4917-4921. M., and Frank, J. (1985) Chromosoma (Bert.) 91, 377-390. U. S. A. 82,2642-2646. Chauvin, F., Roux, B., and Marion, C. (1985) J. Biomol. Struct.
Val. 261, No. 15, Issue of May 25, pp. 7044-7051,1986 Printed in U.S.A.
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1986 by The American Society of Biological Chemists, Inc.
Chromatin Structure NUCLEASE DIGESTION PROFILESREFLECTINTERMEDIATE STAGES IN THE FOLDING OF THE 30-NM FIBERRATHERTHAN THE EXISTENCE OF SUBUNIT BEADS* (Received for publication, September 30, 1985)
P. Roy Walker$, MariannaSikorska, and JamesF. Whitfield From the Cell Physiology Group, Divisionof Biological Sciences, National Research Councilof Canada, Ottawa KIA OR6, Canada
Conditions have been found for the isolation of rat liver nuclei which maintain chromatin in its native state andsuppressesendogenousnuclease activity. Chromatin prepared in this way can be dispersed into buffers containing various concentrationsof monovalent or divalent cations so that the30-nm fiber is either totally or partiallydecondensed. Whenprobed with micrococcal nuclease, the digestion profiles show that although the 30-nm fiber can be cleaved periodically to generate superbead-like particles this only occurs under certain ionic conditions when the fiber is partially decondensed. It is likely that this cleavage pattern reflects the transient exposure of specific nuclease sensitive sites as the 30-nm fiber condenses, rather than theexistence of a specific subunit of a beaded 30nm fiber. The periodicity of these nuclease-sensitive sites appear to be related to the asymmetric distribution of histone H1 molecules along the length of the fiber.
Nucleases, principally micrococcal nuclease (MNasel) which played a critical role in the elucidation of the subunit structure of the 10-nm fiber, are currentlybeing used as probes of the structure of the 30-nm fiber (12-19). These studies have shown that the fiber can, under certain conditions, be digested into a discrete class of particles variously described as nucleomers or superbeads (14, 18) that may represent subunits of a beaded 30-nm fiber. However, there appears to be no consensus on the actual structure of these particlesand, indeed, there is no consensus on the actual number of nucleosomes per particle. Much of this inconsistency and controversy may be due to the fact that the different approaches that have been used to study the fiber may, because of the limitations that these techniques impose on the ionic environment, actually change the structure of the fiber either during isolation of the material or during subsequent analysis. In thisreport we have critically examined the use of micrococcal nuclease as a probe of the structure of chromatin in interphase rat liver nuclei. The data show that the ionic environment in which the nuclei are isolated has a great It is well established that most of the DNA in theinterphase influence over the pattern of digestion by altering the strucnucleus is organized into a fiber of approximately 25-30 nm tures of the 30-nm fiber during preparation of the nuclei. This (see Refs. 1 and 2 for reviews) and that thisfiber is generated is particularly noticeable when nuclei are isolated in the by folding and compaction of the 10-nm “beads-on-a-string’’ presence of the low concentrations of monovalent cations (25 polynucleosome chain. Although the structure of the 10-nm mM) that are used in virtually all conventional nuclear isolafiber is nowwell understood, it is still not clear how it is tion procedures. Moreover, the concentrations of ions in the organized into the30-nm fiber. A number of physical studies, digestion mixture, particularly those of divalent cations, are carried out on chromatinin solution, have lead to theproposal critical since they not only affect the structure of the fiber, that the larger fiber is generated by some form of smooth but also the solubility of the released material. When the winding of the 10-nm polynucleosome chain into a fiber of 30 ionic environment, and hence the degree of condensation of nm in diameter. However, there is considerable disagreement the 30-nm fiber, is carefully controlled, it is apparent that on whether the polynucleosome chain is wound as a solenoid although the fiber can be cleaved by the nuclease in a periodic (3, 4), a helical ribbon ( 5 , 6), or any of a number of other manner it is likely to reflect transient changes in fiber structureratherthan provide evidence for the existence of a similar structures (7-9). Although electronmicrographs that are considered to sup- specific subunit structure of the 30-nm fiber. port these models have been generated, they invariably show MATERIALSANDMETHODS that the 30-nm fiber is irregular or “knobby” in appearance (3, 10, 11).This observation is difficult to reconcile with the Treatment of Animals and Isolation of Nuclei-The animals used above models which would be expected to generate a smooth in this study were males (190-210 g) of a specific-pathogen-free strain fiber of constant diameter. Unfortunately, although the elec- of Sprague-Dawley rat bred in this laboratory. The rats were entronmicrographs do reveal such periodic discontinuities along trained to a 12-h light/dark cycle with the dark period being from 7 p.m. to 7 a.m., and food and water were available ad libitum. Nuclei the length of the 30-nm fiber, they do not indicate anyspecific were prepared as described previously (20) following homogenization underlying subunit structure. of the tissue in 6 volumes of 0.25 M sucrose containing 50 mM Tris*National Research Council of Canada Publication No.25287. The costs of publication of this article were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ T o whom correspondence should be addressed.
