellular RNA from liver. In RNA-driven hybridiza- tion experiments of this sort a simple saturation curve is obtained (Figure 4 of Gottesfeld et al., 1974a). The amounts of DNA found in DNA-RNA hybrids in our previous experiments are given in Table 4. Since the different tracer preparations used reassociated to Liver RNA Pl S2 Liver Chromatin Nonrepetitive DNAa 9.6 3.5e 14.5e RNA Saturation Observedb (% of DNA in DNA-RNA hybrid) 69.5 67.2e 52 .2e Final Reaction of DNA TracerC (%Double Stranded) 13.2 13,7 5. 2 27. 8 Corrected RNA Saturationd (% of DNA in DNA-RNA hybrid) DNA-RNA Hybridizat i on Experiments with Liver Chromatin Fractions Liver S2 67.2e Table 4 . Brain 8.9 aNonrepetitive DNA was isolated from the chromatin fracti ons S2 and Pl (5 min nuclease incubation) and Pl ~ labeled i n v i tro with 125 r (Commerford, 1971; Holmes and Bonner, 1974 ) . Brain t-; 4 bDNA tracers were mixed with whole cell RNA (see Experimental ) at a ratio of 2- 2. 5 x 10 :1 (RNA:DNA ) in 0 0.03 M PB, 0.675 M NaCl, denatured for 5 min at 110°c, and incubated at 74°c for 2-4 days. Rt values of 0 11 4 were obtai ned. Saturat ion is obtained after 1-2 days (Rt> 2.8 x 10 ). The amount of 5 .6-11 .2 x 10 4 DNA tracers were allowed t o reassociate in the presence of an excess of unlabeled 350 nucleotide-long labeled tracer in hybrid f orm was determined as described in Experimental section. C 0 unfractionated DNA, and the amount of DNA in hybrid assayed at high Ct (1 x 10 ). dCorrected saturation value is obtained by dividing the observed saturation value by the fraction of the Data of Gottesfeld et al. (1974). tracer DNA whi ch was able to f orm hybrids. e -76various extents when driven by whole rat DHA, we have corrected the RNA saturation values for the extent of hybridizable DNA in each preparation. Thus, fraction S2 is enriched in template-active sequences over fraction Pl by a factor of 5.4. Since only a minor fraction of the tracer DNA· ever forms hybrids with RNA, it is important to characterize this DNA . We have isolated the DNA from DNA-RNA hybrids by the method of Galau , Britten, and Davidson (1974). In this method , duplex-containing DNA is isolated first by hydroxyapatite chromatography, and then incubated wi th RNAas e. Any tracer DNA which formed DNA-DNA duplexes would still be doublestranded after this treatment; however, DNA from DNA-RNA hybrids becomes single stranded after RNAase treatment and thus separable from DNA-DNA duplexes by hydroxyapatite chromatography . The extent of DNA-DNA duplex formation during the hybr~dization experiment was found to be 2-3%. This is consistent with the value reported earlier (Gottesfeld et al., 1974a) obtained from a self-reaction of tracer in the presence of base-hydrolyzed RNA. The DNA isolated from DNA- RNA hybr::.ds was mixed with an excess of 350 nucleotide-long rat DNA and reassociation kinetics of the tracer determined (squares, Figure 8) . The reassociation curve of the tracer isolated from DNA-RNA hybrids is essentially identical to the reassociation profile of the tracer preparation prior to DNA-RNA hybridization. This shows that the DNA which formed hybrids with RNA was nonrepetitive DNA and not a minor repetitive contaminant of the tracer preparation . We do not know why the DNA which had been in duplex with RNA failed -77to reassociate to completion when driven with whole rat DNA, although this DNA is probably of small size after the many cycles of denaturation and incubation at high temperature to which it has been subjected . We have also used the total duplex containing DNA after DNA-RNA hybridization in a similar experiment (triangles, Figure 8). comprise 17% of this material (3%/3% + 14.5%). DNA-DNA duplexes When driven by whole rat DNA, the total duplex-containing DNA exhibited a significant degree of reassociation at Ct values prior to the reassociation 0 2 of nonrepetitive sequences (Ct< 10 ). These rapidly reacting sequences 0 comprise 21% of the hybridizable material in this preparation (14%/64% final reaction from a computer fit of the data). Thus the fast reacting material could correspond to the DNA which formed DNA-DNA duplexes during the RNA-driven reaction. Melting profiles of reassociated nonrepetitive tracer DNA from fraction S2 have been determined (data not shown). Tracer was allowed to anneal to either whole rat DNA or rat liver RNA to high Ct, and 0 unreassociated material was separated from duplex-containing material by hydroxyapatite chromatography. Melting curves were then determined on the material absorbed to hydroxyapatite. The T of unlabeled m reassociated rat DNA was 84°; the T of the reassociated iodinated m nonrepetitive tracer was 80-81°. The T m of the DNA which hybridized to RNA was similar (17% of this material is in the form of DNA-DNA duplexes). We do not know the reason for this decrease in T although m it might be attributed to decrease of tracer size due to the several cycles of denaturation and incubation necessary to prepare this material, -78or to the iodination procedure. We have also measured the extent of duplex formation of nonrepetitive tracer DNAs in the presence of an excess brain RNA (Table 4). The tracer DNAs were isolated from the liver chromatin fractions S2 and Pl. The percentage of tracer in DNA-RNA duplex (corrected for DNA-D!IA reassociation) was nearly the same for the two chromatin fractions. This indicates that fraction S2 DNA isolated from liver chromatin is not enriched over fraction Pl in sequences that code for brain RNA. This finding is consistent with the cross-reassociation experiment of Figure 7. The saturation values obtained with brain RNA a.re soBewhat higher than values reported in the literature for mouse brain RNA (Brown and Church, 1972; Grousse, Chilton, and McCarthy, 1972). If these latter values are corrected for the amount of hybridiz able DNA present in the tracer preparation (as we have done), then our values are identical to those in the literature. Discussion OveL the past decade several attempts have been made to fractionate interphase chromatin into transcriptionally active and inactive segments . The procedU!"es described to date start with sonicated or pressuresheared chromatin and fractionation is performed either by differential sedimentation (Frenster, Allfrey, and Mirsky, 1963; Murphy et al., 1973) or by chromatography (Mcconaughy and McCarthy, 1972; McCarthy et al., 1973; Simpson and Reeck, 1973). We have chosen to take another approach (Marushige and Bonner, 1971): chromatin DNA is digested -79with the endonuclease DNAase II under conditions of low enzymatic activity for brief periods of time. two ways: Fractionation is achieved in first, DNAase II appears to attack the transcriptionally active regions of chromatin more readily than transcriptionally inactive chromatin (Figure 2) . Second, inactive chromatin can be removed from solution by selective precipitation with either saline- cit r ate (Marushige and Bonner , 1971) or divalent cations (Billing and Bonner, 1972) . Using added radioactive polynucleotides, we find no detectable protein rearrangement during either enzyme treatment or subsequent fractionation (Gottesfeld et al., unpublished). We have repor ted previously (Gottesfeld et al., 1974a) that the active fraction S2 is enriched in nonhistone chromosomal proteins and depleted in histone protein. Histone I (fl) appears to be totally absent from the active fraction. This latter result is in agreement with the findings of Simpson and Reeck (1973) for active chromatin prepared by ECTHA.MchrOI!latog:aphy. Furthermore, nascent RNA appears to cofractionate with S2 chromatin (Bonner et al., 1975). We have also found that the circular dichroism spectrum of the active chromatin fraction is more like the spectrum of "B"-form DNA than the spectra of either unfractionated chromatin or the inactive fractions (Gottesfeld et al., 1974b). Similarly, Polacow and Simpson (1973) have reported that the active fraction obtained by ECTHAM-chromatography of sonicated chromatin is in an extended DNA-like conformation. Mcconaughy and McCarthy (1972) have fractionated pressure-sheared chromatin by thermal elution from hydroxyapatite. These authors -80have shown by RNA-excess hybridization that the early melting fractions are enriched in DNA sequences transcribed in vivo. however, has a major drawback: This procedure, thermal chromatography destroys native DNA-protein interactions and thus only the ective DNA is recovered by this method. An understanding of gene regulation in eukaryotes will require }-..nowledge about both the protein and DNA constituents of active chromatin. We have shown in this paper and previously (Gottesfeld et al., 1974a) that the nuclease-sensitive, Mg ++ -soluble fraction of liver chromatin is enriched fivefold in nonrepetitive DNA sequences which are transcribed in vivo. This was demonstrated by RNA-excess hybridization with labeled nonrepetitive DNA as tracer. It was also shown that the DNA which actually formed hybrids with RNA was derived from the nonrepeated portion of the rat genome {Figure 8). We have shown that the active fraction from liver chromatin is not enriched in single copy sequences complimentary to brain RNA . Similarly, cross-reassociation experiments with fraction S2 DNA from both liver and brain chromatin demonstrate that a low proportion of brain active sequences are present in the liver active fraction (Figure 7) . We believe, therefore, that our fractionation scheme results in the isolation of a tissue-specific transcribed portion of chromatin. We conclude that fractionation does not depend on some general property of chromatin but rather on the transcriptional state of the genome in the particular cell type under study . An estimate of the proportion of transcribed sequences in fraction S2 may be obtained from the RNA saturation data (Table 4). If all -81the nonrepetitive DNA of fraction S2 were transcribed, and transcription was from only one strand, then 50% of the hybridizable tracer should form duplexes with RNA. We found that 29% of the hybridizable tracer DNA from fraction S2 formed DNA-RNA hybrids. This suggests that fraction S2 is 58% l!pure" in transcribed nonrepetitive sequences. If the extent of nonrepetitive transcription in the liver is 3-5% (Grouse et al., 1972; Brown and Church, 1972), then we have purified t ranscri·b e d sequences by a fa.ctor of 6 -to 10 ( 29 to 29 ) over un f rac3 5 tionated chromatin. From cross-reassociation experiments with DNAs from liver chromatin fractions, we estimate that fraction S2 is contaminated with fraction Pl sequences to the extent of about 30-40%. This value is derived from the iegree of reassociation of a labeleu Pl-DNA tracer in the presence of an excess of unlabeled S2 DNA (F°igure 6). In the converse experiment, about 70% of a labeled S2-DNA tracer reassociated with an unlabeled Pl-DNA driver. We interpreted this latter result to mean that not all sequences of S2 DNA are removed from fraction Pl after short times of incubation with DNAase. Indeed, this can be seen in the data of Figure 2; after 5 min of incubation with the nuclease, only half of the S2 DNA is separated from fraction Pl. From the kinetics of reassociation of fraction S2 DNA (Figure 4b), we conclude that this fraction represents a specific subset of sequences of the rat genome. Knowing that the slow kinetic component of S2 DNA is derived from nonrepeated sequences (Figure 8), we calculate that this component comprises 11% of the total single copy complexity -82of the rat. The data of Figure 4b also suggest that fraction S2 contains a subset of repeated sequences. by the data of Figure 9: This notion is substantiated the complexity of 82 middle repetitive sequences is about 5-7% that of Pl middle repetitive sequences (Table 2) . The cross-reassociation experiment of Figure 9b shows that S2 sequences comprise approximately 11% of the complexity of Pl middle repetitive DNA; the kinetics indicate that 82 sequences are in nearly the same or slightly greater abundance as the majority of Pl middle repetitive sequences. Two alternative explanations for the presence of 82 sequences in substantial concentration in fraction Pl are possible: first, the fractionation procedure used (Figure 1) does not remove all S2 chromatin from fraction Pl after 5 min of nuclease treatment (Figure 2). Second, 82 repetitive sequences may be localized in both active and inactive regions of chromatin, and fractionation would never remove these sequences from Pl chromatin. Nevertheless, it is clea~ that fraction S2 contains at most one-tenth of the repetitive sequence complexity of fraction -Pl (Table 2). The experiment of Figure 9d allows us to estimate the purity of S2 repetitive DNA. One-third of the mass of Pl repetitive DNA can form duplexes with 82 repetitive DNA over the Ct values at which 0 82 DNA reassociates (Ct 0.05 to Ct 1). 0 0 The remaining Pl repetitive DNA reassociates at much higher Ct values, indicating that these 0 sequences are present in 82 DNA in extremely low concentration. We calculated earlier that 82 sequences make up 11% of the complexity of Pl repetitive DNA; the kinetics of Figure 9b suggest that S2 sequences -83might contribute more to the mass of Pl DNA than to the complexity. We calculate that S2 sequences could contribute 20% of the mass of Pl repetitive DNA. Thus an additional 10-20% of the Pl repetitive complexity is found in S2 DNA in significant abundance. Most Pl repetitive sequences are present in S2 repetitive DNA in very low concentration--0.2% of the concentration of S2 slow component sequences. This is probably an underestimate; the Ct 2 fractionation step used 0 in preparing S2 repetitive DNA could have reduced the concentration of Pl seq_uences in S2 DNA. We calculate that Pl repetitive sequences could be present in S2 DNA at 5~~ of the abundance of S2 repetitive sequences. In summary, we have shown that fraction S2 consists of a highly selected population of DNA sequences. The major fraction of the mass of S2 DNA (ca. 70-90%) comprises no more than 10-12% of the complexity of the rat genome. Preliminary data suggest that repetitive and single copy sequenc e s are interspersed to a high degree in fraction S2 DNA (Gottesfeld and Bonner, unpublished). This notion comes from an experiment of the type first performed by Davidson et al. (1973). Labeled 1800- nucleotide long S2 DNA (30 sec enzyme treatment) was allowed to reassociate in the presence of an excess of 350-nucleotide long whole rat DNA; after incubation to Ct 10 and hydroxyapatite chromatography, 0 70% of the hybridizable tracer was found in the 0.48 M PB fraction. In comparison, when labeled 400-nucleotide long S2 DNA was incubated .rith whole DNA to the same Ct, only 30% of the DNA was in the hybrid 0 fraction. This suggests that both repetitive and single copy sequences -84are present on the majority of 1800-nucleotide long S2 DNA molecules. This result is not surprising since short period interspersion of repetitive and single copy sequences is a phenomenon common to a wide range of organisms (Davidson et al., 1975a) including rat (Bonner et al., 1973). Furthermore, Davidson et al. (1975b) have found that transcribed nonrepeated sequences in the sea urchin tend to be adjacent to repetitive sequences. Our finding of limited repetitive and single copy complexities in fraction 82 DNA suggests that the distribution of repetitive and nonrepetitive sequences in the genome is nonrandom . Furthernore, fraction 82 is enriched in transcribed single copy sequences and nonrepetitive sequences are probably interspersed with repetitive sequences in fraction 82. This suggests that specific repetitive sequences might lie adjacent to transcribed sequences. At present, we do not know to what extent the repetitive sequences of fraction S2 are transcribed. It is likely that fraction 82 codes primarily for nuclear transcripts (hnRNA). This class of RNA represents about 10% of the total complexity of the rat genome in ascites cells (Holmes and Bonner, 1974a). Messenger RNA complexities tend to be about one-tenth that of hnRNA (Galau et al., 1974; Davidson and Britten, 1973; Hough et al., 1975; Holmes and Bonner, personal communication) . Thus, about one-tenth of fraction S2 DNA probably codes for translated :polysomal messenger RNA, while the bulk of the S2 DNA codes · for untranslated nuclear RNA. Holmes and. Bonner (197lib) and Smith et al. (1974) have shown that many repetitive and nonrepetitive sequence transcripts are on the same hnRNA -85molecules; that is, hnRNA is transcribed from interspersed repetitive and single copy sequences. Messenger RNA, however, is primarily the transcript of single copy sequences (Davidson and Britten, 1973). 