THE JOURNAL BIOLOGICAL CHEMISTRY OF Vol. 255, No. 22. Issue of November 25. pp. 10702-10709.

1980 Printed m U.S.A.

Structure of Sea Urchin Sperm Chromatin Core Particle*

(Received for publication, May16,1980)

Robert T. Simpson and Lawrence Bergman W. From the Developmental Biochemistry Section, LaboratoryNutrition and Endocrinology, National Znstitute of Arthritis, of 20205 Metabolism and Digestive Diseases, National Institutesof Health, Bethesda, Maryland

H , presence of HMG 14 and 17 appears to be necessary for this 1 (8). It is not tissues. We have delineated some effects of the variant known whether core particle structure per se differs for tranH2A and H2B on chromatin by study of the structure scribed versus nontranscribed gene segments; the observation Strpngylocentrotuspurpumtug that HMG 14 and 17 restore DNase I sensitivity to chromatin of the core particle from sperm. The particle contains 146 base p e s of DNA and or monomer nucleosomes suggests that other features (besides equal amounts of the four smaller histones. It sediments these proteins) of chromatin structure at this level of organiat 11 S and has a circular dichroism spectrum similar zation distinguish functionally distinctive gene segments (9). to that of particles containing more typical histones. Weisbrod et al. have suggested that active nucleosomes may The sperm core particle undergoes a shape change at contain only the H3 + H4 tetramer and HMGs, lackingH2A low ionic strength, observed for chicken erythrocyte as particles. In contrast to these similarities, the melting and H2B (9). Isolation of a population of nucleosomes that contains only active gene segments has not been possible to is quite different from that profile of the sperm particle of erythrocyte; both the reversible transition and the this point, precluding a full description of the structure of core irreversible denaturationof the core particle occur at particles that are transcribable. In this context, it is of interest to assess the roles played by highertemperatures.Thespermcoreparticle is digested by DNase I more slowly than the core particle the various individual histones in packaging and stabilizing DNA in the core particle. Since histone sequences in most from chicken erythrocyte. Cutting site maps for the higher species have been conservedstrongly (lo), this too has sperm core particle reveal the same basic organization been of DNA by these histones; however, certain features of difficult. In contrast to this conservation, major variathe map differ significantly from that for chicken tions in lysine-rich histones occur during early embryogenesis erythrocyte, demonstrating a modulation of the canonical in sea urchin. Variant forms of H1, H2A, and H2B associate core particle structure the unusual histones present withDNA at the cleavage stage, others do so during the by in sea urchin sperm. transition from the 8- to 16-cell embryo to mesenchyme blastula, and these eventually yield to the several adult forms at or beforegastrulation (for a review, see Ref. 11).Variant forms of these histones are also present in sea urchin sperm nuclei The core particle of the nucleosome is recognized as the Hl., is more basicthan other Hls; fundamental building block for the first level of organization (12). The H1-like molecule, of nucleoprotein structure in nearly all eukaryotic cells. In a it has nearly equal contents of lysine and arginine (about 20 wide variety of tissues from numerous species, the core particle mole % each) and a ratio of basic to acidic amino acids of consists of 145base pairs of DNA wrapped around an octamer nearly 1O:l (13). Strongylocentrotus purpuratus H2A. miof the four smaller histones, two each of H2A, HZB, H3, and grates more rapidly on sodium dodecyl sulfate gel electrophoH4 (fora review, see Refs. 1-3). With the exception of studies resis than the adult form; the sequence of the histone for 1 ) More striking variants of core particles containing hyperacetylated histone (4), all Psammachinus miliaris is known ( 0 . are the H2B,sof Parechinus angulosus; compared to calf currently available information suggests the essential constancy of the structure of the complex of the histone octamer H2B, these proteins are about21 residues larger. Many of the and its associated 145 bp of DNA. For the case of hypera- additional residues are arginyls located in a highlybasic, cetylated histones, small variations from properties of core extended NHz-terminal region (14, 15). Study of structure of chromatin of sea urchin sperm should particles containing unmodified histones were noted, a dedetail any structural consequences for core particles deriving crease in melting temperature of less than 1C, alterations and of the nuclease susceptibility of the ends of the core particle from the variant H2A, and H2B,. In addition, since sperm is a transcriptionally inactive cell, study of its chromatin strucDNA and the cutting site about 60 bp from the 5 end of the ture might provide insight into the mechanisms used to preDNA (4). clude transcription of chromosomal DNA. We report here At the level of chromatin, at least, the properties of portions of the DNA differ from those of the bulk DNA. In a now such an investigation of the structureof the nucleosomal core classic example,transcribed genes are present in nucleosomes particle from sperm of S.purpuratus. (5) but are more rapidly degraded by DNase I than gene EXPERIMENTAL PROCEDURES segments which are not active in a given tissue (6, 7). The S. purpuratus were obtained from Pacific Biomarine, Venice, CA, a * The costs of publication of this article were defrayed in part by and maintained at 14C in artificial sea water (16). spawning Ws induced by intracoelomic injection of 0.55 M KCl. Sperm were washed the payment of page charges. This article must therefore be hereby marked advertisementin accordance with 18 U.S.C. Section 1734 by low speed centrifugationin sea water and then in 80 m~ NaC1,20 m~ EDTA, and 10 m Tris/Cl, pH 8.0, and stored as a frozen pellet. M solely to indicate this fact. The abbreviationsused are: bp, base pair@);Hl., H 2 k , and H2Ba, Sperm nuclei were prepared as described by Keichline and Waa~arman (17). Core particleswere prepared by suspendingnuclei at a the sperm variant forms of histones H , 1 H2A, and H2B.