HC1, pH 7.5, 150 mMKC1, 5 mMMgC1, and 0.2 mM phenylmethylsulfonyl fluoride. This same buffer was used throughout the nuclear isolation procedure. In some experiments the concentration ofKC1 and/or MgCl, in this isolation buffer was varied as indicated in the figure legends.
7044
The abbreviations used is: MNase. micrococcal nuclease.
Digestion Nuclease MicrococcalNucleaseDigestions-Nucleiwere resuspended a t a concentration of0.8-1.0mg DNA/ml in 10 mM Tris-HC1, pH 8.5, containing varying concentrations of monovalent and divalent cations as described in thetext. Micrococcalnuclease (Sigma or Worthington) a t a concentration of 50 units/ml (1 unitcorresponds to a change of optical density a t 260 nm of 1.0 at 37 "C, pH 8.0) was added and the nuclei digested at 30 "C for the times indicated. The reaction was terminated by adding 0.2 M EDTA to a final concentrationof 10 mM followed byrapid cooling in ice. The suspension was then centrifuged at 25,000 X g for 15 min in the Ti-50 rotor of a Beckman L8-70 ultracentrifuge at 4 "C to generatea supernatant, SI, containing released, soluble chromatin and a pellet, P1, containing unreleased material. Density Gradient Analysis of Released Chromatin ParticZes-Profiles of polynucleosomes were obtained by layering 0.5 ml of S1 supernatant on a 10-35% (w/w) sucrose gradient (prepared in50 mM Tris-HC1, pH 7.6, containing 150 mM KCl) followed by centrifugation in a SW 40 rotor a t 40,000 rpm for 195 min (o't = 2 X 10") at 4 'C. The gradients were fractionated using an ISCO model 185 gradient fractionater connected to an ISCO UA-5 absorbance monitor and fraction collector. 0.7-ml fractions were collected and used for either DNA size or protein profile analysis. Electrophoresis ofDNA Fragments-The fractions from sucrose density gradients were dialyzed overnight against 10 mM Tris-HC1, pH 7.8, + 1.0 mM EDTA using a Bethesda Research Laboratories mini-dialysis chamber to remove sucrose which interferes with the DNA extraction. The samples were then deproteinized by digestion at 30 "C for 2 h with 0.5 mg/ml of proteinase K in 10 mM Tris-HC1, pH 7.8, containing 5mM EDTA and 0.5% sarkosyl. DNA was isolated and purified by successive extractions with phenol, phenol: ch1oroform:isoamylalcohol, chloroform, and finally ether as described by Maniatis et al. (21) followed by precipitation with isopropanol. Electrophoresis was carried out in 0.8% agarose gels in 40 mM Tris acetate buffer, pH 8.0, containing 2 mM EDTA (21) and 0.3 Ng/ml of ethidium bromide a t 100 V for 3 h. Gels were visualized and photographed immediately after electrophoresis using a CAMAG transilluminator. Extraction and Electrophoresis of Proteins from Gradient Fractions-Gradient fractions were made 20% (v/v) in trichloroacetic acid and stored at 4 "C overnight. The precipitated proteinswere pelleted and washed once with 90% acetone containing 0.1 N HCl and then with acetone before being dissolved in sample buffer and loaded onto 15% (w/v) polyacrylamide, 0.1% sodium dodecyl sulfate gels. After the electrophoresis the protein bandswere visualized by staining with Coomassie Blue. Chemical Measurements of Total Protein and DNA-Samples of S1, P1, or gradient fractions were extracted for total protein or DNA measurements using the Lowry and diphenylamine reagents, respectively, as described previously (22).