7hus, the sequence organization of S2 DNA reflects the sequence organization of hnRNA. Britten and Davidson (1971) have postulated that the control of transcription in eukaryotes is mediated through the binding of activator molecules to sites on repetitive DNA, promoting the transcription of neighboring sequences. This model and its consequences have been reviewed recently by Davidson and Britten (1973). The model vould predict that, associated with transcribed single copy sequences, one should find a limited complexity of repetitive sequences. The results reported herein are consistent with the predictions of the Britten-Davidson model; however, we hasten to point out that these results should not be taken, at this time, as a confirmation of the Britten-Davidson model. Experi:nental Preparation and Fractionation of Chromatin Chromatin from either frozen rat liver (Pel-Freeze, Rogers, Ark.) or rat Novikoff ascites cells was prepared by the method of Marushige and Bonner (1966). Chromatin from frozen rat brain (Pel- Freeze) was prepared by the same method except the crude chromatin was purified by two cycles of sedimentation through 1 M sucrose-10 mM -86Tris-Cl (pH 8) prior to centrifugation through 1.7 M sucrose. The pellets obtained after ultracentrifugation were washed once in 10 mM Tris-Cl (pH 8) and then dialyzed overnight against at least 200 vol of 25 mM sodium acetate ( pH 6. 6) at 4 °C. Fra.ctionation was per- formed as described previously (Gottesfeld et al., 1974a); a brief description of the method is found in the Results section. Rat DNA Preparations Total Rat DNA. DNA was prepared fron either crude rat liver or rat ascites chromatin by standard techniques (.Marmur, 1961). In the case of ascites DNA, the chromatin was prepared from cells which had bee::, labeled in vivo with 3tt-thymidine. of Dr. J. S. Sevall. This DNA was a gift DNA was sheared by two passes through the Ribi- Sorvall :i;ressure cell at 50,000 p. s. i. DNA· was sheared in moderate salt (either 0.12 M PB or 0.1 M NaCl) to avoid thermal denaturation during this treatment. After shearing, the DNA was extracted once with phenol (saturated with 10 mM Tris-Cl, pH 8) and then with chloroformisoamyl alcohol (24:1, v/v). DNA fron Chromatin Fractions. DNA from chromatin fraction S2 was prepared in a much gentler manner than described above for total rat DNA. Im,~ediately after fractionation the chromatin was carefully pipetted into a large screw-top test tube containing an equal volume of Tris-saturated phenol. The tube was then laid on its side on the lab bench and incubated at room temperature for 20-30 min. The -87tube was then gently turned upright such that no mixing of the phenol and aqueous phases occurred. The aqueous phase was removed with a pipette and transferred to another tube above an equal volume of fresh Tris-saturated phenol. This process was repeated five more times with phenol and then 2-3 times with chloroform-isoamyl alcohol (24:1, v/v). The final aqueous phase was dialyzed against 10 mM Tris-Cl, 10 mM EDTA, pH 8. Ribonuclease (preincubated at 80°C for 10 min) was added to 25 µg/ml and the mixture incubated at 37°C for 1 hr. NaCl was added to 0.5 Mand sodium dodecyl sulfate to 1%. Pronase (autodigested for 1 hr at 37°C) was added to 25 µg/ml and the mixture was incubated at 6o 0 c for 1-2 hr. Phenol and chloroform extractions were repeated as described above. The final aqueous phase was dialyzed against 10 rnM ammonium acetate and then examined by electron microscopy by the aqueous ammonium acetate method of Davis, Simon, and Davidson ( 1971). The bulk of the DNA from any given preparation was not used for electron microscopy and was lyophilized. DNA from fraction S2 was purified further by chromatography on Sephadex G-200 in either 0.1 N NaOH or 50 mM Na Po 4 (Britten et 3 al., 1974). The excluded fractions were pooled, neutralized, dialyzed against ammonium acetate and then lyophilized. In an early experiment it was found that without G-200 chromatography fraction S2 DNA (5 min enzyme treatment) reassociated to only 70%. It was found that approximately one-third of the DNA applied to a G-200 column (as measured by A ) was included and hence was less than 100-150 260 nucleotides in length. -88DNA from chromatin fractions Pl and P2 was sheared in the RibiSorvall pressure cell as described above for total rat DNA . Middle Repetitive DNA. The middle repetitive component of either rat liver or 3H-ascites DNA was prepared in the following manner: DNA was dialyzed against ammonium acetate, lyophilized and redissolved in an appropriate volume of 0.12 M PB. The DNA solution was placed in a Reacti-Vial, denatured in a boiling water bath for 5 min, and incubated to a Ct of 0.1. 0 Nonreassociated DNA was separated from duplex-containing DNA on hydroxyapatite (Britten and Kohne, 1968). The single stranded fraction was dialyzed against ammonium acetate, lyophilized and redissolved in 0.12 M PB. The DNA was incubated to a Ct of 50, and double stranded DNA was separated by hydroxyapatite 0 . chromatography. The final duplex-containing material was termed middle repetitive DNA. After this procedure, the single strand length of the DNA was determined to be 300 nucleotides (determined by alkaline sedimentation velocity, see below). The specific activity of the ascites middle repetitive DNA was 10,000 cpm/µg. repetitive DNA was labeled in vitro with 125 Rat liver middle 1 by the Commerford method (1971) as modified by our laboratory (Holmes and Bonner, 1974). cific activities of 1-10 x 10 Spe- 6 cpm/µg were obtained. Middle repetitive DNA of fractions Pl and S2 were prepared in a man..'1er similar to that described for whole rat DNA except different Ct values were used. 0 Highly repetitive DNA was removed from sheared Pl DNA after incubation to Ct 0.05. 0 The single strand fraction -89from a hydroxyapatite column was then allowed to reassociate to Ct 0 20 and the double strand fraction was collected. Highly repetitive DNA was removed from S2 DNA after incubation to Ct 0.02, and middle 0 repetitive DNA was isolated after incubation to Ct 2. 0 These middle repetitive DNAs were labeled in vitro by nick translation with DNA polymerase I. The following methods were suggested to us by Mr. Glenn Galau and Dr. Francine Eden. Prior to labeling, Pl middle repetitive DNA was incubated to Ct 55 (without denaturation before 0 incubation)~ S2 middle repetitive DNA was incubated to Ct 12. These 0 Ct values are approximately 250-300 times the Ct 1/2 values calculated 0 0 for pure niddle repetitive DNAs. After incubation, the DNAs were dialyzed at 4° against 4 £ of 50 mM PB, 1 mM EDTA, pH 7.55 (Schachat and Rogness, 1973). For each sample to be labeled, 0.4 mCi of 3tt- TTP (45 Ci/.cmole, Nuclear Dynamics) plus lQ µmoles of each of dATP, dCTP and dGTP were lyophilized in a conical centrifuge tube. To this, 2-5 µg of DNA (in 0.25 ml of 50 mM PB, 10 mM MgC1 , 1 ml'! EDTA, 2 pH 7.55) was added. The reaction was started by the addition of 5 units of DNA polymerase I (Boehringer-Mannheim) per ~g of DNA. After i~cubation at 12° for 22-27 hr, the reaction was stopped by the addition of EDTA (pH 8) to 50 mM and chilling on ice. Approxi- mately 3 ml of 0 . 03 M PB-0.1 M NaCl was added; the sample was denatured at 97° (3-5 min) and applied directly to a 1 ml hydroxyapatite column equilibrated in 0.03 M PB-0.1 M NaCl. The column was washed with 0.03 M PB-0.1 M NaCl until the effluent contained less than 1000 -90cpm per 20 µl (when counted in 5 ml of Aquasol). then eluted with 0.12 Mand 0.48 M PB. The column was About 30% of the radioactivity incorporated into DNA eluted in the 0 . 48 M PB fraction; this material 6 was discarded. Specific activities of 0 . 5 to 3,5 x 10 were obtained . Labeled middle repetitive DNA of fraction S2 was cpm/µg DNA further purified by chromatography on Sephadex G-200 in 0.05 M Na Po . 3 4 The excluded fractions were pooled, dialyzed against 0.01 M ammonium acetate and lyophilized. Nonreuetitive DNA. Chromatin from rat liver was treated with DNAase for 5 min and fractionated as described above. from fractions S2 and Pl. above. The Pl-DNA was pressure sheared as described The DNA was denatured and incubated in 0 . 48 M PB to an equivalent Ct of 2.5 x 10 0 DNA was prepared 2 (S2-DNA) or 1.5 x 10 3 (Pl-DNA), and the single stranded fraction was isolated by hydroxyapatite chromatography. This material was dialyzed, lyophilized, dissolved in 0.48 M PB, and allowed to renature as before. The DNA that remained single stranded after two cycles of purification was allcwed to renature to a Ct of 10 0 4. The res-ulting duplex-containing DNA was isolated by hydroxyapatite chromatography, dialyzed and concentrated by lyophilization. The· purified nonrepetitive DNA was labeled in vitro by the modified Commerford method. Specific activities of 1-10 x 10 6 cpm/µg were obtained . Further Treatment of Tracer Preparations. In order to remove nonhybridizable material, some tracer preparations (labeled with 125 1) were purified on Sephadex G-200. In some cases foldback or -91zero-time binding DNA was stripped from the ations. 125 I-labeled DNA prepar- This was accomplished by denaturing DNA in 0.03 M PB and pipetting the denatured DNA directly onto a hydroxyapatite column equilibrated in 0.03 M PB-0.1 M NaCl at 6o 0 c. The time for absorption onto the resin (1-2 min) was such that equivalent ·c t values of no 0 greater than 10 -4 were obtained. The colu..>n.n was washed with 0.03 M PB-0.1 M NaCl, and the single stranded DNA was eluted with 0,12 M PB. DNA Reassociation Kinetics Kinetics of DNA reassociation were monitored by estimation of duplex formation via hydroxyapatite chromatography (Britten and Kohne, 1968). DUA preparations dissolved in PB were sealed in glass capillary tubes, denatured in a boiling water-ethylene glycol bath (1:1, v/v), and allowed to incubate for various times. Incubations were carried out at 24°c below the T for rat DNA at the particular sodium ion m concentration used. The T was estimated by the equation of Mandel and Ma_~ur ( 19 68 ) . For 0 , 72 M Na+ (0.48 M PB) the T m m of rat DNA is 96°c and incubations were carried out at 72°C; for 0.18 M Na+ (0.12 M PB) the T m 1-1 Na + is 86°c and incubations were at 62°C; for 0,075 (0.05 M PB) the T m is 8o 0 c and incubations were at 56°c. All Ct values reported herein have been normalized to Ct values equivao 0 lent to those obtained under standard conditions (0.12 M PB, 62°C). Normalization of the data was performed according to Britten et al. (1974). After incubation, the contents of the capillary tubes were -92diluted into 0.03 M PB, 0.1 M NaCl, and the solution was applied to a hydroxyapatite column equilibrated in the same buffer at 6o 0 c. Approximately 1 ml of packed DNA-grade hydroxyapatite (Bio-Rad BioGel HTP) was used for each 100 µg of· DNA applied to the column. The column was washed with 4-6 column volumes of 0 . 03 M PB, 0.1 M NaCl; single strand DNA was eluted with 0.12 M PB and duplex-containing DNA was eluted with o.48 M PB. An extinction coefficient of 6600 was used for double stranded DNA and 8850 for single stranded DNA. DNA-Excess Reactions Reassociation experiments where unlabeled DNA was used in excess over a ~abeled DNA tracer were carried out in a manner similar to that desc~ibed above. Labeled tracer and driver DNAs were incubated in 0.03 ~ PB plus NaCl to the desired final sodium ion concentration. HydroX'Japatite chromatography was performed as described above. the case of 125 In 1-labeled DNA tracers, iodine is lost from the DNA during thermal denaturation and incubation at high temperatures. The free iodin.e in the reaction mixture is not absorbed to hydroxyapatite, and must be washed from the column with 0.03 M PB-0.1 M N~Cl. In most DUA-excess experiments, 1-ml hydroxyapatite columns were used, and two 8- to 10-ml fractions were collected (0.12 M PB and 0. 48 M PB). The fractions were chilled on ice; 50 µg of bovine serum albumin was added as a carrier, and nucleic acids were precipitated with 7% c1 ccOOH for 10-20 min. 3 The precipitates were collected on filters and the filters were washed with cold 7% c1 ccooH followed 3 -93by 60% ethanol. The filters were dried in a vacuum oven at 6o 0 c and counted in a toluene-based scintillant. RNA-Driven DHA-RNA Hybridization Experiments Total cellular RNA from rat liver and brain (Pel-Freeze frozen tissue) was isolated by a modified hot phenol-sodium dodecyl sulfate extraction procedure (Moriyama et al. , 1969) . The RNA was sheared to a weight-average length of 1000 nucleotides by two passes through the Ribi-Sorvall pressure cell at 30,000 p.s.i. The RNA size was determined under nondenaturing conditions in the analytical ultracentrifuge. The Spirin equation was used to relate s weight of RNA (Spirin, 1963). 