Sea urchin sperm chromatin contains forms of

H2A,and H2B which M e r from those present in adult differential DNase sensitivity of active genes

Downloaded from www.jbc.org at HINARI, on April 23, 2010

10702

Particle

Core

Sperm

10703

DNA concentration of 2 mg/ml in 10 m MgCL, 1 m CaClz, and 10 M M

m Tris/CI, pH 8.0, warming to 37"C, and digesting with staphylococcal nuclease (Worthington) using a concentration of 400 units/ml M for 10 min. Digestion was terminated by addition of EDTA to 25 m and cooling to 0C. Nuclei were pelletted, washed once in 0.5 M NaCl and 25 m EDTA, pH 7, and repelletted. The salt-washed, digested nuclei were lysed by suspending in 0.9 M NaCl and 25 m EDTA, pH M 7, and debris removed by centrifugation a t 2,000 X g for 5 min. Sucrose gradients, isokinetic for a particle with density of 1.51 g/cmJ at 4OC, were prepared with a meniscus concentration of 15% (w/w) M sucrose and contained 10 m Tris/CI and 1 m EDTA, pH 8.0 (18). The gradients were overlaid with a shelf equal to thesample volume of 0.9 M NaCI, 4% (w/w) sucrose. Centrifugation in the SW 41 rotor was at 40,000 rpm for 22 h; gradients were emptied through a flow cell in a Zeiss PM6 spectrophotometer. After dialysis to 0.25 m M EDTA, pH 7, particle preparations were adjusted to contain 0.1 M NaCl to precipitate internally cut nucleosomes, centrifuged, and the soluble material stored at 4C. Chicken erythrocyte (Pel-Freez) core particles were prepared as previously described (19). Histones were extracted from nuclei with 0.4 N H ~ S O I a t4C. dialyzed to water, and lyophilized. Gel analysis of histones from core particles was made by direct solubilization of the particles in sample buffer. Electrophoretic analysis was as described (20, 21). DNA was purified by phenol extraction and analyzed as previously described, using 5% polyacrylamide gels for resolution of native, double-stranded DNA and 12%polyacrylamide gels containing 7 M urea for resolution of denatured, single-stranded DNA fragments (22, 23). Sizing standards for 5%gels were an Hae I11 digest of @X174RF DNA (Bethesda Research Laboratory). When necessary, core particles were labeled at the 5' end using [y-:"P]ATP (New England Nuclear) and polynucleotide kinase (Miles) (23). Rates of digestion and mapping of cutting site susceptibility were done as previously detailed (19, 23). For determination of the kinetics of digestion of total core particle DNA by DNase I, aliquots of a digestion mixture were precipitated at 0C in 0.5 M HCIO, and 0.5 M NaCI, centrifuged, and supernatant absorbances determined. Circular dichroism measurements were made as previously described (24), with a Cary model 61 spectrometer. Thermal denaturation of core particles was measured in 1 m sodium cacodylate, pH M 7.1. Samples with A260 = 0.25 to 0.3 were monitored during heating in a Cary model 219 spectrophotometer equipped with thermal probe accessory. Samples were heated with a heater-circulator driven by a NesLab programmer. Data points were digitized a t 0.2"C intervals with a Numonics Co. graphics calculator, andderivative melting profdes were calculated by linear least square fitting the tangent to the melting curve, using 11 to 16 points around each data point. Sedimentation analyses were performed in a Beckman model E analytical ultracentrifugeequipped with ultraviolet scanner and multiplexer. Observed velocities, analyzed as least square regression lines, had standard deviations of the slopes of less than 1%of the regression coefficient. Data were corrected to 20OC; a l measurements were made l between 19.5 and 20.4"C. The solvent was 0.25 m EDTA, pH 7, plus M NaCl to the indicated ionic strengths, except for the lowest ionic strength which was in 0.08 m EDTA. M
RESULTS