of Chromatin
7045
i I
I
I
0
20
IO
30
50
40
60
DIGESTION TIME (minutes) d
v,
0
t
I
0
I
I
I
I
,
L "
2 3 4 5 0 50 100 MAGNESIUM POTASSIUM CONCENTRATION(mM)
I50
FIG. 1. The concentrations of monovalent and divalent cations in the isolation buffer influence the activity of endogenous nuclease(s) during subsequent incubations. In A, nuclei were isolated in homogenization buffer containing either 25 or 150 mM K+ (in thepresence of M%+).In B, the buffer contained 25 mM K+ and the indicated concentrations ofMg", whereas in C it contained 5mM M e and theindicated concentrations of K+. Digestions were carried out in the absence of exogenousnuclease for the indicated times in A or for 5 min in B and C, and the percentage ofDNA released measured as described under "Materials and Methods."
When the monovalent cation concentration of the isolation buffer was increased (in the presence of 5 mM Mg2+),the nuclease activity was further diminished, and at concentrations 2100 mM it was essentially inactive. Furthermore, even though some endogenous nucleases are dependent upon ea2+ and/or Mg2+ (23), the addition of these cations to digestion mixtures (up to 5 mM in the case of Mg2+and/or 1 mM for Ca2+)did not stimulate this suppressed nuclease activity (data RESULTS not shown). Therefore, if chromatin is maintained in a conSuppression of Endogenous Nuclease Activity-Liver nuclei densed state by isolating nuclei in buffers containing physiohave particularly high levels of endogenous nuclease which logical levels of monovalent cations (2100 mM K+) endogecomplicate the interpretationof structural studies using other nous nuclease activity can be effectively suppressed and will nucleases. For example, when nulcei were isolated in the not interfere with subsequent digestions using MNase. presence of a low concentration of monovalent cation (25 mM Effects of Ions on the Morphology of Isolated Nuclei-The is the concentration used in nearly all conventional nuclear morphological appearance of nuclei isolated in buffers of isolation procedures) and digested at low ionic strength in the various ionic composition was examined in the light microabsence of added nuclease, virtuallyall of the DNA was scope under phase contrast optics. When 25 mM KC1 was released (Fig. L4). The curve was triphasic suggesting that chromatinfractions with differingsensitivitiesto the nu- used in the isolation buffer, a M$+ concentration of 5 mM clease(s1 were being digested. However, when the ionic was required to generate morphologically intact nuclei. Howstrength of the isolation buffer was increased to 150 mM (with ever, when the monovalent cation concentration of the isolaKCl) there was virtually no release of DNA even when nuclei tion buffer was increased to 150 mM a good yield of morphowere incubated at low ionic strength for as long as 60 min logically intact nuclei could be obtained at Mg2+concentrations as low as 0.1 mM. These nuclei were somewhat swollen (Fig. 1A). Fig. 1, B and C demonstrate that the concentrations of both with a "homogeneous" nucleoplasm. As the concentration of monovalent and divalent cations in the isolation buffer af- Mg2' in the isolation buffer was increased from 0.5 to 2.0 mM fected the sensitivity of the chromatin to the endogenous nuclei with increasingly granular nucleoplasm were obtained, nuclease(s1. Magnesium ion concentrationsgreater than 2 At a physiological concentration of 5 mM, the nuclei had a mM lowered the endogenous nuclease activity, rendering it very granular nucleoplasm indicating that much of the chrocapable of digesting only 15-18% of the DNA at 5 mM matin remained highly condensed during isolation. These magnesium (thestandard isolation buffer concentration). observations are generally similar to those made by Olins and