20 , w to molecular Mixtures of sheared RNA and nonrepetitive DNA were heat denatured and incubated in 0.03 M PB, 0,675 M NaCl, 1 nL\1 ED~A, pH 6.5, at 74°c. The ratio of RUA to 125 1-labeled DNA 4 was 2-2.5 x 10 :1, and the RNA concentration was 20 mg/ml. were carried out in sealed capillary tubes. Reactions Hybridization reactions were terminated by diluting the samples 20-fold in 0.03 M PB, 0.1 M NaCl at 60°C followed by application to a hydroxyapatite column equilibrated in the same buffer at 60°C. apatite .,,-as used per 100 µg of RNA. About 1 ml of packed hydroxy- At 0.03 M PB, 0.1 M NaCl, all nucleic acids are absorbed to the resin. Single stranded DNA and the bulk of the RNA is eluted in 0.12 M PB; DNA-RNA hybrids, DNADNA duplexes, and about 20% of the input RNA is eluted in 0.48 M PB. The samples were precipitated and counted as described above. The extent of DNA-DNA self-reaction was estimated in two ways. Ten mg of liver RNA was hydrolyzed in 300 µl of 1.125 N NaOH for 3 hr at 37°C. At the end of this time 200 µl of 1.69 N HCl, 0.075 M PB was added to give a final NaCl concentration of 0.675 Mand PB concentration of 0.03 M. A small a.mount of 125 ~-labeled nonrepetitive DNA (in distilled water) was added, and the mixture was lyophilized. The dried material was taken up in 500 µl of distilled water; 50 µl aliquots were sealed in capillary tubes, denatured and incubated at 74°c. above. The a.mount of DNA in duplex form was estimated as described DNA-DNA duplex formation at zero-time was approximately 2%; this amount did not increase during a 96 hr incubation. The second method of estimation of DNA-DNA duplex formation during RNA-driven hybridize.tio!l was the method of Galau et al. ( 1974). DNA-RNA hybridiza- tion in R?TA excess was carried out as described above. The O. 48 M PB fra~ticn from a hydroxyapatite column was dialyzed against 6 t of 0.05 M PB; RNAase A (preincubated at 80°C for 10 min) was added to 10 \.lg/"11.. The mixture was incubated overnight at 37°C and a second hydroxyapatite column was run. The material which elutes in 0.12 M PB is the DNA which was in DNA-RNA hybrid, and the DNA in the 0.48 M PB fraction is in DNA-DNA duplex. By this method, DNA-DNA reassoc- iation was estimated to be 2-3% at high Rt values. 0 For the isolation of la.Yger a.mo'1.Ilts of DNA from DNA-RNA hybrids, 2 µg of 125 1-labeled S2 nonrepetitive DNA was mixed with o.4 mg of liver RNA and hybridization was allowed to proceed for 1.8 hr (Rt= 23,600). 0 This corresponds to a tracer Ct of 118; 5% of the input DNA was in DNA-DNA duplex 0 -95and 13% of the input DNA was in DNA-RNA hybrid. Estimation of DNA Size DNA size was determined in any one of three ways. Double stranded DNA was visualized in the electron microscope after spreading from aqueous ammonium acetate by the method of Davis et al. (1971). ¢x174 DNA was used as a standard. Phage Single and double stranded lengths were determined by sedimentation velocity in the analytical ultracentrifuge by the methods of Studier (1965), Weight-average single strand lengths were also determined from isokinetic sucrose gradients under alkaline conditions. The gradients were formed according to Noll (1967) with the following parameters: 38.9% (w/v); V . = 6.1 ml. mix CTOP = 15.9% (w/v); CRES = Gradients were centrifuged in the SW 50.1 rotor at 48,ooo rpm for 16 hr at 20°C. • A marker of sheared calf thymus DNA was the generous gift of Ms. Maggie Chamberlin. DNA had a number-average length of 320 nucleotides, as judged by electron microscopy, and an observed sedimentation value of 5.4s under alkaline conditions. This Acknowledgments We wish to thank Mr. Mahlon Wilkes for performing the electron microscopy for us, and Mr. Ralph Wilson for technical assistance during the early portion of this work. We are very grateful to Drs . Roy Britten, David Holmes, Barbara Hough and William Klein, and to Mr. Glenn Galau for numerous discussions and evaluation of this work . These studies were supported by the United States Public Health Service (Grants GM 86 and GM 13762). -97References Billing, R. J., and Bonner, J. (1972). Biochm. Biophys. Acta 281, 453-462. Bonner, J., Garrard, W. T., Gottesfeld, J., Holmes, D. S., Sevall, J. S. , and Wilkes , M. ( 1973) . Cold Spring Harbor Symp·. Quant. Biol. 38, 303-310. Bonner, J., Gottesfeld, J.M., Garrard, W. T., Billing, R. J., and Uphouse, L. (1975). In Methods in Enzymology, 40, s. P. Colowick and N. O. Kaplan, eds. (New York: Academic Press), in press. Britten, R. J., and Davidson, E. H. (1971). Quart. Rev. Biol. 46, 111-138. Britten, R. J., Graham, D. E., and Neufeld, B. R. (1974). In Methods in Enzymology, 39, S. P. Colowick and N. O. Kaplan, eds. {New York: Academic Press), pp. 363-418. Britten, R. J., and Kohne, D. E. (1968). Science 161, 529-540. Brown, I. R., and Church, R. B. (1972). Develop. Biol. 29, 73-84. Cairns, J. (1963). Cold Spring Harbor Syrop. Quant. Biol. 28, 43-46. Commerford, S. L. (1971). Biochemistry 10, 1993-2000. Davidson, E. H., and Britten, R. J. (19T3). Quart. Rev. Biol. 48, 565-613 . Davidson, E. H., Hough, B. R., Amenson, C. S., and Britten, R. J. (1973) J. Mol. Biol. 77, 1-23. Davidson, E. H., Galau, G. A., Angerer, R., and Britten, R. J. (1975a), in press. Davidson, E. H., Hough, B. R., Klein, W. H., and Britten, R. J. (1975b). Cell, in press. -98Davis, R. W. , Simon, M. , and Davidson, N. ( 1971). In Methods in Enzymology, 21, part D, S. P. Colowick and N. O. Kaplan, eds. (New York: Academic Press), pp. 413-428. Frenster, J. H., Allfrey, V. G., and Mirsky, A. E. (1963). Proc. Nat. Acad. Sci. USA ?0, 1026-1032. Galau, G. A., Britten, R. J., and Davidson, E. H. (1974). Cell _g_, 9-20. Gottesfeld, J. H. , Garrard, W. T. , Bagi, G. , Wilson, R. F. , and Bonner, J. (1974a). Proc. Nat. Acad. Sci. USA 71, 2193-2197. Gottesfeld, J.M., Bonner, J., Radda, G., Walker, I. O. (1974b). Biochemistry 13, 2937-2945. Grouse, L. , Chilton, H. D. , and McCarthy, B. J. ( 1972) . Biochemistry 11, 798-805. Holmes, D. s., and Bonner, J. (1973). Biochemistry 12, 2330-2338. Holmes, D. s.' and Bonner, J. ( 1974a). Biocnemistry 13, 841-848. Holmes, D. s. , a."1d Bonner, J. (1974b). Proc. Nat. Acad. Sci. USA 71, 1108-1112. Hough, B., Smith, M. J., Britten, R., and Davidson, E. (1975). Cell, in press. Mandel, :M . , and Marmur, J. ( 1968) . In Methods in Enzymology, 12, part B, s. P. Colo;;ick and N. O. Kaplan, eds. (New York: Academic Press), pp. 195-206. Marmur, J. ( 1961) . J. Mal. Biol. l, 208-218. Marushige, K., and Bonner, J. (1966). J. Mol. Biol. 15, 160-174. Marushige, K., and Bonner, J. (1971). Proc. Nat. Acad. Sci. USA 68, 2941-2944. -99McCarthy, B. J., Nishiura, J. T., Doenecke, D., Nasser, D., and Johnson , C. B. (1973). Cold Spring Harbor Symp. Quant. Biol. 38, 763-772 . Mcconaughy, B. L. , and McCarthy, B. J. ( 1972) . Biochemistry 11, 998-1003. Moriyama, Y., Hodnett, J. L., Prestuyko, A. W., and Busch, H. (1969). J. Hol. Biol. 39, 335-349. Murphy, E. C., Jr., Hall, S. H., Shepherd, J. H., and Weiser, R. S. (1973). Biochemistry 12, 3843-3852. Noll, H. (1967). Nature 215, 360-363. Polacow , I., and Simpson, R. T. (1973) . Biochem. Biophys. Res. Commun. 52, 202-207. Schachat, F. H., and Rogness, D. S. (1973). Cold Spring Harbor Symp. Quant. Biol. 38, 371-381. Simpso~, R. T., and Reeck, G. R. (1973). Biochemistry 12, 3853-3858. Smith, >~. J., Hough, B. R., Chamberlin, M. ~., and Davidson, E. H. ( 197 4). J. I.fol. Biol. 85, 103-126. Sober, H. A. (1968). Handbook of Biochemistry (Cleveland, Ohio: Chemical Rubber Co.), pp. H-58. Spirin, A. S. (1963). Prog. Nuc. Acid Res . .!_, 301-345. Studier, F. W. ( 1965) . J. Mol. Biol. 11, 373-390. Wetmur, J. B. , and Davidson, N. ( 1968) . J. Mol. Biol. 31, 349-370. Young, E.T., II, and Sinsheimer, R. L. (1965). J. Biol. Chem. 240, 1274-1280. -100- V. STRUCTURE OF TRANSCRIPTIONALLY-ACTIVE CHROMATIN To be submitted for publication in the Proceedings of the National Academy of Sciences. -101- ABS1'RAC~' Rat-liver chromatin has been fractionated into transcrip- tionally-active and inactive regions [Gottesfeld et al. (1974) Proc. Nat. Acad. Sci.. USA '/1, 2193-2197) and the distribution of nuclease-resistant complexes in these fractions }ms been investigated. About half of the DNA of both fractions is resi stant to attach by the endonuclease DNas e II. The nuclease-resistant struct,.u-es of inact ive chromatin are DNA-histone complexes (v-bodies ) which sediment at 11-138 ; species a r e fom1d in active chromatin. Two nuclease-resistant The se particles sediment at 15 and 20S, respE,ctively, and contain DNA, RNA, histone and nonhistone c:hromosoma.l proteins. Thermal denaturation studies suggest that the DNA of the act:ive fraction is complex0d with nonhistone protein. -102- A regular repeating unit in chromatin was first suggested from the X-ray diffraction studies of Pardon, Richards and Wilkins (1): A series of reflections were observed in the X-ray patterns of native and reconsti tuted nucleohistoncs, but not in the X-ray diffraction patterns of DNA or histon e s by themselves. It was proposed that the chromatin fiber 0 is organized int o a regular supercoil of pitc h 100-120 A. This model, although wi<lely accepted for some time, has now come under question. 0lins and 0lins (2) have observed regular spac ings of chromatin particles (termed v-bodies) in water-swollen nucl ei centrifuged onto electron 0 microscope grids. These particles are 60-80 A in diameter and are joired by thin fil aments 15 A in diameter. These results have been confirmed and extended by other laboratories (3- 5 ). Nuclease digestion studies also support a subunit or part iculate structure for chromatin: and Burgoyne (6) and Burgoyne~.!. al. Hewish (7) have found that an endogenous Ca ++ - Mg++ -activate d endonuclease of rat and mouse liver nuclei cleaves chromatin DNA at regular intervals; fragments of DNA at integral multiples of about 200 base pairs were observed. Many workers have now studied the di gestion of isolated chromatin with exogenous nucleases (8-14), and the results are gene r ally cons istent with the subunit or "beads-on-astring" mode l of chromatin structure (2,15). Furthermore, isolated particles fr o;a nuclease-treated or sonic ated chromatin resemble v-bodies in the electron microscope (1 6-18). Recent evidence obtained from neutron scattering of chromatin (19) 0 suggests that the 110 A reflection seen in the X-ray pattern does not ari s e fr om a DNA repeat, but rather from a regular spacing of a protein core. Baldwin et al. (19) and van Holde et al. (20) propose that chromatin DNA is wound about the exterior of the protein core. Noll's -103- finding of JJuclease-sensitive sites every 10 base pairs along the DNA in chrorn:-J .ti n appears to support this notion (21). '!.'be work on chromatin structure cited above has been carried out • with whole dJJ:oruatin, w1fractionatcd with respect to transcriptional activity. Since only a minor portion of the DNA in any given cell type is transcriptionally active, the properties of unfractionated chromatin primarily reflect the structure of inactive regions. We are interested in whether the template-active fraction of chromatin is organized in a structure similar to inactive regions or whether it is in a different conformation. Previous work from this laboratory has shown that it is possib.1 e to separate chromatin into active and inactive fractions (22-24). Transcriptiom1.lly-act:i. ve chromatin appears to be preferentially attacked by the endonuclease DNase II; separation of the chromatin fractions is accomplished by centrifugation and selective precipitation of inactive chromat in with MgC1 . 2 The active fraction diff2rs from whole chromatin or inact:i.ve chromatin in that it po ssesses a decreased histone to DNA ratio, lacks histonc I, and has an j_ncreased nonhistone to DNA ratio (24). The active fractian comprises a subset of both single copy and repetitii.e DNA sequences and is enriched five-fold in sequences complementary to cellulftr RNA (24). In this communication we report that both active and inactive chromgtin fractions contain nuclease-resistant particles; however, the nuclease-resistant structures of inactive chromatin are DNAhistone complexes while the nucl ease-resistant particles of active chromatin are DNA-RNA-protein complexes made up largely of nonhistone chrornosom'.11 protE>ir.s. ... -104- Ml•~I'HODS Chrornati!1 fractionation. Chromatin was prepared from rat liver by the method of Marush:i ge ~-nd Bonner ( 2'..i). The chromatin pellet aft er sucrose gradient centrifugation was washed once with 10 mM Tris-HCl (pH 8) and dialyzed overnight at 4° against 200 vo1umes of 25 mM sodium acetate (pH 6. 6) . Fractionation was carried out as diagrrunmed in Fig. l; details of this method have been published previously (24). Preparation of chromatin subunit s. Nuclease-resistant particles from rat-liver chromatin were pr epared as follows: DNase II was added to 10 units per A unit of chromatin in 25 mM sodium acetate (pH 6.6). 260 Digestion was carried out at 24° and was terminated after 90 min by raising the pH to 7.5 with 0.1 M Tris-HCl (pH 11). Nuclease-resistant subunits from 'chromatin fraction Pl were prepared by homogenizing the pellet fraction in 25 mM sodium acetate (pH 6.6) and redigesting with DNase as described above for whole chromatin. Undigested chromatin (about 20% of the input DNA) was removed by centrifugation at 21,000 g for 10-15 min. The supernatant wa s layered on isokinetic sucrose gradients in S\·125 . 1 cellulose nitrate tubes. The gradients were formed according to Noll (26 ); the parameters were CTOP = 15% (w/v), CRES = 34.2% (w/v), and VMIX = 31 .4 ml. (pH 8). All solutions contained 10 mM Tris-HCl Centrifugation was at 25,000 rpm for 36-42 h. Gradients were analyzed with an ISCO UV Analyzer and chart recorder. Fractions from these gradients were rerun on a s econd isokinetic sucrose gradient. Sucrose was removed from the fractions by dialysis against 10 mM Tris-HCl (pH 8) and the sample was concentrated with an Arnicon Minicon device prior to l ayeri ng the sai!lple on the gradient. The parameters for the -105- second isokin e tic sucrose gradient were C = 5.1% (w/v), CRES = TOP 31. 4% w/v, and VMIX = 9. 4 ml. All solutions contained 10 mM Tris-HCl (pH 8). Centrifugation was in the SW41 rotor at 39,000 rpm at 4° for 16.5 h. The gr a dients were analyzed spect rophot ometrically as described above. Subunits we re ri.lso prepared from chromatin devoid of hi stone I. Removal of this histone wa s accomplished by extraction of Virtis-sheared chromatin (45 v, 90 sec) with 0.5 M NaCl at 4°. The resultant nucleo- histone was pelleted by centrifugation in the Ti 50 rotor at 50,000 rpm for 18 h. The pellet was homogeni zed in 10 mM Tris-HCl (pH 8) and treated in the same manner as whole or Pl chromatin. Bed~ gest ion of Chromatin Fraction S2 . Chromatin of fraction S2 was digested 1;1:ith nuclease in three different ways: the DNase II present in fracti on 8 2 was reactivated by adding EDTA to 20 mM and lowering the pl! to 6.11 with dilute HCl. Alterna tively, aliquots of chromatin fract:ion 82 were dialyzed against either 25 mM sodium acetate (pH 6.6) or 5 mM sodium phosphate (pH 6.7) containing 2.5 x 10 MgC1 . 