-HI,

H3 H2B' HZA' H4-

-HZB, -H3

-H2As -H4

Downloaded from www.jbc.org at HINARI, on April 23, 2010

FIG. 1. Chicken and sea urchin sperm histones. Sodium dodecyl sulfate gel electrophoresis of ( A ) inner histones of chicken erythrocyte, ( B ) histones of core particle from S. pupuratus sperm, and (0total acid-soluble proteins of S. purpuratus sperm nuclei.

TABLE I Amino acid composition of HI-likehistones

acid sperm Current study sperm

Ozaki (13)
mole %

Calf thymus

LS Y His k g ASP Thr Ser Glu Pro GlY Ala Val Met Leu Ile Tyr Phe

27.8 20.1 0.9 19.5 1.8 3.9 7.4 2.9 10.1 9.3 3.7 20.4 4.2 0.7 2.1 2.0 0.8 0.3

21.2 0.8 19.0 1.8 4.1 8.1 2.8 3.2 3.2 24.4 5.0 1.1 2.1 1.8 0.8 0.4

0.0 2.0

2.7 5.5 6.5 4.2 6.9 23.3 4.9 0.0 4.1 0.9 0.6 0.6

Sea urchin sperm chromatindiffers in three compositional and one known structural feature from most adult chromatins. The histone equivalent of H1 in S. purpuratus sperm (Fig. 1) is more highly basic and contains much more arginine than typical Hl's of other tissues (Table 1).The amino acid analysis of our Hl,, prepared by solubilization in 5% perchloric acid, is nearly identical with that reported by Osaki for the most slowly migrating component on acid-urea gel electrophoresis of S.purpurutus sperm acid-soluble proteins (13).Electrophoretic mobilities of H2A, and H2B, are different from those of chicken erythrocyte H2A and H2B, representative of a variety of adult tissues (Fig. 1). The structure of chromatin in sea urchin sperm is distinguished by the long repeat length for the chromatin subunit structure. The DNA ladder obtained after hydrolysis of sperm nuclei with staphylococcalnuclease corresponds to nucleosome spacing a t about 260 base pair intervals (Fig. 2), in good agreement with the results of others (17, 25).

While the early points in a digestion of sperm nuclei by nuclease yield a monomer DNA length of about 200 bp (mean size), with time, the size of the major portion of DNA in the monomer band becomes about 145 bp (Fig. 2), equivalent to the DNA size in the core particle from other tissues. Identity of coreparticle DNA lengths in spermandother tissues suggests that either 1) H3 + H4 provide the primary forces which determine the length of DNAwrappedaround the histone octamer, as has been suggested (26, 27), or 2) interactions of DNA with the globular portion of the histones, made up of the COOH-terminal two-thirds of the proteins, determines the core particle length; these regions of the histones are closely similar for the four inner histones of sperm and chicken erythrocyte (10). In order to isolate core particlesfrom sperm chromatin, we havehad toresortto removal of the lysine-rich histone equivalent, Hl,, prior to sucrose gradient fractionation. In

10704

Core

Sperm

Particle
two very sharp steps, centered at and 1 mM, when studied 7.5 by diffusion and sedimentation methods(29). In contrast,Wu et al. found a single transition,centered at 1.3 mM, using transient electric dichroismas an experimental method (30). Coreparticles cross-linked withformaldehyde (29) or dimethyl suberimidate (30) did not expand under the stress of lowered ionic strength, suggesting that partial disruption of the quaternary structure the histone octamer necessary of was for the conformational transition.Wu et al. fit their data for the core particle a t low ionic strength with a model particle which is a 178-A diameter disk, 60 A high, with about one turn of DNA wrapped around the histones (30). We have studied this conformational change in sperm core particles, using sedimentation velocity measurements toindicate changes the particle shape. ionic strength is lowered in As from 100 m to less than 1 mM, the sedimentation coefficient 11.0 of the sperm core particle decreases from to 9.5 S (Fig. 4). The transition is smooth, with no indication of the abrupt transitions noted by Gordon et al. (29); rather the data resemble thetransition observedby Wu et al. (30) by electric fit dichroism. Our data are well by a noncooperative transition from a particle with sedimentation coefficient of 11.0 S a t higher ionic strength to one with sedimentation coefficient of 8.2 S a t very low ionic strength, centered at an ionic strength of 1.2 m. The exact nature of the core particle a t very low ionic strength is not known; whatever it in detail, the present is data clearly show that thepresence of variant forms of H2A and H2B in sperm core particles uersus chicken or calf does not preclude the conformationalchange. A second type of reversible conformational change occurs for core particles on heating. Weischet et al. showed that about 30% of the total hyperchromicity developed on denaturation of core particles occurred as a separate transition centered a t about 6OoC in 1 mM sodium cacodylate. The at remainder of the DNA denatured 75-76"C, concurrent with