7046
Digestion Nuclease
Olins (24) in theirstudies onthe effects of cations onisolated nuclei. The Ionic Composition of the Isolation Buffer Determines the Overall Sensitivity of Nuclei to MNase-Since the ionic composition of the isolation buffer so obviously affected the gross morphology of the nuclei, it was expected that it would also affect the sensitivity of the chromatin to exogenous nucleases. Fig. 2 shows that when nuclei were isolated in the presence of 150 mM KC1 and 5 mMMgC1, and digested with MNase inthe presence of the same ionsso that thechromatin was maintained in a highly condensed state then only 5% of the DNA was released even after 60 min of digestion. In contrast, if nuclei were isolated in the presence of 150 mM KC1 and 0.5 mM MgClz and digested in theabsence of added cations so that thechromatin fibers were dispersed then DNA was rapidly released, with 70% being released within 10 min. However, the digestion reached a plateau after 20 min, and even when the reaction was continued for 1h, approximately 20% of the chromatin remained in the nucleus. Since it is conceivable that some cleaved material may not be able to escape from the nucleus, particularly at high ionic strength, we attempted to extract material from such nuclear pellets by resuspending them in 20 mM Tris-HC1 + 10 mM EDTA, followed by a second spin. However, no significant further release of chromatin could be obtained indicating that there was no trapping of cleaved material. Furthermore, if nuclei were isolated with highly condensed chromatin (150 mM K'/5 mM M F ) and digested in the absence of added cations then thechromatin rapidly dispersed and was readily digested (Fig. 2, dashed line). Thus, theionic environment that the nuclei are exposed to during isolation has a critical impact on the nature of the chromatin fibers and the accessibility of this material to nuclease probes. In subsequent experiments in thisstudy, nuclei were isolated in the presence of 150 mM K' to suppress endogenous nuclease activity and in thepresence of different divalent cation concentrations to manipulate the degree of chromatin condensation. The Influence of Cations in the Digestion Buffer on the Sensitivity of Chromatin to MNase-Some constraints are placed upon the use of nucleases as probes of chromatin structure by the fact that they generally require a divalent cation for optimal activity and the presence of this divalent cation may affect the structure of the substrate and/or the solubility of the released material. For example, MNase is a
z r
g 100-
a
g
80-
W -!
j,
-
< ;*-
a
/$."
-
8
5 20oi .jL""
.. GI
o/o
20;
1
20 40 DIGESTION TIME (rnlnules)
€0
FIG. 2. The concentrationsof monovalent and divalent cations in the digestion buffer influence the sensitivity of chromatin to micrococcal nuclease.Digestions were carried out for the times indicated on nuclei isolated and digested (in the presence of 50 units/ml of micrococcal nuclease) in buffer containing the following 150/5 -+ 150/5; c"., K+/M%+concentrations (mM): A-A, 150/0.5 -+ O/O; O---O, 150/5 -+ O/O, where the numbers preceding the arrow refer to K+/MgZ+in the isolation buffer and the numbers after the arrow refer to cations in the digestion buffer.
80
e
0
(7 If
2 40
I
-
a F-cT'; ?