2 -4 M CnC1 2 and 2.5 x 10 -4 M DNase II was added to the chromatin in sodium acetate buffer to 10 uni ts p,:r A unit of chromatin ; staphylococcal nuclease was added 260 to the chromat i n in sodium phosphat e buffer to 50 units P:r ml. actions wer<2 c?.rried out at 24°. Re- Aliquots were taken at various times to test fm: the production of tri chloroacetic acid-soluble material (measured by absorbance of the supernatant at 260 nm after centrifugat:ion at 27,000 g at 4° :for 15 min). Sucrose Gradient Analysis of F'raction S2 Olromatin. prepared and fractionated as described above. Chromatin was The active fraction S2 ;,as concentnd,ed on a,n Ami.con Minicon device, l _ayered on a 5-24% isokinetic -106- sucrose grad:i.ent, and centrifuged in the 8W41 rotor at 39,000 rpm for 17.5 hat 4°. '!'he parameters for the gradients were CTOP = 5.1% {w/v), CRES = 31.4% (w/v), arid VI-llX:: 9.'1 ml (26). All solutions contained 10 mM Tri s- HCl {pH 8). DNA Size Estimation. Single-strand DNA lengths were estimated by velocity sedimentation in alkaline sucrose gradients. The parameters for the isokin etic gradients were CTOP = 15,9% {w/v), CRES = 38.9% (w/v), and VMIX = 6.1 ml (26). All solutions contained 0.1 N NaOH. Chromatin samples were suspended in 0.1 N NaOH, 2% sod:i.um dodecyl sulfate, 2 M urea, and 100-200 µl aliquots were layered on each gradient. Centri- fugati.on was in the SW50 .1 rotor at 48,000 rpm for 16 h at 20°. DNA molecular wci r,hts were det ermined relative to a standa rd sized by electron microscopy ( 320 nucleotide-long, 5.4S, calf thymus DNA; a gift of Ms. M. Chamberlin). Double-strand lengths were determined in the analytical ultracentrifuge ( 27) . Optic a l 'l'hcrmlJ.1 Denaturation Profile s were obtained with a Gilford 2400 Spectrophotometer, A.ll s&.mples were dialyzed against 0.2 mM EDTA (pH 8) prior to melting. Analys i~; of Chromatin Composition. content wa s det ermined as described (28). Histone and nonhistone protein Protein was analyzed by sod:inm dodecyl sulfat e-disc gel elec:trophoresis (29) and by acid-urea gel electrophoresis ( 30) . DNA and RNA were determined by the methods of Schmidt and Tannhauser ( 31). -107- Enzyme ~. DNase II (E.C. 3.1.4.6) and micrococcal nuclease (E.C. 3.1.4.7) from staphylococ cus were purchased from the Worthing ton Biochemical Corp . RESULTS ChromaUn Fractionation. DNase II preferentially attacks a minor portion of chromatin DNA; the: proportion of DNA in this fraction varies depending upon the source of the chromatin, but corresponds quite closely to the measured template activity of the particular chromatin (2 3). fractionation sc heme used herein is diagramme d in Fig. 1. The After 5 min exposure to DNase II, 15% of rat-liver chromatin DNA remains soluble after centrifugat ion (fracti on Sl). Mg ++ About 11% of the total DNA is - soluble and is found iri fraction S2. This DNA comprises a subset of whole genornal DHA sequenc es and is enriched rive-fold in transcripti onally-active sequences (24). The DNA has a double-strand length of about 700 bnse pairs and a single-strand l ength of 200-600 nucleotides (range of observed values ). About 1-3% of this DNA i s acid soluble. After 30 min exposure to DNR.se II, nearly 80% of the chromatin is found i n fr action Sl , and 20-24 % is found in fraction S2. The !32 DNA averages 50 nuc leotides, but half of this DNA is acid soluble. A more detailed descripti on of the kineti cs of DNase II digestion will appear elsewhere (Gottesfeld .s'...t al., in preparation). Table I lists some of the properties of the chromatin fractions: the compos iti ou of fraction Pl is similar to that of unfractionated ratliver chromatin (24-25). Fraction S2, however, is enriched in RNA and nonhistone prote in and has a r educed content of histone protein. As re- .... -108- Fie;. J. Fractionation sc:heme. The yields of Di~A in each fraction are the ;;; :> ~,n and. standard deviation for· 11 determinations. -109- Chromatin (370 µg/m I DNA, 25 mM Na Acetate, pH 6.6) DNase II (100 units/ml) Incubate 5 min at 24° Centrifuge (27,000g, 15 min) t I Pellet Supernatant 0 0 ! 84.6±4.8% of DNA 2 mM MgCl2 Centrifuge (27,000g, 15 min) I t Pel let Supernatant 8 @ 4.1±2.5% 11.3 ± 3.9% 0 I rl rl I Table I. Chemical Coml?.osition of Chromatin Fractions. RNA 1.73 Total Protein 0.96 1.15 1.60 <0.05 0.58 Com.1222ition Relative to DNA (w/w) DNA 0.05 1.01 0.61 Sample 1.00 - 2.21 0.60 S2 Subfractions 0.3-0.4 2.94±0.06 0.54 0.72 3.2 1.35 Nonhistone Protein Pl Chromatin 1.00 0.25 0 . 24 Histone Protein* 12 S" Subunits 1.00 0 . 88±0.06 11 S2 Chromat in - 1.00 0. 3-0 . 7 3-5S 14.9±0.8s(16) 1.00 } 1.00 + 20.2±0 . 7s (16) + -111.:.. Tahle I. l<· Footnote. Chromatin was separated into acid-soluble (histone) and acid-insoluble (nonhiston e ) protein (28). For Pl, "12S" and total S2 chromatin, all acid soluble protein indicated is histone; however, we do not know what proportion of acid soluble protein of the S2 subfractions is histone protein. +Sedimentation values± S.D. determined from 16 gradients equivalent to those in Fig. 3. -112- ported previously (2l1), fraction S2 lacks his tone I. The melting char acteri~tics of the chromatin fractions are pr esented in TabJ. e II. The mean temperature of thermal denaturation (rr ) of the Jll Di~A in P.1 chromatj_n is 69°; however, s~veral transitions can b e recognized in a derivative plot of the melting do.ta . Li and Bonner (32) have investigated the melt ing of pea bud nucleohistone and have assigned the transition at 42° to r egions of free DNA, the transitions at 52-60° to DNA complexed with nonhistone protein, and the t ran sit i ons at 66 and 81° t o DNA complex~d with histone protein. Table II lists the fraction of total hyperchromicity in each of five melting transitions observed in rat--livcr chromatin. About two-thirds of the hyperchromicity of Pl chromat in DNA comes from melti ng transitions IV and V, the histoneinduc ed transitions. In contrast, S2 chromatin DNA has a T of 55° m and about two-thirds of the hyperchromicity is due to nonhistone-induced DNA melting and fr ee DNA mel~ing transitions. This result is consistent with the chemical compositions of these chromatin fractions. It should be noted that S2 chromatin exhibit s a lower final hyperchromicity than either DNA or Pl chl'omatin ('J.'able II). The range of hyperchromicities obs erved for S2 chromatin is 7-16% with a mean and standard deviation of 11. 4 ± 3. 5% ( 6 determinations). This reduced hyperchromj.ci ty cannot be due sol ely to the TINA content of this fraction; perhaps much of the DNA in S2 chromatin is unpai r ed or is in DNA-RNA hybrids. The source of the decreased hyperchroruicity is removed by the combination of pronase and ribonuclease treatment; purifi e d DNA from fraction S2 exhibits 70-75% of native hyperchromicity. Table II. Melting Properties of Chromatin Fractions. Hyperchromicity (61°) III 0 IV (67°) 0.15 0 (76°) Fraction of Hyperchrornici ty in Mel ting Trari"sition* (52°) II 0 0.25 Sample I (42°) 0 0.21 o,. 58 (76°) 0.42 (81°) V 1.0 0.21 (%) 35 0 . 18 m 42 ll (oc) DNA 55 0.34 Average Melting Temperature (T ) S2 Chromatin 0. 31 I (Y) 0.20 0 0.11 0 0.03 0 33 0 69 35 Pl Chromatin 77 rl rl I "12s" Subunits *Derivative plots of the data were calculated according to Li and Bonner (32) and the relative area under each transition was determined. -114- Subunit Structure of Chrr,matin. Rat-liver chromatin and fraction Pl chromatin have l)een digestuu with DNase II for extended periods of time (90 min), and th e resulting soluble chromatin analyzed by centrifugatfon in isok:i.netic sucrose gradients (Fi g . 2). About 40% of the input DNA sedimu 1ts extremely slowly; the bulk of this DNA is acid soluble and hence has been reduced to oligonucleotides by the nuclease. Fig. 2 presents data for whole chromatin depleted of histone I (Fig. 2A) and for Pl chromatin (Fig. 2B). Most of the chromatin sediments at about 11-138, with some material sedimenting more rapidly. The proper- ties of nuclease-resistant particles from unfractionated chromatin have been described in detail elsewhere (8-14, 16, 18). properties for the particles from Pl c~romatin: We find similar the particles are composed of equal amounts of protein and DH/I. (by weight), and the protein compl ement is &,lmost entirely histone (Table I). Subunits from native chromatin sediment slightly.more rapidly than subunits from histone Idepleted chromatin (Fig. 2). The DNA of subunits isolated from native chromatin is 211 ± 9 nucleotides in length while subunits from .histone I..,. deplet ed chromatin contain DNA 166 ± 16 nucleotides in length. These value s were estima tca by sedimentation velocity measurements under alkaline conditions. The nucl ease-resistant particles resemble \!-bodies (2) in the electron microscope (18). The DNA in the nuclease-resistant particles is highly stabilized against thermal de naturation; me,lting transitions at 76° and 81° can be recognized in the derivative plot of the data (Table II). The overall T i.s 77°, about 35° above the T of deproteinized DNA in the same m m solvent. -115- Fig. 2. samples. Sucrose gradient sedimentation of DNase II-treated chromatin A. Histone I-depleted chromatin was digested as described in Methods; ~he soluble chromatin was layer e d on isckinetic gradients and centrifu6 cd for 42 hr in the SW25.l rotor. Fractions were pooled as indicat ed und rerun c:'l. 5-24% isokineUc gradients (inset). B. Fraction Pl chror.iat in was digested and soluble chromatin was centrifuged for 36 hr. :r·ractiorw we.re poolC;d a,; indicated and rerun (ins et) . -116- Absorbance (260 nm) .l> ,, ~ 0 (") (X) 0 ::, l zC: I\) (J) 3 en Absorbance (254 nm) CJ" ct) ~ -~ ... CD N 0 ~ 0 3 CD - 0 0 ... 3 Absorbance (260 nm) --! _., 0 -0 .,, .CD (Jl ~ - 0 n {>J (J) 0 ::, 0 z Absorbance (254 nm) C: 3 CJ" ro (Jl ~ I\) 4-• CD ... CD 0 0 0 0 ... ::: _., 3 3 (X) (J) Subunit Structure of Ac tive Chromatin . We now ask whether nuclease- resistant particles occur in transcriptionally active regions of chromatin. Chromatin fr om rat liver was treated with DNase II for 5 min, fractionated as before (Fig. 1) , an d S2 material analyzed on isoki netic sucrose gradients (Fi g. 3, lower curve). About half of the UV-absorbing material appli ed to the gradient sediments at 3-5S; greater than 90% of this material is acid precipi table after the 5 min nuclease treatment. Two more rapidJ.y sedimenti ng peaks are seen in the gradient of S2 chromatin: one at 14-15S and another at 19-21S. . These gradients were 3 calculated for particles of density 1.44 g/cm, and so the observed sedimentation coefficients couJ.d be in error if the particle densities 3 are .very different from 1.4-1.5 g/cm . About 6% of the . input nucleic acid pellet ed during the c entrifugation. It was ~ound previously that histone I-deficient subunits sediment at 11.5 ± 0.3S (Fi g. 2A) .. Since chromatin fraction S2 lacks histone I, we would expect that if this fraction contains nuclea se-resistant "v-body" structures, these particles would sedim ent at ll-12S. No material was seen at this S value in fraction S2 chromatin (Fig. 3); however, fraction S2 produced by a 5 min DNase digestion might contain these particles in the form of dimers and hi ghe r multiruers (>14S). Therefore, fraction S2 was redigest ed with DNa~;e and analyzed as before (Fig. 3, upper curve). Redigest ion was accompli shed by reacti vating the DNase present in fraction S2 by the addition of EIYI'A (pH 6.11, see METHODS). The EDTA #- presumably acts by chelating inhibiting Mg ions. chromatin was carried out at 24° for 20-120 min. Reincubation of S2 Upon analysis in iso- kinetic sucrose gradients, no significant changes were observed in the >lOS region of the gradients. ... -118- Fig. 3. Sucrose gradj_ent sedimentation of template-active fraction S2 chro1M.1;~in . Curve A: ChroIT'atin was fractionated (Fig. 1) and S2 material centrifuged for 17.5 hr at 39,000 rpm in an isokinetic gradient. Curve B: Fraction S2 was isolated and DNase II reactivated by the addition of 20 mM EDTA (pH 6.4). Incubation we.s for 1 hr at 24°. The reaction was terminated by raising the pH to 8 with 0.1 M Tris•HCl (pH 11), and the sarnple was centrifuged fer 17,5 hr at 39,000 rpm in Bn isokinetic sucrose gradient. -119- Absorbance (at 254 nm) ..,_ (>J I U1 (/) m 0 ➔ -+ 0 3 -120- On redig er;tion of S2 chr0matin, it was found that 50-60% of the UVabsorbing material became acid-soluble (Fig. 4). per formed in three ways: Redigestion has been reactivation of DNase II; addition of fresh DNase II under ~t a ndard conditions of enzyme treatment; and addition of staphylococc al nuclease. In all cases, 50-60% of the input S2 chromatin A b ecame acid soluble. 260 When the reaction approached completion (50% digestion), the solutions becaJne turbid; this was observed with both nucleases. The nuclease-resi s tant chromatin DNA was found to have a weight-average single-strand length of 120 nucleotides. These observa- tions on S2 chromatin are similar to those of Clark and Felsenfeld (8-9) with unfractionated chromatin. The chromat in species at 15 and ·20s both exhibit T 's of 55° m (0.2mM EDTA, pH 8) and reduc ed hyperchromi cit ies relative to DNA or Pl chromat in. The material at 3-5S exhibits the same Tm as protein-free • DNA, l12°. The ch emi cal c9mposit ion, of the s ubfrac tion s of S2 chromatin have been dete rmined (Table I). The material at 15 and 20S is enriched in both RNA and nonhistone chromosomal proteins. -121- Fig. 4. Kineti c s of di ges tion of chromatin fr a ction S2. ;;·a s frs.ctior.at cd as describe d (Fig. 1). nucleases in three ways : Chromatin S2 chromatin was incubated with rea ctivation of DNase II (o ); addition of fresh DNas e II ( 1:1) ; ad.dit ion of st aphyloc occ a l nuclease ( Cl). Aliquots were taken at Vhr ious time s to t est for th e production of trichloroacetic acidsoluble mat e r ial (measured by absorbance at 260 nm). was 7% (w/v). The TCA concentration -122- % TCA Soluble A260 N (>J (Jl 0) 0 0 0 0 ~N 00 -f -· 3 (0 ---3 0CD~0 :::, 0) (X) oo .......... .......... t> .£.o ........ t> DISCUSSION It appears that the template-ac:tive fraction (S2) is organized in a fashj_ on similar to that of inactive chromatin; th a t is, the active fracti on is composed of r egions of nuclease-senGitive and nucleaseresistant DNA, However, nuclease-resistant segments in transcription- ally :inactive chromatin are due to histone-DNA interactions (the\)bodies), while the nuclease-resi stant portions of active chromatin arise from the interaction of DNA with both histone and nonhistone chromosomal proteins. Thermal denaturation of S2 chromatin (Table II) shows that S2 DNA 'is complexed primarily with the nonhi stone proteins (32). These results shed new li 6ht on the find:ings of Felsenfeld's laboratory (8-9,33). These inve s tigators have reported that portions of the globin gene are found in both nuclease-sensitive ("open") and nuclease-resistant ("closed") regions of reticulocyte chromatin. The present collllllunication suggests that Felsenfe.ld's method of isolation of "open" and "closed" segments of chromatin (33) should not discriminate between transcriptionally active and inactive chromat in regions, in agreement with the globin result ( 33). l:<'elsenfeld' s data indicates that regions of the globin gene, are always cuvered by protein (33) but makes no distinction between hi stone or nonbistone protein. }'rom the present data, we speculate that this protein might be nonhistone. -124- The distribution of nuclease-resistant complexes in templateactive chromatin parallels th at of inac tive chromatin; however, thermal denaturation and circular di chro ism studies (34,35) sugge st that active chromatin is in a more ext,~nded , DNA-like conformation than inactive chromatin. Nuclease-resistant particles from whol e chromatin exhibit a nonconservative circular dichroism spectrum, suggesting that the DNA may be in a conformation different from that of the B-form in aqueous solution (11). Protein-induced folding of B-form DNA might also account for the observed circular dichroism spectrum. The electron microscope has revealed differences in the structure of transcriptionally active and inactive regions of chromatin. Ribosomal genes in the act of tran- scription are the length of their tnrnscription product ('pre-rRNA) (36). On the other hand, DNA complexed with hi stones in the \1--body configuration is one-seventh, the length of deprotcinized DNA (4). The basic fiber diameter of i nac tive chromatin is 100 A; however, active chromatin has a fiber diameter of about 30 A (37-38). Thus both physical chemical and E.M. studies suggest that the DNA of active chromatin is more extended than the DNA of inactive chromatin. It is likely that this is why active chromatin is mere su:c: ceptib1e to nuclease; this differential sensitivity to nuclease forrns the basis of our fractionat:(.on procedure (Fig. I). We have founct that i.solated \1-bodies and V-body containing chromatin i s precipit ated by divalent cations (Gottesfeld et al., unpublished), while activ e chromatin remains soluble. We propose that the v-body is both the basic unit of structure and the physical mechanism of gene repression in the eukaryotic chromosome. The subunit structure of active chromatin appears to be different from that of inactive chrom~t in. Further work on the active fraction should yield -1 2.5 - much yaluable information about the structure and mechanisms of the euknryotic gene regulatory apparat u s . -126- ACKll ,J\{l,EDGr:1-SNTS 1s'e t,:•,uri :: Tlr, ! , . J1r:.uva,; fo :c J·,c Jp with r~el elect rophoresis. Helpfu1 di,;,,·.tss i cn~; w·i 0; 1, Tu· . K. T'c·tc1·~; , lir . W. \·/hc a tley and 1-'rr. R. Murphy u rc 1;~·:1tAfull;y c1ck:1c,wleu c,)d . 'l 'h i s work v as supported by the U.S. Public Health Serv ice (GH 86 and Gi•1 13762). -127- References \ 1. Paraon , J.F., Wilkins, M.H. F ., & Richards, B.M. (1967) Nature 215, 508-509. 2. Olins, A.L. & Olins, D.E. (1974) Science 183, 330-332. 3, Woodcock, C.L.F. (1973) J. Cell Biol. 59, 368a. 4. Griffith, J.D . (1975) Seier~ 187, 1202-1203. 5, Outlet, P., Gross-Be1lard, M. & Chambon, 6. Hewi_sh, D.R. & Burgoyne, P. (1975) Cell 11, 281-300. L.A. (1973) Biochem. Biophys. Res. Connun. 52, 504-510. 7, Burgoyne, L.A., Hewish, D.R. & Mobbs, J. (19711) Bioch e.m. J. 143,67-72, 8. Clark, R. & Felsenfeld, G. (1971) Nature 229, 101-106. 9, Clark, R. & Felsenfeld, G. (19'14) Biochemistry 13, 3622-3628. 10. Axel, R., Melchior, W., Jr., Sollner-Webb, B. & Felsenfeld, G. (1974) Proc. Ns.t. Ac acl . Sci. USA 71, 4101-4105. 11. Rill, R. & van Jiolde, K.E. (1973) J. Biol. Che.in. 248, 1080-1083. 12. Sahasrabuddhe, C.G. & van Holde, K.E. (1974) J. Biol. Chern. 249, 152-156, 13. Noll, M. (1974) I,atQre 251 , 249-252. 14. Oosterhof, D.K., Hazier, J.C. & Rill, R.L. (1975) Proc: Nat . Acad. Sci. USA 72, 633--637. 15, Kornberg, R. (1974) Science 184, 868-871. 16 . van Holde, K.E., Sahasrabucldhe, C.G., Shaw, B., van Brugger, E.F.J. & Arnberg, A.C. (1974) Biochem. Biophys. Res. Commun, 60, 1365-1370. 17, Senior, M.B., Olins, A.L. & Olins, D.E. (1975) Science 187, 173-175, 18. Gottesfeld, J.M., Kent, D., Ross, M. & Bonner, J. (1975) In: Florida Colloguim on Molecular Bioloey, eds. Stein, G. & Stein, J. (Academic -128- Press, New York), in press. 19, Baldwin, J.P., Dosely , P.G., Eradbury, E.M. & Ibel, K. (1975) Nature 253, 245-248. 20. van lloJde , K.E., Sahasrab,1dclhe , C.G. & Shaw, B.R. (1974) Nucleic Acid Res . 1, 1579-1586 . 21. Noll, M. (1974) Nucleic Acid Res . 1, 1573-1578. 22. Marushige, K. & Bonner, J. (1971) Proc. Nat. Acad. Sci. USA 68, 2941- 2911 4. 23. Billing, R.J. & Bonner, J. (1972) Biochem. Biophys . Acta. 281, 453-lr62. 24. Gottesfeld , J.M., Garrard, W.T., Bagi, G., Wilson, R.F. & Bonner, J. (1974) Proc. Nat. _Acad. Sci. USA 71, 2193-2197. 25. Marushige,.K. & Bomier, J. (1 966 ) J. MoJ.. Biol. 15, 160-174. 26. Noll, H. (1967) ~ature ~15 , 360-363. 27. Studier, F.W. (1965) J. Mol. Bi£l.. 11, 373-390. 28. Bonner, J., Chal key, R., Dah'Tlus, M., Fambroueh, D., Fujimura, F., Huang, R.C.C., Huberman, J., Jensen, R., l'l.aru.shige, K., Olenbusch, 1-I., Oli.vera, B. & Widholm, J. (19 68) In: Methods in Enzymology, eds. Grossma.n, L, & Moldave, K. (Ac edem ic Pre ss , New York) Vol. 12, part B, pp. 3-65. 29. King, .J. & Laemmli, V.K. (1971 ) J. Mol. Biol. 62, 465-477. 30. Panyi.m, S. & Chalkley, R. (1969 ) Arch. Biochem. Bi.ophys. 130, 337-346. 31. Schmidt, G. & Tannhauser, S.J. (1945) J. Biol. Chem. 161, 83-89. 32. Li, H.J. & Bonner, J. 33. Axel, R., Cedar, H. & Felsenfeld, G. (1973) Cold Spring Harbor Symp. (197J) Biochemistry 10, 1461-1470. Quant. Biol. 38, 7"73-7 83. -129-- 34. Gottesfeld, J.M., Bonner, J., Radda, G. K., & Walker, I. 0. (1974) Biochembstry 13, 2937--29~5. 35, Polac ow, I., & Simpson, R. T. (1973) Biochem. Biophys. Res. Commun. 52. 202--207 . 36. Miller, O. L., & Hamkalo, B. A. (1972) In: Molecu.J.ar Genetics and Dev~~~pmental Biology_, ed. Sussman, M. (Prentis-Hall, New Jersey) pp. 183- 199, 37, Griffith, J. (1970) Ph.D. Thesis, California Institute of Technology, Pasadena, California. 38, Mill er, 0. L. (1965) Nat. Cane. Inst. Monographs 18, 79-89. -130- VI. BIOPHYSICAL STUDIES ON THE MECHANISM OF QUINACRINE STAINING OF CHROMOSOMES The material in this chapter is reprinted by permission of the American Chemical Society from Biochemistry, Volume 13, pp. 2937-2945, 1974. -131- [Reprinted from Biochemistry, (1974 ) 13, 2937.) Copyright 1974 by th e American Chemica l Society and repr intcd by permission 0f the coprright owner. Biophysical Studies on the Mechanism of Qui na crine Staining of Chromosomest J.M . Gottesfeld,• J. Bonner, G. K. Radda, and I. 0 . Walker ABSTRACT: The nuorescence of quinacrine was measured in solution in the presence of interphase chromosomal material (chromatin) and in the presence o f chromatin which had been fractionated into extended and condensed regions (cuchromatin and hetcruc hrom a tin) . Quinacrine nuorescence is quenched most effectively by the cuchromatin fr acti on. intermedi a tel y by unfractionated c hromatin, and least e ffectively by the heterochromatin fract ions. These differences a rc a boli s hed when t he nuorescencc of quinacrinc is meas ured in the presence of DNA isolated from chromatin a nd the chromatin fraction s. Spectrophotometric titrati ons indica te that the assoc iation cons ta nts for quinacrinc binding to the various chromatin fraction s differ by onl y a factor of (wo, and th a t the number of sites per nucleotide av~.i lab lc for yuin. ,crine binding a( sa(uration a rc nearly identi ca l for all fr act iom. Ci rcular dichroism spec trosco py suggest <:d (hat the conformati on o f the D N A in the euchromatin fr action is most like th ;, t o r prot,,in-frcc DNA in aqueous solution (' ' B"-form DNA ) while the DNA in t he hc tcrochromatin fractions is partial ly in th e "C" conforma tio n. These res ults sugges t th a t protein - DNA interactions in chromatin are responsible for the nu orescc nce patterns observed and t ha t chromosome banding with quin.ocrine might arise from diffe re nces in pro1cin - D N A int crac ti0ns (a nd DNA conformatio n) along the chrornatids of m eta phase c hromosom es . C as persson el al. ( 1968) and others have shown tha t the nuorescent dye qu inac rine mustard stains specific regi ons of chromosomes with a ve ry brillia nt in tens it y, leaving other areas of chromosomes re lativel y dark . It was origina ll y thou gh t th a t the linea r d ifferenti a t io n of chromosomes int o nuorcscc nt ba~ds a nd poorl y staining interband reg ions was due to the specific alky la tion of g ua nine res idues by the mu stard fun c ti on o f the d ye. The finding that qu inacri ne itself pr0duces identical banding pa ttern s su gges ted that alkylation of guanine res idues was not the prima ry mec ha ni sm o f nuore.scencc s ta ining of chromoso mes (Cas persson el al., 1969). Ell ison a nd Barr ( 19 ./2) ha ve suggested th a t enhancemen t of quinacrine nuoresce nce might be a fun c ti on of base ratio . .vith (A + T)- ri ch regio ns nuorcscing brightly. W eisblurn ~n<! de H aseth (1972) a nd Pac hma nn and Ri g le r ( 1972) have investigated quinacrinc n,10rescencc in 1,i1ro with a se ries of natural and synthetic po lynu c lcotidcs. and found th a t A •T base pairs a rc responsible for fluoresce nce enha ncem en t. Gu a nin e resi dues were sh0wn to give r ise to qu e nchir.g of quinac ri ne l'l uorcsccnce. Th ese da ta , and severa l other lines of e vid e nce (Sc hreck el al. , i 9 7); l.umho lt a nd Mohr. 1971 ). s ugg ested that the nuoresce nt. bands observed with q u inacrine- ; tained chromosomes are indeed ( A + 1) rich. Th is inves tigation was undert a ken to de te r mine w hether quin ac ri,,c n ourescc ncc in the prese nce o f isolated chru mosomal mat er ial is due solc :y to intrachron,osomal di!Taenccs in DNA base co mpositi on . or whether DNA-protein int eractions in chromat in play a role in producing ba ndin g p,1 ttcrn s. Chromatin l,~s been fra c ti o 11:.Jcd into extend ed and condensed regions (euchromatin and hctcroch,omatin) (Uillin g a nd Bonner. 19 72; Bunner el al., 1974; Gottcsfcld el al., 1974). It has been su ggest ed that the e xtended ch romatin fracti o n correspo nds to t Frnm lhc Division nf Biol(1gy, C aliforni a Institute of Technology, Pasa de na . Ca lirorn ia 91109 (J . M . G . ). and the Department of Aiochcrni stry . lJ ni ve r, ity or (hf.ird. South Parks Road. Oxrord. O.X I JOU. f:ngLln<l . J<, ·,·t'it'l!d J),~c,·mh<'r ~'n. /9 7J. J. M . <i. c,rn·-..,1::-. hi 'i grat1tutlc to tht.: U. S. - U . K . FJuc11i1l 11;tl Commission and the U . S . Lh-pa rlm cnt of S1.11.: for fin :1111.,:i~tl surp, 1rt ,1rhilc at O xftJ rd { t-11 1>:>rit!htHa y-; !\ct. P. 1. :S7 -.~~ h). Thi s ,qirk Y. ;1, suppo rll'c.l. in r•irt. hy 1h c U . S. Public Health Service (Gra nt ( i ,\1 Xt, a nti (i M. 1371>]). a nd , in pa rt, by the S..:icncc Rc~ca ri. : h Cuu1H.:il of Great Brit ai n. BtO C flFMISTRY , \'01. 1 3. NO . 14, 19 7 4 2937 -132TAIILE 1: Di s1ribulion uf DNA i11 the Chrcn>alin f'rnc1ion s. '.;-~ of Tola! DNA Sample Ral chromatin Euchromatin {S2) Heterochronwtin {l'l) Hcterochrom:,tin (1'2) Drvsophilil chromc1 tin Euchromatin (S2) Heterochromatin (P l) Heterochroma tin (1'2) 5-min DN:1 sc 15-min DN:,sc Exro,11rc Exposu re II 3 ,le 3 9" 84 (, + 4 . 8 4 . I ·- 2 5 o• 25 0 + 3 2.1 9 ,J·, 2 9 50 . 6 ± 3 . 3 25 6 cJ, 3 . 2 ' 28 2 t I .9 45 4 + 5. 8 a Mean of 11 dett-rn,inatior.s ± SD. 'Mean of 9 determinations± SD. 'l\kan of 3 d e tcrn)inat: <.'ns ± SD. the templ a te nctivc portion of chrn111:iti11 in vivv a nd that the hcterochromatin fractions correspnnJ tu th e rcprcs,cd or tcmpbte inactive por!ion of chrolTlat in in 1·ivo (Bonner er al., 197 3, I 974; Gottesfcld rt al., I 974). Flu c,resce ncc and bind inc d ~ta have been obtain ed for the interacti,,n of quinacrinc with chromatin a nd the chr0matin fractions . Sirniricant differences in the quantum yieid of quinacrine buund to these fraction s were found; these differences were found to be due to protein - DNA interactions and not differe nces in DNA base compos ition. In add iti on, circulu d ic hroi sm st udi e.s of the chromatin fractions have yielded in formation on the pro1c,n-ind1a:ed co11forma1ional alteration of DNA in th ese fractions. Materials and Methods Preparation of Chroma/in. Chromatin from rat liver was prepared by th,., method of Maru ,hige and Bonner ( I %6). Drosophila embryo chrumatin was a gift of Dr. S. C. R. Elgin (E lgin a nd Hood , 1973). The crude chromati n from both rat ar.d Drosophila"'" purified by centrifugation through I :1 M sucrose as dcsc:ril>d by Marushige and Bonner ( 1966) . The sucrose pellet wa s ho moi;en ized in 0.0 I M Tris-Cl (p l I 8) ancl centrifuged at 27 ,SQOg for 2u min. Frac1ionatio11 of Chroma/in. Purified chromatin was homogenized with a motor-driven glass-Teflon homogenizer (20 strokes at JO strokes per minute) in apprvximatcly 20 - 30 ml of 0.01 M Tris-Cl (pH ~) The homogen;,t c was dialyzed overnight vs. 6 I. of 0.02S M sodium ac,·ute (pl I 6.6). After dialysis. the dialysate was homogenized :is before with th e m0tordriven glass-Tefion hum ogmize r. Th e volume of the chroma tin solution was ,1rljustcd to give 10 A 21,:, unit.;/m l w hen the absorbance of a 11 ,i!iq uot -( 100 µ I) "a, me;, ,.urcd in 0.9 N NaOH. DNase II (Worthington. HDAC) w:is 3dded to IO u nits of enzyme / Air,o unit uf chromatin ( I 00 un its o f cn,yme/1111). The digestion was a !J,11,rd to proceed for 'i 15 min at ambient temperature (24°) . .'\ t the end of the ineubati0n pc-riod. the pH uf the ch rom at in suspension was broug~t to 7.'i w ith 0.02 .\,I TrisCl (p l-I 11 ). The chromatin suspension was co,1kd 011 ice and then cent rifu ged at 27.500g fo r 15 min at 0 -- 4° . To the supe rnatant, 0.2 M MgCI: was added dropwi,e to give a final concentration of O.tlil] \t. After JO-min stirring at 0 - ,1 °, 1he sc 1:,tion was centrifLl~ c·d apin at n,:-OOg f,ir 15 min. The final supn11;,t,!lll fr()m ahl\'e ;, lcrmcd S2 (,., xtcndcd cuchrom:llin) . 1\ 5- rnin 11ucle;4sc trc:ttn1L•rit of rat li v,;r r hroma- tin yie lds l(J -- 1~'\, of the lt1tal ,:hro111;iti11 nucki c '" ·id in S2. The DNA prcp:ircd fnllll th,· S~ d1101na 1i11 fr;iction (5-min DNase trcatn:ent) is duuhlc st randeJ :ind 700 base pairs in 2938 BtOCIII ~l l S TRY, \· UL 13, NU 1 4, 1974 GOTTESFELD er al. length ( number-average le ngth). Longer expos~rc to DN;i ; e 11 yields up to 25% of the chromatin D Nt\ in S2. \\'hen prepJ red from Drosophila embrvo c hroma tin , t he euchrornatin fraction also contains 25% of the total chromatin DNA ( 15-min di:;,,stion). Table I lists the Jmoun ts uf nucleic acid found in the fmctions obtained from both rat liver and Drus ·iphi/a chroma tin. The pellet heterochro matin fractions are tamed PI (first pellet) a nd 1' 2 (second pellet). Gottcsfcld and Bonner ( 1974) have shown that th e Pl fract ion from rat chr<imatin is rel atively inaccessible to enzyme attack. Thu , the D:'\ A ir. th is fract ion is like ly to be more highly compacted and may correspond to th e constitutive or cent r ic heterochroma1 1n in metaphase chromosomes. Measurement of template activity aPd D:--: .