:
5
10
calcium-requiring enzyme ( 2 5 ) , and the presence of this ion in reaction mixtures may influence the pattern of digestion (the fact that the enzyme was active in the absence of any added Ca2+in the experiments in Fig. 2 may be attributed to the contamination of commercial preparations of the enzyme by low concentrations of calcium (25)). We explored the effects of adding low concentrations of Ca" (0.1 mM) to digestion mixtures in which eitherthe chromatin was dispersed (no othercations present) or highly condensed (150 mM K+/5 mM Mg"), and the resultsare summarized in Fig. 3. Allthe nuclei used in these experiments were isolated in the presence of 150 mM K+ and 5 mM Mg. As expected, 0.1 mM Caz+ gave a 6-fold stimulation of the already rapid rate of DNA release when the digestion was carried out at low ionic strength (O/O in Fig. 3) with 60% of the DNA being released in 2 min. However, during prolonged digestion the released material appeared to be processed to a Ca2+-insolubleform and was precipitated. Thus, after 60 min only 20% of the DNA remained in solution. Calcium ions also gave a 2-fold stimulation of the rateof digestion of condensed chromatin (150/5 in Fig. 3), but the reaction soon reached a plateau and less than 20% of the total DNA was liberated. This material did, however, remain soluble in 0.1 mM Ca2+. Clearly, the pattern of digestion is affected by the concentration of calcium ions. Short-term digestions in the presence of low concentrations of the cation may be satisfactory, but prolonged digestions in the presence of 20.1 mM Ca2+ are likely to result in theselection of specific chromatin fractions that can remain in solution, and these may not accurately represent chromatin fibers in situ. It is also noteworthy that the 10 mM EDTA used to stop the digestion in this, and virtually all other published studies (which are usually carried out in the presence of 1 mM Ca2+),could not solubilize the calcium-precipitated material. Density Gradient Profiles of Material Released by MNaseWe first examined the profiles obtained from nuclei prepared with condensed chromatin but digested under conditions in which the 30-nm fiber would disperse (Fig. 4A) in theabsence of added Ca2+.Under these conditions 40% of the DNA was released inthe 5-min incubation, andthis material sedimented over a wide range of molecule sizes, with some of the material being pelleted. Therefore, when digestions were carried out on dispersed chromatin oligonucleosomes ranging in size from 2 to >50 nucleosomes were released indicating a random cleavage of the chromatin fiber. When digestions were continued for 60 min, virtually all the chromatin that was released (82.3% of total nuclear DNA) was processed to mononucleosomes and shortoligonucleosomes,although there
e 6080 - #
v
0
of Chromatin
1;
'x,,
0
""_ l+l
' X b .
'
10
20
XI
,
,
40
M
150/5
:Iy
- 60 -80 - 40
,o,o'd"Q
0 IO 20 DIGESTION TIME (mmulesl
60
;
(+I
0
-20 0
" " " " " "
, 30
40
(-)
, 50
:-o 60
FIG. 3. The effect of calcium ions on the amounts of DNA released by micrococcal nuclease.All the nuclei were isolated in the presence of 150 mM K+ and 5 mM M%+ and digested with micrococcal nuclease as described under "Materials and Methods" either in the absence of added K+/MgZ+(left panel)or the presence no added Ca2+; of 150 mM K+/5 mM Mg2f (right panel). U, 0"-0,0.1 mM CaClZ.The percentage of DNA released was measured as described under "Materials and Methods."
7047
Nuclease Digestionof Chromatin
o/o
I
-co2+
I
50
d
rn
I
3)
: c)
10
“I
R L
. B
0
‘.
DIRECTION OF SEDIMENTATION
30 W m
-
I-
rn D
FIG. 4. Sucrose density gradientprofiles of S1 supernatant
20
material releasedby micrococcal nuclease. All the nuclei were isolated in the presence of 150 mM K+ and 5 mMM$+ and digested for either 5 min (solid line)or 60 min (dashed line)in the absence of added K+ or M e . 0.1 mM Ca2+ was either absent (left panel) or present (right panel)in the digestion buffer. Gradient centrifugation and fractionation were carried out as described under “Materialsand Methods.”