-\ RNA hybridi za tion studies (G o ttesfcld er al., 1974) show that the pellet or helerochromatin fractions correspond to the template inactive fraction of in terphase chromatin. Table I indicates that very little nu cleic aci d is found in the second pellet fraction (P2) after a 5-min enzyme treatment. Bonner er al. ( 1973, 1974) h11ve given the chemical compusitions of the fractions obtained from ra t chromatin . Evidence for the validity of this fractionation procedure is a lso presented in Bonner er al. ( 1973 , 1974) and in Got tcsfeld et al. (1974). Prepararion of DNA . DNA was prepared from chromJtin a nd the chromatin fractions in the followin g manner. The chromati n preparations were first extracted with phenol (saturated with 0.0 1 M Tris- C l, pH 8) unti l no visible material remained at the inte r face. The aqueous phase was extracted repeatedly with an equal volume of c hloro form - isoamyl alcohol (24: I. v / v). The aq ueous phase was then dialyzed overnight vs. 0.0 I M T ris-Cl (pH 8), containing 0.5 M NaC l. The dialysate was treated with ribunuclease (Worthington, RAF; preincubated at 80° for JO min) at 50 µg / ml for I hr at 37°. After ribonuc:lease treatment, EDTA (pH 8) was added to 0.01 Mand sodium dodecyl s ulfate to 1% (w / v). Pronase (Calbiochem, CB; preincubated a t 37° for I hr) was a dded to 50 µg/ml, and the mix: ure was incubated a t 60° fo r 1- 2 hr. After enzyme treatment s, the DNA was phenol and ch lorofor m-isoamyl alcoho l extracted as before. The final DNA solution was d ia lyzed extensively vs. distil led water and then lyophili zcd. E.r1imario11 of Nucle ic Acid Concenrrar ion.,. DNA concentrations were de termined spectrophotome trica ll y using an <i60nm value of 6600. DNA concentrations of chroma ti n sa mples were estimDted spectrophotometrically on solutions diluted with I N NaOH (final NaOH concentrations were 0.9 -- 0.98 N). An A 26 o nn, reading of 1.0 corresponds to 37 µ g/ml of chromatin DNA. Fluore<cc11ce Measuremenrs . Fluorescence ,pectroscopy wa s perfo rmed with a Hitachi Perk in -E lmer spec,ronuorimetcr Model MPF-2.'\. /\II measurements were made at 24 -- 25° . rlu ore,cencc of quinacrine in the presence of nucleic acid and nucleoprutcin preparations was measured at a quinacr ine concentration of 2.0 X 10- 6 \t. The buffers used in th is study were 0.1 M sodium phosphate (pH 6.8 -7. 0) and 0.01 M Tris-Cl (pH 8). F luorescence values reported herein a rc given relative to t he fluore scence of quinacrine mea s ured in the same solvent. Quin ac rin e was the dihydra te hydrochloride (mo! wt 506.9) obtained from the Sigma Chemica l Co ., S1. Loui,. Mo. The e xcit~tion wavcl~ngth was 424 nm, and an emission .s pec tru m was obtained for each sample. The excitation and emission sl it s we re maintained at 6 nm. To test th e effect of chromosoma l prot eins on the fluorescence of a quinacrinc -c hrnm.llin complex. l'r<111;isc wa, "d d cd 10 the mixture and t he tluorc,cencc output of the so lu tion was munitorcd with time . To 0.5 ml of a solution eon!aining rat I i j_ -133MI: C II A N IS M O I' () U I N ACR I NF ST A IN I N c; liver PI fraction he1croch run1:1t in (at 83 µg/ml of chrnmalin DNA) and quinacrinc (al 2.0 X 10 · " M) in 0.01 ., 1 Tris-Cl (pH 8) , 50 µI of a 200-µg/ml sulu1ion o r Prun ase (p rc ineubatcd at 37° for I hr) was added ,11 l ime ,ero. The e,ci1ation wavelength was 424 nm . and em is, iun was mon itored fur JO min a t 495 nm . No significant in str ument dr ift was obse rved wit h quinacrine a lone measured at interva ls ove r a 30-min period . Pronase (in the absence o f nuckoprotcin) had no effect on the nuorescence of quinac r inc. Measurement of Quinacrin,· Binding to DNA. C!iro111ati11, and the Chromatin Fractions. Owing to the fact th,11 quinacri nc nuoresce1.ce is affected differently by A•T and G ,C base pairs (Wcisblum and de H ascth , 1972: Pachmann and Rigler, 1972), nu orescence titrat ion, with natural DNAs cannot be used to est imate associa1ion constant.s or the numbe r of sites per nucleotide avail:.iblc for dye binding. To study the interaction of quinacrinc with D:-.A and chromatin quanti1a1ivcly, we made use of the fac t that the optical extinction coefficie nt of quinacrine is reduced on binding to DNA. Pachmann and Rigler (1972) have show n t h~t both A-T- a nd G-C -conta ining polynucleotidts cause a hypoc hro mic shift in th e absorption spect rum of quinacrinc. Blake and Peacocke ( 1968) have presented a de tailed trea tme nt of th e use of spectrop hotometric titrations and the Scatcha rd plot in studyi ng the binding of aminoacridines to nuc leic acids. Brieny, if one plots r /Cf vs. r, w here r is defined as the ratio of moles of bound dye (quinac rin e) to moles of DNA nucleotides and Cf is the molar concentration of free dye, the n a straight line is obtained for each class of bind in g sites present if the s ites within each class are equiva lent and independe nt of one another. The s lope of the straight lin e is -Ka, the assoc iation co nsta nt , a nd the intercept on the abscissa is ii. the maximal nu mber of moles of dye bound per mole of DNA nucl eot ide. The molar ex tin ction coe ffic ien t of bound quinacrine, t 420 nmb, was obtained experimenta ll y for each DNA and nucleoprotei n sa mple . At a constant conce ntration of quinacrine (4.4 X 10- 5 M) in 0 .0 1 M Tri s-C l (pH 8) , t he nucleic acid concentration was varied over t he range of 1.0 X I 5 to 3.0 X 10- 4 M. The a bso rba ncies of th e quinacrine-DNA or quinacrine-nucleoprotein compl e xes were measured at 420 nm with a bla nk of DNA or nuclcoprotein in the abse nce of quinacrinc. This se rves to correct for li g ht s<.::ll leri ng a t high nuclcoprote in concentrations. The a bso rhance a t 420 nm was obta ined from a n absorption spectrum performed with a Cary Model 11 recording spec t rophotometer. Al l abso rpti on meas urements were made a t ambien t temperat ure (24_0 ). The value of <<2onmb was o btained fr om the absorption value at infinite nu cleot id e co ncentration . This latter value was obtai ned from a computer fit of a double-rec ip roca l plo1 of th e experimen tal da 1a by the met hod of least sq uares. T able 11 lists th e exti nct ion coeffi cients fo r free q uin ac rine and for quinacrinc bound to va rious DNA an d nuclcoprotein pr epa ra1ions. Data for the Scatchard plot were obtained from exper im ents where the co11ccntratiun of quin,1crinc was va ried ove r a 15fold ra nge at a fixed concent rat ion of nucl e ic acid (DNA or chromati n). These exper imen ts were perfor med as foliows . To 3.0 ml of a solution co n1ai11i11g DI\; ,\ o r nud eoprote in (ranging in co11centra1ion frorr, 1.-1 10 2.5 X 10- 4 , ; in IJNA nu c leotides) in 0.01 M Tris-C l (rl I !:. 0). 5- l<l 20-µI ali4uots of a 5.0 X I o -J M solution of quin:icrine were adJeJ. After each audition of quinacrinc an ab,c1rp1ion spectru m was taken. A bla nk con1aining DN ,\ or nuclcopro1cin in the ab,cnce or quin.1crinc was used to compe nsate fur light sc:t tt c ri ng. The abso rption spectrum of free quinacr in c at e;1ch concenlration ( 1.0 - 15 X 10-~ M) was al,u ob1ainet.l. 13cl'r's law was round ltl be obeyed o- T,,111. i: 11: E, tinctio n Coc!l icicnts of Qui nacrinc (Free:) and Qu in aa inc Bo un d tu Rat Liver DNA, Chromatin, a nd Chromatin Fractions. Sample• E~·!O n,11 Quinacrinc, free Q uinacri ne , bound to DNA Chromatin E uchromatin Heterochromatin 7 . 6 X 10 3 3 9 X 10 ' 3.8 X 10' 3.9 X 10 3 4 .1 X 10 3 • All measurements were made in 0.01 M Tris-Cl (pH 8). over this range of qu in acrine concentrations . The molar fraction of bound q uin acrine was obtained from the absorbancc reading s as desc ribed by Blake and Peacocke ( 1968 ). To obtain Ka, the association constant, a nd r., the ma xima l number of moles of quinacrin e bound per mole of DNA n ucleot ides, a computer fit of the Scatchard plot by the least-squares m et hod was pe rformed. Circular Dichroism Spectroscopy. Circular d ichroism s pectra we re obtained wit h a Roussel-Joua n Dichrograph 11 a nd with a Durrum -Jaseo ORD/UY-5 . Spect ra were recorded at room te mperat ure with a sample cell of I -cm path length. Spectra were scanned from 350 nm dow nward to lower wave le ngths until t he noise leve l became too high to record meaningful data (usually around 210 nm). CD spec t ra were obta in ed wi th samples at A 16 unm 1 cm values of 0.8- 1.5 . Spectra arc repo rted in terms of the difference in extinc1ion coefficien ts for left a nd right circularly polari1cd light, respectively . , 1 - ,, is d efined as (A 1 - A ,)/ le, where A 1 a nd A, a re ,he absorbancies for left and r ig ht circularly polarized light. respcctiv~ly, J ,s the path length of the sample in cent imeters. and c is t he molar concentration of the samp le. Al l spectra prescntet.l herein a rc reported in terms of molar concentra tions of DNA nucleoti des. Thermal Denaturation. The mean temperature of thermal den atura ti on ( Tm) of th e DNA samples was determined with a Gilford spectrophotometer - multiple sam ple abso rbance recorder equipped with a Haake ci rc ulating bath. The temperature of the bath was increased at a linear ra te of a bout 0.5° / min. DNA sam ples were dialyzed agai nst sta ndard saline citrate (0.15 M NaCl-0.0 15 M trisodi um citrate) prior to dena turation. Determination of DNA Size. The s ize of th e DNA prepa red from chromati n and the chromatin fractions was determined by electron microscopy. The D NA samples we re spread from aqueou, ammonium acetate by the method of Davis et al. ( 197 I). Resul1 s Quinocrine Fl uorescence. FLUORESCENCE. OF O1.!:",iACR INE ON RINDING TO DNA AND SYNTIIETIC POL.Y:--IUCU'OT I DES. W hen cxci1cd at 424 nm, t he nu orcsccnce maximum of quinacrine in a4ucous so luti on is 490 - 500 nm . T he ~d diti un of naturally occur rin g DNAs of moderate A-T contcn l (le" than 60% ,\,T) o r G,C-containing po ly nucl co1ides to a solut ion of quinacrine causes t he quenching of fluorescence . O n ihe other hand, (A + T)-r ic h D NAs and polynuclro1idcs cause an enha ncement of 4uinacrin c lluo rcsce nce (Wei sbl um a nd de ll ascth, I 972; Pach mann and Rigle r, I 972). Nei l her qucr. ... hing no r enhanceme nt of nu orcscencc affects 1he shape or the position of th e max imum of 1hc q uinacri ne emis., ion spectrum. IJIOC'lll·. M I STRY, VOi. 1.1 , NO 1 4, 1 974 293 9 GOTTESFEI.O <'I a/. -134- >,-. J) z 1,0 ,-. w ~ W0.8 w u L> w w (/1 (/1 g o.4 a: '"g u u0.6 • w a: J ~0.2 W I- > <{ 0. 4 0 .2 <{ J J '---- ' - -- ' ~ - - ' - - -L - - . ; -- -1----',------'---Q 0.1 0.2 o.3 o.4 o .~ s w a: w c,; NUCLEOTI DE CON CE NTR,HION (M) x 10 3 0 0.1 0 .2 NUCLEOTIDE CON CENTRATI ON (M ) x 10 3 >- >,-. '= (/1 (/1 z 1.0 w ,.o I- 1- z z w wO.A u u 0.8 z z w 0 w ~0.6 • (fl w a: u 0.6 w a: gOA 0 ::, J ,._ L,,. 0.4 J w ~02 ?: ,-. I- <{ <{ J w a: 0 .6 u. I- Z 0 J u. w 00 z Z' 0 02 J 0 0.1 0.2 0.3 0.4 0 .5 w S Ct: NUCLEOTIDE CONCENTRATION (M) X 10 3 FIGURE I: Rel..itivc nuorcscence intensity of quinacrinc measured in the presence of increasing concentrntions of unfractionatcd chromatin (e), S2 euchromalin (A). and Pl hctcrochromatin ID) . The d:tt3 for P2 hcterochromatin (not shown) were essentially identical with thJI for unfractionatcd chromatin . Chromatin fraction.5 were isolated a ftl·r I 5. min exposure to DNa sc. Quinacrinc flu orcscc:ncc was measured in 0 . 1 M pho,phatc buffer (pH 6.8) as described in Methods. Data arc in- cluded for (a) rat liver (top panel) and (b) Dro.rophila melanogaster chromatin (bottom panel) . "S" indicates computer fit value of nuorcs- 0 0 .1 0 .2 0. 3 0 .4 O.~ NUCLEOT ID[ CON CENTRATION ( M) x 10 3 FIGUR E 2: Relat ive nuorcsccncc intensity of quinacrinc measured in the presence 01 increasing concentrations of D1'iA prepared from S2 euchromatin (A) and Pl heteroehromatin (D) . DNA wn , prcpued from chroma tin fr action s obtained a fta 15-min exposure to D:'\:.i sc as dc~cribt·d in Methods . Flcorc sccncc \.\a s measured at a quinacrinc con- centration of 2 X 10-• Min 0. 1 M phosphate bufrer (pH 6.8 ) as described in Methods. Data are included for (a) rat liver (lop panel) and (b) D. melanogastl'T chromatin (bottom panel) . "S" indicates computer fil va lue of nuorcscence at infinite nucleot ide concentration . cence at infinite nucleotide concentration . Pachmann and Rigler ( 1972) have presented absorption and emission spectra of 4uinacrine alnnc and quinacrinc in the presence of both (A + T)- and (G + C) -rich polynuclcotidcs . QUINACRINE FlUORESCENCt IN THE l'Rl'SENCE or CHROMATIN AND CHROMATIN l"RACTIONS. Figure I presents the data obtained from nuorcscence titrations of a standard amount of quinacrine (2 .0 X I 6- M) with varying amounts of chromatin and fractionate<l euchromatin an<l hctcrochromatin. Data are presented for two experiments, one performed with rat liver chromatin :ind another with chromatin prepared from 6- to 16-hr embr yos of Drosophila melono{?as/er. The two chromatin preparati ons !!ave the sa me qualitative results. Euchromatin fr action ; from both rat liver and /)ro.wphila embryo c hrom:itin quenched 4uinaerinc nuorcsccnee 10 a greater extent than did either unfra ctionatcd chromatin or heterochromatin . The curves appear to be simple first-order saturation curves . We have performed a fit of the dat a by the lea stsquares method (straight-line fit of an inverse plot) lo obtain nuorcsccnce values at saturating concentrations of chromatin. The saturat i0n value is the relativ e nuorescence of completely bound quinacrine . These result s arc also shown in Figure I. The relative nuoresccncc of quinacrinc bound to cuchromatin is 0.19 for /)rosophila and 0.28 fur rat liver. Thus, the lluorcscencc of 4uinacrinc is 411cnc!tcd bv 'II XO% on binding to cuchromatin; 011 th,· other hand, quinacrine fluorescence Is 4ucnchcd by only 22 35% on binding to h,·tcrochr,,lllatin . o- 294(} BIO C II FM IS TRY , VO I 13 , NO 14, 1974 Quenching by the nucleoprotein samples die! not shift the emission maximum of quinacrinc by more than 5 nm. PROTElr, - D\;A INTERACTIO.'~S AND QUI N ACRINE FL UORESCENCE. We now attempt to answer the following question : Are the fluorescence patterns observed with .:iui nacrincnucleoprotcin complexes due to differences in the base composition of the D\:A of the chromatin fractions. or arc they due in part Ill differences in r,rotcin-DNA interaction s1 The first half of th is question can be answered directly by determining the b~se composition of the DNA isolated from the chrorr.atin fractions. Furthermore, the fluorescence of quinacrir.c cai, he measured in th e presence of these DNA samples. The base comrosition of a D~A prepar~tion ( o f 400 base pairs or g. ,eatcr i n length) c~n be determined by mcasur i n!! its Tm (mean temperature of thermal <lenaturation) in standard saline - citrate t .'v1andel and l\f armur, 1968). Ri t DNA (42'X, G-C) theoretically ,hould exhib i t a Tm of 86° . We observed a Tm of 85° fer r at liver D N A . The DNA prepared from mt liver euchromatin (5-min DNa:.c treatment) exhihited a Tm oi 84°. Similarly, the D .'-<A of rat liver Pl hetcrochromatin had a Tm of 85° . To a fir st approximation, therefore. the base composition of the D:\A~ from the chromatin fr action s arc identical (with 42 ± 2% G,C). When the fluorescence output of yuinacrinc w~s measured in the r'c srnce of varying :,mounts of cuchromatin anti hctcrochromatin DNA. the satur.1tion curves of Figure 2 were obtained . !);,ta for cuchrom;1tin and hctcrochromatin DNA of rat -135M [' C II /\ N I S M 0.55 O I' Q U I N /\ C' R I N F S T -~ I N I N G • !' (/) f- Z 0.50 ::, w u Z0.4:> w u (/) ~ ~04 0 0 ::, 0 N " ...J <! LL 0.35 <J 0 -"---~--~-~--~ --~---~ 24 28 4 8 12 16 20 TIME (MIN) FIGURE 3: Time course of Pronasc treatment of a quinacrinc - hetcro- chromatin mixture. /\t time ze ro (arrow). 