U J
6 -
z
”I 0
10
-
2
I was some evidence of a shoulder in the 10-12 nucleosome 3 DIRECTION range (Fig. 4A). 0 50 100 150 OF In thepresence of 0.1 mM ca2+ (aconcentration of calcium SEDIMENTATION POTASSIUM that greatly stimulates the activity of MNase (cf. Fig. 3), but CONCENTRATION (mg) is likely to have only a minimal impact on chromatin strucFIG. 5. Sucrosedensitygradient profiles of material reture), approximately 80% of the total DNA was released in the 5-min digestion (Fig. 4B). About half of this material was leased by micrococcal nuclease at different monovalent cation Nuclei were prepared in buffer containing 150 mM released as monomers and shortoligomers, and about half of concentrations. K+ and 5 mMM$+ and digested with micrococcal nuclease for 5 min it sedimented as a broad peak centered around10-12 nucleo- in the presence of the indicated mM KC1 concentration. The leftpanel somes. When theincubation was continued for 60 min, much illustrates the gradient profiles and theright panel the percentage of of the higher molecular weight material precipitated (Figs. 3A DNA released during the 5-min digestion. and 4B) leaving only mononucleosomes and subnucleosomal material in solution. above indicated that condensed chromatin was relatively reThe shoulder on the gradient profile obtained when chro- sistant to nuclease attack, we attempted to probe the conmatin was digested rapidly (5-min digestion in the presence densed fiber with MNase by prolonging the digestion times of (=a2+)suggested that if the 30-nm chromatin fiber was and/or stimulating its activity with low levels of calcium ions. attacked by the nuclease while it was dispersing in the low The results are summarized in Fig. 7 for salt-condensed ionic strength digestion buffer, then the shoulder may repre- chromatin and Fig. 8 for magnesium-condensed material. Low sent anunderlying subunit component of the fiber. To inves- concentrations of Ca2+(0.1 mM) produced a 3-fold increase in tigate the existence of such a component further, we carried the amount of chromatin released during the first 5 min of out digestions on nuclei in which the chromatin was prepared digestion (Fig. 7, left panel), and this material sedimented in a condensed form and then allowed to reach different with a profile that was similar to that obtained for partially intermediate levels of condensation in the presence of either decondensed chromatin (Fig. 7, center; cf. Fig. 5). If the monovalent (Fig. 5) or divalent cations (Fig. 6). Fig. 5 shows concentration of Ca2+was increased to 1.0 mM, the activity the profiles for 5-min digestions carried out at five different of the enzyme was further stimulated, and although therewas K+ concentrations.When the chromatin was either fully no further increase in the amount of DNA released during dispersed or fully condensed (i.e., in the presence of 0 mM or the 5-min incubation period most of it was converted to 2100 mM K+, respectively), the nuclease appeared to cut the mononucleosomes and shortoligomers. chromatin fiber at random indicating that there was no unWhen digestion was continued for 60 min, chromatin was derlying subunit structure. However, at intermediate salt con- released even in the absence of added Caz+ with a similar centrations the chromatin showed an elevated sensitivity to profile to that obtained in the shorter digestions with Ca2+ the nuclease (40 and 60 mM K+, Fig. 5B) and superbead-like ions (Fig. 7, right panel). A 60-min digestion in the presence profiles were obtained. Similar results were obtained when of calcium rendered virtually all the released material to divalent cations, a t much lower concentrations, were used to mononucleosomes and subnucleosome species. maintain chromatin at intermediate levels of condensation Five-min digestions of magnesium-condensed chromatin in (Fig. 6). the presence of0.1 and 1.0 mM calcium released material These results indicated that during condensationof the 10- with similar profiles to those of salt-condensed chromatin nm fiber to a 30-nm fiber, and its subsequent compaction, it (Fig. 8). Furthermore, a 60-min digestion in the presence of folds in such a way that a number of sites arranged periodicalcium ions converted all the material remaining in solution cally along the fiber are exposed and become nuclease sensi- to mononucleosomes or subnucleosome species. However, a tive. As condensation proceeds these sites abruptly become 60-min digestion of magnesium-condensed chromatin in the less accessible and thefiber is, once again, cleaved randomly. absence of added calcium ions produced material (10-15% of Nuclease Studies on Condensed Chromatin-Although the total DNA) with a larger average size than thetypical superresults of the condensation/decondensation studies described bead profile.