50 µI of a 200-µg/ml solution of Pronasc w~1s added to 500 µI of a solution con taining quinacrine (2 X 10-• M) and ra t li ver Pl hetcr-xhromat in (8J µg / ml of nucle ic acid). The he1crochro111:i tin sa mple wJs that obtained after 5-min ex- -0.2 posure of chromatin to D i\Jsc. Other experimental de tails arc given in Methods. Fluorescence is c>..prcsscd in arbi1ra ry units . 350 400 450 500 WAVELENGTH ( nm) ( Figure 2a) and Drosophila ( Figure 2b) are show n. As expected from base comrositi on determinations. the data fo r euchromatin and het erochromatin DNA (from th e s2me organism) are identica l. The nuorescencc titration curves for unfractionated DNA were also found to be the same as those for euchromatin and heterochromatin DNA from the same organism. The nuorescence titrations for euchromatin ( Figure I) most closely resemble those for deproteinized DNA ( Figure 2). a nd the rel ative nuorescencc of quinacrine at saturating concentrations of rat cuchromatin (Figure l a) was the same as that value at sat urating concentration of rat DNA ( Figure 2a). Since base composition differences cannot rea sonably account for the data of Figure I , some ot~r component of chromatin must be responsible. In order to examine the role of chromosomal proteins, the Ouorescence of a quinacrine - heterochromatin mixture was monitored with time after add it ion of Pronase to th e sol ution (Figure 3). Over a 20-min period a marked decrease in Ouorescence output was noted. This is the change one would expect as heterochromatin is deproteinized and transform ed into DNA. Pronase by itself did not alter the nuorescence of qu inacrinc . F urthermore , chromosomal prote in s by themselves (hi s tone and non-histone proteins) did not affect quinacrine nuoresccnce under our experimental conditions (0.01 :vt Tris-Cl (pH 8 ) or 0.1 M sodium phosphate (pH 6.8)). Thus it appears that protein - DNA interact ions in chromatin are respons ible for the observed r~sults ( Figure I). We have indicated that the nuoresccnce of quinacrine bound to heterochromatin is substa nti al ly greater than the nuorescence of quinacrine bound to euchromatin or DNA. We attribute th is difference to the different quantum yie lds of the bound species. This conclusion rests on the value of relative nuorescence a t saturating nucleotide concentrations . However. if the fractions are vastly different with respect to either binding affinities for quinacrine or the number of sites per nucleotide available for binding. the co nclu s ions reac hed above must be reasses~ed. Thcrdo re, a qua nt ita tive estimation of these binding parameters has been obtained. Binding of Q11i11acrine 10 Chromatin and DNA. SPECTRAL CHANGES O N BINDING. Quinacrine in aq ueous solution has absorrtion maxima at 345, 424, and 446 nm . On binding to D:--JA. the absorption maxima arc shifted to longer wavelengths; moreover, the entire spectrum undergoes a hypochromic shift. This is characteristic of the interaction of planar FIGURE 4: Visible difference spectra of quinacrinc in the presence of near satu raling concen tratio ns of rat DNA (0 .3 1 X tQ- 3 M in nuclcotidcs J a nd chroma I in (0 .23 X Io- , M in nucleot ides) as compared to fre e quinacrinc. J\bsorprion spectra were mea sured at a '-!Uin:l cri nc concentration of 4.4 X 10- 5 Mi n 0.01 M Tns-CI (pH 8). DNA; solid line: chromatin ; dash,,d line. dye molecules with nucleic acids (Blake a nd Peacocke. 1968 ). Pachmann ·and Rigler ( 1972) have presented data which in dicate that both (G + C)- and (A + T)-rich polynucleotides ha ve the same effect 011 the abso rption spectrum of quinacrine as naturally occurring D NA ( <60% A,T). On binding to either chromatin or the chromatin fractions, the absorption spectrum of quinacrine undergoes the same hypochromic shift as on bincing to rat DNA . Absorption diffe rence ~pectra a re presented in Figure 4. The absorption spectra of quinacrine - DNA and quinacrine-chromatin complexes were measured at near saturating nucl eotide concentration s. Under these conditions nearly all the dye pre~ent in solution is bound. Spectra were taken with two cuvets in the reference position of the spectrorho tometer; one cuvet contair.ed quinacrine alone while th e second cuvet contained either DNA or chromatin. An instrument baseline was obtained by measuring the absorption of quinacrine with an idc ntiotl cuvet containing quinacrine in the refere nce position. The absorption difference spectra of quinacrine-DNA a nd quinacrine-chromatin complnes are nearly identical ( Figure 4). Thus it appears that quinacrine forms a similar complex with chromatin as with DNA . Lerman ( 1963) has shown that the mode of interaction of quinacrine with DNA is intercalation. It is tempting, therefore, to speculate that the mode of interaction of quinacrine with chromatin is also via intercalation of the dye between the stacked bases of DNA. Since the absorption spectrum of quinacrine undergoes the same transition upon interaction with DNA or chromatin, s pectrophotometric titrations can be used to investigate the binding of quinacrine to DNA a nd nucleoprotei ns quantitatively. A spectrophotometric titration of quinacrine with increasing concentrations of rat chromatin. euchromat in, and heterochromatin was performed in order to obtain the value of the extinction .:ocfficient of bound quinacrine . Figure 5 and Table 11 present the results of this experiment. The values of , 42 o nmb are nearly identical (±3.5%) fo r DNA. chromatin and the chromatin fractions. Even though the absorption values at sat- 2941 -136(; 0 TT E S F E I. D e I a / . TABLE 11 1: Data from the Scatchard Plots. n (Sites/ Samrle DNA Chromatin Euchromatin Heterocht omati n 0 20 ' - - - - - - - - - - 1 - - - ' - - - - - - - - ' - - J - • 0 .1 0.2 0.3 S NUCLE OTID E CONC EN TRATION (M) x 10 3 FIGURE 5: Spec1ropho1,,mc1ric titration ofquinacrinc (4.4 X 10-i M) with increasinR. concentration:-:. of rat chromatin(•>. Pl heterochroma- tin (D). ar.d S2 euchrqn1a1in ( A). The absorption -,f 4uinacrine in 0.01 M Tris-Cl (pH 8) in the absence of nuclcoprotcins was 0.335 al 420 nm ... S.. indicates com puter fit value of ah'-orption a t infinite nucleotide concentration. Chrum::itin fractions we re obtained after 15-min ex- posure 10 o;-;ase. uration are nearly identical for the nucleoprotein preparations, the spectrophotometric titrations arc somewhat different. An equivalent hypochromic shift in the absorption of quinacrine was found at lower concentrations of cuchrornaiin than heterochromatin; that is, more hcterochromatin than euchromatin was required to produce the same decrease in absorption. This holds true for nonsaturating conce ntrations (less than 3 X J0- 4 M in nucleot ides). Again, quinacrine fully bound to chroma tin, euchromatin or hcterochtomatin has essentia ll y the same extinction at 420 nm. SCATCHARD PLOTS. Spectrophotometric titrations were 1.5 .,. I Q X 1.0 u. u ':::- 0.5 0 L - - L - - - - - - - ' - - - - -- 0 0.025 ~,.....__ __ _. 005 1- 1c;1 RI : 6 : SLah.:lu rd rli1t, of dat :1 \lhtainc:J frnn1 ., pl·\:1ruphotorrn:tr ic ti1r ,11iilr.'i of a -.:L11HbrJ ;1mount of ON,\ or t.:hrnrnatin \\Ith 111nc;1..ing cmt.:ullr;1ti11 ;i , ,,1 q1: in:h.:r111c he.:,: \li.,:th uds): I)'\: :\ ~OL u11 fi: 1l' litl11 :1tc<l chrurn ~11in 1•J : S2 cut.:hron1.1tin (A); anJ Pl hctcr,1\.'.hror11at in ( □). Chrom:1tin frJt.:tiu11' \\l.'.rl' ob t~1inl·d after I '.'-min ~,r o ~1:rl· \o D Nas1.; . 2942 BIOCHH,IISTRY, VOL. 13. NO 14, 19 74 K" Kd 4.0 2 6 3 5 2 0 2.5 3 .9 2 .9 5.0 (10·' M - 1) (!O··• ~1) Nucko- Sites/JO tide) 13nsc Pairs ·-· ----- -- - - 0.051 0 . 047 0 . 049 0.047 102 0.94 0 98 0 . 94 alsv performed under co nditio ns where the conccntr:ition of nucleic a cid was fixed and the conce ntration of quinacrinc was va ried over a 15-fold range (see Methods). The results of such titrations were plotted according to the method of Sc,!tchard ( I 949). Figure 6 illustrates S cat chard plots for rat DNA, ch romatin, euchromatin and hcterochromatin. The intercept on the r axis (the absc issa ) is fl, the maximal number of moles of quinacrine bound per mole of nucleotides, and the slope is -K.,. the association constant. The values of ii obtained for DNA , chromatin a nd the chromatin fractions are nearly identical (Table 111); the average value of fl was 0.048 ± 0 .00 2. This corresponds to about one site available for quinacrinc binding per turn of double-helical "B"-form DNA . The association constants obtained from the Scatchard plot s are listed in Table Ill . The values of K, for the int erac tion of quinacrine with DNA and for quinacrine with hcterochromatin differ by only a factor of two. The values of K 0 for euchromatin and unfractionated chromatin lie between the values for DN/\ and heterochromatin. The data of the Scatchard plots ap pear to lie on one straight line; thi s ind ica tes that only one class of binding sites for quinac rine exists in either DNA or chromatin (Blake and Peacockc, 1968) . This finding was quite surprisi ng since A,T and G,C base pairs affect the nuorescencc of quinacrine so differently. From the Sottchard plots we must con ·clude that quinacrinc bind s to A•T a nd G·C base rairs by the same thermodynamic proces s. Furthermore, sin ce the maximal numbe r of sites available for quinacrinc bind ing {ii) is the sa me for DNA, chromatin a nd the chromatin fr ac t ions, chrQmosomal proteins d o not occupy potential bind ing si tes. Chromosom a l proteins do , however, decrease the assoc iation constant for the interaction of quinacrinc with th e nucleic acid in chromatin. These data a re quite similar to those obtain•:d by Sic11pson ( 1970) for the interaction of a reporter molecule wi1h DNA and chromatin. The number of sites available for binding the reporter (ii) was found to be the same for DNA anJ chromatin; however, the association constant for the interaction of th e reporter with chro1c1a1 in was r,ne-half th a t for the interacti,1n of the reporter v.ith DNA. Pachmann and Ri g ler 1. 1972) and Modest ;ind Sengupta ( 1973) have also presented results of binding cxrerimcnts with quinacrine and D;",;A . Using nuorcsccnce polari ,a tion, Pachmann and Ri gle r ( 1972) report an a,sociation co11,;tant of 6 - 10 X 10 5 M- 1 and ii= 0.08 - 0.16. Th ese ,·,tlucs arc i,1 close agreement with those of Figure 6 and Table Ill. On the 01h,·r hand, Modest a nd Sengupta ( 197 3) have repo rted v:ilucs ,if K = 1.6 X 10' M- 1 and n = 0.72. The,c data we re obtained by ultr:ifiltration and probably renen the ionic int eract ion o f quinacrine with ON /\ phosph a tes. T he spectroscopic bindinp. <iJta re ported herein and by Pachrnan11 and Rigler ( 1972) rencct binding 1·ia intcrcal:i1ion. The second (ionic) cl,t.,s ,,r binding sites wa s not dctcct~d in our experiments ( Figure 6): higher dye to nuclc,llide ratios, ht,wl·vcr, ha,-c revealed this cla" of binding . -137MECHANISM 01 QUINACRINE STA INI NG FSTIMATJON OF fllNDl"G PAR ,\MF TJ'RS IT<O~ I Fl.LO· s.o Rl'SCENCE DATA . We have mentioned before 1h;;1 the binding paramelcrs /(, and 11 c;111r1'Jt be dc1a111ined indq1enden1ly from nuorcsccncc titrat io n, with natura l DNJ\ , due to the fact th :, I J\-T and G-C base pirs have different cffccls on q11in;;crinc nuoresccn<:e . Helene <' I 11/. ( 1971). however. have shown that the pr0du c1 of K"il ,;in be obtained dircctlyly from a 1i1rntion of a nuorcsccnt molecule with Dl\ A. Eqt•'1tion 6 of Helene el al. ( 197 I) can be modified 10 give 3 genern l equ a tion for the binding of nuoreseent molecules to DNA 4. 0 - I 3.0 Et 2.0 20 0 This equation assu:nes th3t i classes of binding sites exist. cJ>r is the qu a ntum yield of free quinacrine: <PR; is the quantum yield of quinacrine bound to sites of .:lass i; :ind t/> is the observed qu an tum yield. cf>F/ (<bF - <b ) is equal 10 I/ !:>f where c.f is I minus the quantum yield cJ> / ef, 1. Equation I describes a straight-line plot of I/ .:0./ 1·s . I/ C D"-'A· Figure 7 illustrates such a plot for rat DNA and fur poly(rG)•poly( rC). The ratio of y intercept to slope is 'f. ,- K;",,,. From Figure 7. ~,A·/ h; is 2.07 X 104 M - 1 for rat D;',;A, and 2.21 X 10' M - 1 for poly(rG)-poly(rC). From the Scatchard plot for rat DNA, K" is 4.0 X 10·' M- 1 and,, is 0.0 5 1. The rroduct (K,11) is 2.04 X 10 4 M-1. Thus by two independent measurements (nuoresccncc and spcclrophor.ometric titrations), v.c arrive at the same value of Aj1. This indicates that the va lues of K, and n obtained from the Scat..:h.trd plots arc reliable. We must reemphasi1e that nuorescencc titrations cannot be used to estimate K, and n independently for nalural DNAs . Binding parameters can be estimated, however, from fluorescence titration , of quin acri nc with ,ynthetic hom opolymer duplexes. From a Scatchard plot of the fluorescence titration d~ta for poly(dA)·poly(dT) (Weisblum and de'IL!seth , 1972), a K" of 3.3 X 10 5 M- 1 was calculated. This value is quite similar to the value of K, calculated from th e spec1ropho1ometrie titra , tion for rat DNA (4 .0 X 10 5 ~1- 1). Furthermore, the value of K. for poly(rG)-poly(rC) is approxima tel y 4.4 X 10' M- 1. Thus quina crine appears tu have about th e same affinity for A,T and G-C base pairs. Moreover, the number of sites per nucleotide available for binding (ii) appear to be the same for all polynucleotides and nuclcoproteins investigated (11 = 0.048 ± 0.002) . DNA Conformation . CIRCULAR DICI-IROISM SPECTROSC OPY . In recent years circulJr dichroi,m spectrosco py has come into wide use as a tool for investig,,1;11g nu clei c acid conformat io n and protcin -inrluced al teratio n, in DNA conformation. We have n:easured th,, CD spc<:tra of c·hromatin. euchrnmatin, hetcrochr<Jm;itin . and D~ ,\ under a v;,rie ty of model conditions. Figure 8 illustrates the CD spectra for these nucleoprotein s a nd DNA isol;i tcd from rat liver. The CD spectrum of DNA in aqueous sol ution is characterized by a major positive band at 27 5 nm and a ma jor negative· bitnd d 24 5 nm with a 60 4'.; 1/[DNA] (x 10- 3) FIGURE 7: Plot o [ l/.;1/ vs. I/ DNA conccn, rati on for rat DNA (9 ) and rol)(rG)-p,,1,(rC) (f,il) _ The data for p,,ly(r(;) ,poly(rC) we re obta ined fr om Wci, blum an~ de ll ascth ( 1972). I / .c,f i, dd"i ncd as the inve rse of ( I - re l:ni\·I.! nuore~cence int cn~ity ) . Th e \''.'.l lu c of l; K./h 1 was 2.~I X 104 M - 1 for poly(•G),poly(rC) and 2.07 X 10 4 M- 1 [0r rat DNA. the CD spectrum of chromatin are the protein CD band at 220 nm and the nucleic acid shoulder at 245-250 nm. The reduced value of the CD band at 275 nm has been interpreted by ma,,y authors as evidence for the a ltered conformation of DNJ\ in chromatin. It has been suggested that the DNA in chrom.1tin is partia lly in the " C" con formation (Johnson el ai .. I 972). DNA is thou ght to exist in the "C" confo rmation under a variety of conditions: for exa mple, Li sal ts of DNA at 95% relative humidit y ( Marvin et al., I 961; Tt:nisSch ncicier an·d Maestre, 1970) and DNA in ethylene glycol -I crossovt!r at the w avc kn gth ur m ~L\imal ;1bsor pti on. 258 nm . A second positi l'e band is cen tered at about 220 nm. The CD spectrum of clirom:itin di ffas fro m th,1t ,, 1· D:-SJ\ in several respects. The absolute rnJgnitude of the po,itivc band at 275 nm is redu ced from that of D'.\A in chromat in. Bv treating chromatin with dcprntcinizing ;ircnts (sodiu1:1 dn,kcyl sulfJte. NaCl , or rro teolv tic cn,)mcs) , one c;in lra ;;s r,,r,n the CD spectrum frum that of i.::hronutin to one r~sC"m b!ing the CD spec- trum or DN".,\ (Shih and F;;sman. I 970: Si u:p,un. 1977.: I kn son and Walker, 1970) . Bdow 250 nm, the major k.tturcs of 2,;o 300 2so 32 0 WAVELENGTH (nm) I 1Cit1Kl ·. X: Circ.:ular dic.:hroi , m srx:,,:tr~1 of r:1 1 D NA (0), chrom ~1tin {X). S2 1.:uchrorn.1lin (A). Pl h..:lcrochrom:nin ( □) in (l.01 M Tri,-CI (pH X) ~111d r.:l D'.'.°;\ in 9(Y;,, cthyknL· gl~L·ol (e). The c.:hrn111..1 1i n f rac.:lilrn!\ w1.:r1: ob1:1in1.:<l ;.fie:- 15-rnin cxpo.-..urr.: ll> 0:\.a.,r.: . lllOCIIEMISTl<Y, VOi. Ll, NO 14, 1974 2943 -138ba:-.c cnmpositiuP %H ronforrnation 2 64 -0 01 1.40 100 0 ("C" form) 53 2.00 76 0.83 32 (Nelson and Johnso n. 1970) show characteristics of the "C" form . The CD spectrum of ON/\ in e thyl ene g l)CO I (Figure 8 a nd Nelson a nd J ohnso n, 1970) is nearly id,·11ti<.:al the spectrum of film s of Li sa lt D:--J/\ known to be in 1he "C" form (Tunis-Schneider and \1cestre, I 970). We will t herefore refer to the CD spectrum of DNA in ethvlene g ly.;o l as a "(". for m spectrum. The B - C tran sition is characteri?ed iJy an a Im o.s t complete loss of th e 275-nm posi t ive band . 1\li11or CD band s arc seen at 28 .J and 295 nm. respectively. /\ negative shoulde r is seen at 265-270 nm. The major negat ive band of DN !\ is increased in the "(".form srcctrum; the value of ..i,24~ nm for "B"-form ONA is -1.96, \\hilc Ll<N Sn m in the "C"-form spectrum is -2 .67 . The values of Ll< ,1 5 ,.,, for DNA ("B"- .111d ·• (".form spectra) anci the nuclcoprotcin samp les a rc lis ted in Table IV. The spectrum of euchromatin is most like th at or "B"- form DNA while heterochromatin e xhibit., a spectrum more like that of the "C'-form of DNA. The spectrum of unfrac t ionatcd chroma tin is intermediate between those of cuch ro matin and hctcrochromatin . John son <'I al. ( 1972) have assigned a per cent " B" contribution to chromat in DNA from the CD spectr um. If we assume that all the base pairs of rat DNA in a4ueous solution are in the "B" form. and that all the base pairs of rat DNA in 90% ethylene g lyco l a re in the "C" form, th en we can ascribe a per cent "B"-form conformation to the nucleic acid of the nucleoprotein samples. Sin ce protcin•does not contribute to the CD s pectrum of D~A above 250nm. we make use of ..iq1 5 nm in making the assignment of conformation . Our only reservation in making these ~ssig nrncnts is !hat we do not know what effects, if any, light scatteri ng may have on the circul:!r dichroism spectrum of c hromatin (Dorman and Maestre, 1973). The hctcrochromatin frJction (Pl) s howed cons iderably more light scattering than either euchromatin or unfractionated chromatin (AJ2o/Aioo?. 0.1 for hcterochromatin; Ai :o/A21,o <0.02 for cuchroniatin and unfractionatcd chromatin) . It i, clear that th e CD spc.:t rum of heter0chromatin does not fit either th e "B"- o r "C '-form ., pcctra above 28:i nm; this tailing of the positive CD b:.ind to lont!cr wavckngth, is mos t likely due to ligh t scatle ring (Do rm ;i n and Maestre. 1911) . Thus. we do not know to w hal deg.rec uur at;sign mcnt of per cent "B .. con - formation to hctc,ochron1e11i11 D;\A is innu •,·ncc·d bv scattering artd'acts; nc vcrt hcl (·,s. the Cllnforrr.ati,ms of D N ,\ 1n the chromatin fractions arc ck:1rl 1 diffcr~nt (Pol.,c,,w and Simpson, 1973) . Conclusions In this study, we have conce rned ourselves with the biophysical basis of difrcrenti :il ,t:iining of chromosomes with quinacrine. We have chose11 lo 11urk with fractionated intcrphasc hetcroc hromatin and eu chroma tin . We have shown that the quantum yield of quin:icrine bound l•.l ci1her unfractionatcd chromatin or hctcn.1L·hn 1111:1t in i!-1 ~ignificant ly grea ter lh~tn the quantum yi dd or quin:1,:ri ni: bt,u nJ lt) ,:i1hcr cuL·hnm1a ti11 or ON ,'\ in aqut:nus ·"nlutin ,1. DL"protcini1:nion :Jbl)lishcs thi s dif- 2944 13 I O C ll I' ~I IS ·1 RY, v O I. , I 3. e r al. ferc·n tial nuorcscence : furthermo re, no differences in DNA T/\ l!U: 1v : CD and Nu,·lcic Ac id Confornialion. DNA, aqueous solut ion DNA, 90~~ ethylene glycol Chromatin Euchroma tin Hetcrochrornatin GUTTESFELD NO . I 4, I 9 7 4 Wl'rc: found bct-.H·cn th (· euchrom ali n ar~d hctcroc·hrnrnatin fr;i ctions. These rc.,u its c:t s t doubt on whether the banding pattern s sc·cn with quinacrinc-staincd chromosomes a rc solclv a refk.:tion of intrachromos0mal differences in DNA base c.omposi tion. It must be stressed . however. th at this work does 1101 dispute the findi11gs of Wci sb lum and de Ha :,eth (1972) and Pa,hmann and Rigler (19n) pertaining to the in ,,i1ro base speci fici ty of qu ina c rin e nuoresccnce with purified nucleic ac id s. In fact, the very s harp (and very narrow) nuoresccnt b:inds see n on the Y chromosome of Drosophila (Vosa , I 970) co uld indicate the loc.:alization of (A + T)-r,ch sa tellite DNA . Quin ac rine bands are generally thought lo be locali zed in t he hcte rochromatic regions of chromosomes (Vasa~ 1970; Adkis.son er al .. 197 1; Gagne et al.. 1971). Our results s uggest that the conformation of the DNA in the heterochromatic porti on of chromosomes is altered from that of the "B" to the ·•c· fo rm. Chrom os0111:1 I proteins, in particul a r the histones. are thought to induce and maintain thi s change in nucleic acid conform ,Hi o n (Simpson , 1972; Shih and Fasman, 1972; Henso n and Waikcr, i 970; Simpson a nd Sober, I 970). Fur t hermore . our resu lts s ugges t th a t quinacrine bound to hetcrochrl,matin is highl y fluorescent while quinacrine bound t0 euchroma tin _is only weakly fluorescent. The altern a tion of hcterochromat,c ar.d euchromatic regions aiong the chromatics of mclap hasc chromosomes would, therefore, lea d to fluorescence ba nding p~ltcrns. Caspcrsson er al. (1972) have shown that the mi1otic chromoso mes of several cell types of the sam e organi s m have the same banding patterns. Thus, banding docs not renect the cenetic activity of cells per se. More likely, banding reflect s the ~rdcrcd pack agi ng of DNA into the chromosome. On e of the mos t rcmukable fe a tures of chromosome banding techniques is the similarity of band s produced by a variety of methods . . Lee er a l. ( 1973) have identified four groups of techniques which produce similar patterns: these arc ( 1) quinacrine tluorcscencc; (2) Giemsa in combination with alkali-heating techniques; (3) Giemsa in combination wi th ,iny one o f several proteolytic enzymes: and (4) Gicmsa in combination wit h protein-den a turing s ubsta nces such as 5 M urea a nd a nionic and nor,ionic detergents. Unti l recently, it was thought that the Giemsa - a:kali h~ating technique (number 2 above) rcnectcd the differential ch romosomal loca li zation oi cla sses of repetitious a nd nonrcpetitious DNA. Comings er al. ( 1973) a nd Stockert and Li sa nti ( 1972) have ,how n th at in _,i1r1 renaturatior\ of chromosomal DNA (repetitive and sing le copy) is com plete withi n a few mi~utcs . Thus, preferenti al rea ssoc iation of repetitiou s DNA cannot explain the bandin g ;:,atlcrns produced with G ic.s ma. C,,mings et al. ( 1973) have alsn shown that the G-ba11cing tc-.: hn iq ues rrn, ovc ,-cr v 1i1tle [):'\,\ ur protein from the chromo,,omes. H.ir:, h treatment of chromosomes with a lkali or prol,rngeJ c.xrosurc to proteolytic enzymes abolishes G b.1r.ds :iltocclhcr. The rcs~lt s of G-b:inding techniyue,; en,!'l ovi~g mild treatment with protco1 1 tic c n?ymcs and detergent s tend to suppo rt the idea that chromos-,mc banding ratterns do n,it reflec t intrachromosonwl Ui ffrrcn ,;cs in D:\.A b~tsc r:1tio . Thc~c agents (en/\'mcs and dcnaturants) prcsur:wbl y act only on the rrot cin comro~cnt of .:hronios,;mcs The fact that these techniques give ri~c to b::rnc.!inr, !'):.t'tcrns wh! ch arc si;ni l::i;- to t_ho~c o~- tained ,,. rth quinacrine s uppo rt th e co nclusion th a t va r1.1t1 ons 1n protein D~:\ interactions along the chrnma1ids o f rnclaphase chromo.,omcs arc n:spon\iblc for banding . Rcci:ntl ; . Rodman and T:thili a ni ( 1973) h:I\~ shown thal 1·c11lgcn staining of mouse chromosome., rewa h a bandincZ pattern sir,1ilar to that -139\1ECIIANISM OF QUINACRINE STAINI NG obtaintd with Gicm,a o r q ui nac rine-m ustarci . These au thors p,istulate that " the lncali za ti o~ of h ul gcn d~rk a nd li gh t st.tin, representing rela tive D N,\ dcnsitit:,, renccts the reg io~a l protei n association of the D NA." In conclusion. we suc!£est that chromosome banding with qu ina crin e rcnects differences in prot ein - DNA intera cti ons, a nd D:'<'A conforma tion. along the ch rom:ll ids of mcra ph:ise chrom osom es. Acknowledgments Help ful d isc uss ions with Dr. Angeli ne Douvas, Mr. Willi am Pca rsor. , a nd Or. C hris Skidmore arc gratefully ac knowl edged. We al so wish 10 tha nk Dr . S. C. R. Elgi n for generously supplying us with Drosophila chromatin, Mr. Mahlo n W ilkes for performing the electron microscopy, and Mr. Ralph!- . Wilson for the computer programs. References Adkisson, K. P., Perreault. W . J ., and Gay, H. ( I 9i I) , Chromosoma 34, 190 205. Billing. R. J .. and Bonner, J. ( 19 72), Biochim . Biophys. Acta 281, 4 53-562. Blake, A .. a nd Pca cocke, A . R. (I 968). Biopolymers 6, I 22 51253. Bonner, J., Garra rd . W. T. , Gottes fcld , J .. Billing, R . J ., and Uphouse . L. ( 1974), Me1hods En::ymol. (in press). Bonner, J ., Garra rd , W . T. , Gottcsfrld, J., Hol mes. D.S .. Sevall. .I . S., and Wilkes, M . (1973), Cold Spring Jlarhor S y mp . Quan /. Biol. 3/i, (i n press ). Ca spersso n, T .. De la Cha pelle, A., Schroder, J ., and Zech, L. ( I 972) , Exp. Cell Res. 72, 56 -5 9. C aspersson , T ., Farber, S., Foley, G . E .. Kud ynowski , J ., Mod est, E. J ., Simon sson, E., Wagh, V ., and Zech, L. (1968), Exp. Cell Res. 49, 21 9-22 2. C aspersso n, T ., Zech , L.. Modest, E. J ., l'oley, G . E., Wagh , V., and Simonsson, E. ( 1969), Exp . Ceil Res. 58, 128--140. Comings, D. E. Avelino, E, Okada, T. A., a nd Wyandt, 1-1. E. • ( I 973), Exp. Cell Res. 77, 469-493 . Davis , R. W., Simon, M .. a nd Davidson, N. (1971), Methods Enzymol. 21D, 413-428. Dorman , B., and Maes tre , M . F. (1973) , Proc. Nol . A cad. Sci. U.S . 70. 255 - 259. Elgin. S . C. R .. an d Hood, L. ( I 973) , Biochemis1ry 12, 49844991 . Ellison , J. R., a nd Barr, H. J . ( 1972). Chromvsoma 36. 375 390. Gagne, R., Tang uay. R .. a nd Labe rge , C. (1971), Nature (London) , Nee, H,o/. 23:. 29- 30. Gottesfcld, .I . M . a nd Bon ner, .I . (IG74). in rr~parati on. Gottcsfr!d, J. M ., Garrard. W. T., Oag i, G .. Wil son, R. F .. a nd Bo nn•:r, J. (19 7-1), Proc. Na/. Acod. S ci. U.S. ( in press). Helene. C.. Di .nicol i, J .- L. , a nd Brun, F. ( 197 1). /Jiochemi.11r;• 10, 3802 -3809. Hen so n. P., a nd Walker , I. 0 . (19 70), Eur. J. Biochem . / 6, 524 - 53 I. Johnso n, R. S ., Cha n, A., a nd Hanlo n, S. ( 1972), BiochemiSlry // , 4347-4358. Lee. C. L. Y., Welch. J .P. , and Lee, S. H . S. (1973), Nat ure (London), Ne>i· Biol. 241, 142 - 143. Lcrm;i n, L. S. ( 196 3 ). Proc. Na/ . A cad. Sci. U. S. 49, 94 - IO I . Lomh olt, B.. a nd Mohr, J . ( I 971 ), Na/U re (London), Sew Biol. 23-1. I 09- 1 I 0. Mandel, M .. and Marmur, J. ( 1968). Methods Enzymol. 12B, 195-206. Marus hi gc, K., a nd Bonner, J. ( I 966). J . Mo/. Biol. 15. I 60174. Marvin, D. /\ ., Spencer, M., Wilken s. M . H . F .. a nd Hamilton , L. D. ( I 96 I). J . M o/. Biol. 3, 547 -5 65. Modes!, E. J., a nd Sengupta, S. K. (1973) , in Chromosome Identification. Nobel Symposium 23, Caspcrsson, T .. a nd Zech, L. , Fd ., New York, N. Y. , Aca demic Press, pp J2733 3. Nelson, R. G ., a nd Johnson. W. C. ( 1970), Biochem . Biophy s. Res. Commun . 41, 211 - 216. Pachm a nn, U., and Rigler, R . ( I 972) . Exp. Cell Res. 72. 60 2- 608. Polacow , I., a nd Simpson. R. T . ( 1973), Bioc~em. Biophys. Res. Commun. 52. 202-207. Rodma n, T. C. , and Tahiliani, S . ( I 973) , Ch romosoma 42, 37-56. Scatchard, G . ( I 949), Ann. N. Y. Acad. Sci. 51, 660-672. Schreck . R. R., Warburton , D .. Miller. 0 . J., Be iser, S. M ., and Erlanger, 8. F. (1973), P.roc. Na1 . Acad. Sci. U.S. 70, 804 - 807 . Shih, T . Y., and Fasman, G . D. (1970). J . Mo/. Bio l. 52, 125129 • Simrson , R. T. ( 1970), Biochernis1ry 9, 48 14- 4819. Simpso n, R. T . ( 1972). Biochemis1ry I I . 2003 - 2008. Simpson, R. T ., a nd Sober, H. A . ( 1970) , Biochemistry 9, 3103 - 3109. Stockert, J . C., and Lisa nti , J . A. (1972), Chromosoma 37, 117-130. Tunis -Schneider, M. J .. and Maest re, M . F. ( 1970), J. M o /. Biol. 51, 521 -541. Vasa, C. G. ( 1970) . Chrom osoma 30, 366 -3 72. Weisblum . B.. a nd de H aseth, P. L. ( 1972), Proc. Nat. Acad. Sci . U.S. 69, 6 29-63 2. BIOCHF. MISTRY, VOi 13. N O . 14 . 1974 2945 -140- FURTHER DISCUSSION Since this publication appeared, Comings et al. (1975) have presented results obtained from a similar study of the interaction of quinacrine with polynucleotides and chromatin fractions. In agreement with our results, Comings et al. observed that chromatin fractions, isolated by differential centrifugation of sonicated chromatin, differ in their ability to quench quinacrine fluorescence. DNA prepared from the various chromatin fractions quenched quinacrine fluorescence to the same extent. Equilibrium dialysis experiments suggested that the chro- matin fractions differed in their ability to bind quinacrine; the poorest binding fractions were enriched in nonhistone proteins. In agree- ment with the observation that acid treatment (histone-extraction) does not alter Q-banding of metaphase chromosomes, histones had little effect on the quenching of quinacrine fluorescence by DNA. Agents that cause DNA to undergo the B ➔C transition (salts, ethylene glycol) also prevent quinacrine binding and fluorescence quenching. Comings et al. have concluded that the most pale staining regions in metaphase chromosomes are due to a decreased binding of quinacrine and that this inhibition is predominantly due to nonhistone proteins. Our results are in general agreement with those of Comings et al.; however, we concluded that quinacrine banding was due to protein-DNA interactions and DNA conformation. Our data did not indicate which class of chromosomal proteins -141- might be responsible for quinacrine staining. In light of the results of Comings et al., it appears that the proteins responsible for banding are the nonhistones. Furthermore, the results of Comings et al. cast doubt on our hypothesis of the relationship between DNA conformation and quinacrine fluorescence. It seems more reasonable that poorly staining regions of chromosomes should bind less dye than the brightly staining regions. Nevertheless, our results and those of Comings et al. show that DNA-protein interactions are central to the mechanism of quinacrine staining of chromosomes. -142- REFERENCE Comings, D. E., Kovacs, B. W., Avelino, E., and Harris, D. C. (1975) Chromosoma 50, 111-145.