-
7048
Nuclease Digestionof Chromatin
magnesium-condensed chromatin, there seemed to be a relatively sharp transitionbetween tetranucleosomes and pentanucleosomes. Tetranucleosomes appeared to be ina more open configuration, and therewas extensive cleavage between the monomers, whereas pentanucleosomes (and higher oligomers) were more stable. Proteins Associated with ReleasedParticles-Acrylamide gelsof proteins extracted from gradient fractions of either salt-condensed or magnesium-condensed chromatin are shown in Fig. 10. The mononucleosomes and small oligomers released from both preparations were depleted in histone HI, whereas the polynucleosomes at thepeak of the profiles were enriched in H1. Furthermore,although many nonhistone proteins, including proteins with electrophoretic mobilities similar to those of high mobility group proteins, HnRNP particle proteins, and many higher molecular weight species were released, virtually all of them remained at the top of the gradient indicating that neither the H1-depleted oligonucleosomes nor the H1-enriched polynucleosomes have significant numbers of these nonhistone proteins associated with them. This was further demonstrated in Fig. 10 (lower panel) which shows that virtually all of these nonhistone proteins were released in the absence of any significant digestion of DNA. DIRECTION I I I 0 I 2 It appears, therefore, that thereis a labile structure containing OF the ribonucleoprotein particles and many nonhistone proteins SEDIMENTATION MAGNESIUM including high mobility groups 1,2,14, and17 which is rapidly CONCENTRATION (mg) FIG. 6 . Sucrosedensitygradientprofiles of material re- released when nuclei are incubated at the elevated temperaleased by micrococcal nuclease at different magnesium ion ture used in these digestions. concentrations.This experiment was carried out asdescribed in the Histone HI-depleted Nucleosomes-We examined in more legend to Fig. 5 except that the indicted mM MgClz concentrations detail the conditions under which histone H1-depleted or H1were used in the digestion buffer instead of monovalent cation. free nucleosomes were released in an attempt tofind out if a particular ionic environment selected for these particles. Fig. It appears, that the condensed 30-nm fiber can be probed 11shows acrylamide gels of S1 supernatantsfor 5-min digeswith MNase revealing an underlying periodicity. However, tions (tracks 1 and 2) carried out on dispersed chromatin in only 10-30% of the total DNA is accessible indicating that the presence or absence of Ca2+ (cf.Fig.3A). Histone H1most of the condensed fiber is organized into a nuclease containing materialwas released in thepresence and absence inaccessible form. of Ca2+, although there was a great stimulation of release in Size Analysis of the DNA in Particles Released by Digestion the presence of Ca" ions. However, after 60 min of incubation of Condensed Fibers-The distribution of DNA in thegradient (tracks 3 and 4) only H1-free nucleosomes remained in solufractions of 5-min digests of either salt-condensed (Fig. 9A) or magnesium-condensed (Fig. 9B) chromatin was analyzed tion when the digestion mixtures contained 0.1 mM Ca2+.In by agarose gel electrophoresis. In both cases most of the contrast, when Caz+was not present all the nucleosome and released DNA was present in the fractions in the middle of polynucleosome material remained in solution. Similar results the gradient and was 2000-2500 base pairs in lengthwhich is were obtained when condensed chromatin was used (tracks 5equivalent to 10-12 nucleosomesjparticle. This material was 10) with H1-free nucleosomes being selected for during proessentially intact, although in the case of the salt-condensed longed incubations. Theseresults indicate thatHI-containing nucleosomal chromatin some of the DNA was cleaved in between a small percentage of the nucleosomes, but this did not lead to dis- monomers or oligomers appear to be initially soluble in buffers ruption of the particle. A relatively high proportion of the containing Ca2+ions but are eventually rendered insoluble. It smaller oligomers were cleaved between the nucleosomes in- is not clear how the H1-containing nucleosomes are processed dicating that the linker DNA in these oligomers was exposed to the Ca2+-insolubleform, nor whether there is a subset of and much more sensitive to the nuclease. In the case of H1-free nucleosomes that are initially released and remain in
-
FIG. 7. Densitygradient profiles of the products of digestionof saltcondensedchromatin by micrococcalnuclease. Nuclei were isolated in the presence of 150 mM K+/5 mM Mg2+ and digested in the presence of 150 mM K+ with or without the indicated mM concentrations of Ca2+. The left panel illustrates the time course of digestion in the presence or absence of 0.1 mM Ca2+and the center and right panels indicate the gradient profiles of material released after 5 min and 60 min of digestion, respectively.
9 ;; 0
5
''
-0 Iuo
0
85 t-
m
c n l
60-
k! 5
2
OFl
-.
t+J
-
gr" sz 2 2
2 0 - ! /"---(-1
-
- 4
n
c'
40-
*-
,-e' I
-. ""__