1 32 64 x IO2 Umin ml" 6

- 4-mer - 3 -mer - 2 -mer

- I-mer
FIG.2. Digestion of S. purpurntus sperm nuclei with staphylococcal nuclease. Sperm nuclei, prepared as described under "Experimental I'rocedures," were digested at 37C with staphylococcal nuclease for the indicated product of time (min) X enzyme concentration (units/rnl). DNA was isolated and electrophoresed on
a X polyacrylamide gel.

Downloaded from www.jbc.org at HINARI, on April 23, 2010

contrast to nuclei from other cells we have studied, where addition of, or dialysis against, EDTA suffices to release core particles and their multimers after staphylococcal nuclease digestion,only nonsedimentingDNAandnohistonesare released from sea urchin sperm nuclei by such treatments. Addition of /3-mercaptoethanol, mechanical homogenization, and even sonication did not effect nuclear lysis and release of core particles. Apparently Hl,, variant smaller histones, or non-histones maintain the intregrity of the sperm nucleus even after extensive DNA digestion. To minimize the length of time, samples are exposed to the salt concentrations necessary for H1, dissociation, and to enhance resolution on the gradients, we have utilized a discontinuous gradient centrifugation method. Samples after digestion were lysed in 0.9 M NaCl, dissociating Hl,, but almost no other protein (in contrast to the resultsof others (28)).The sampleis applied to a sucrose gradient a t low ionic strength which is overlaid by a shelf of 0.9 M NaCl, 4% sucrose. The shelf allows chromatin particles to sediment away from H1, before entering thelower ionic strength of the gradient, where random association of these two species would be expected to occur. Resolution of such gradients is equal to that conventional gradientsused of for preparation of more easily isolated core particles (Fig. 3). The core particles contain DNA of length about 145 bp, with some tailing to longer sizes, equal amountsof the four smaller sperm histones, and no H1, (Fig. 1). The circular dichroism spectrum of the sperm core particle in 0.1 M NaCl is closely similar to that of chicken erythrocyte core particles, with a maximum ellipticity at 282 nm of 1800 A 150"cm2/dmol of phosphate. The sperm core particle is homogeneous on analytical ultracentrifugation and has sedimentation coefficient a of 11.0 S at ionic strength 0.1. In order to assess the role of the variant histones in core particle stabilization, we havestudied two conformational transitions for the sperm core particle. Both conformational changes, induced by lowering ionic strength orby heating, are reversible for other core particles and havebeen suggested as potential candidatesfor conformational changeswhich might accompany chromatin transcription(29-32). As ionic strength is lowered to less than 30 mM, the core particles of chicken erythrocyte or calf thymus undergo a reversible conformational change, unfolding into a less compact structure. Gordon et al. found the unfolding to occur in

- BOTTOM

TOP

FIG.3. Sucrose gradient fractionation of a staphylococcal nuclease digestof sperm nuclei. Sperm nuclei were digested with 400 units/ml of nuclease for 10 min at 37"C, lysed with 0.9 M NaCI, and centrifuged on the gradients described under "Experimental Procedures" for 22 h at 40,000 rpm and 5C. Sedimentation is from right to left.

Sperm Core Particle

II

10705

further analysis. The second and third, at temperatures of 68.5C and 73C, respectively, are due to melting of about 10% and 20% of the DNA.While their sum is similar in magnitude to the premelting transition observed for the erythrocyte core particle, the sperm transitions are displaced to temperatures about10C higher, closer to the temperatureat which the entire structure of the core particle is disrupted.

v)

$ IO i

9
>

/
-3
50

-2 -I LOG (I) FIG. 4. Change in shape of the sperm core particle at low ionic strength. Sperm core particles were sedimented at various ionic strengths, as indicated, at near 20C and a DNA concentration of 50 pg/ml. The points are experimental, and the c w e i theoretical s for a two-state transition with sedimentation coefficients at high and verylowionic strengths of11.0and 8.2 S, respectively, with the transition centered at log ionic strength = 2.9.
disruption of histone secondary structure and presumed dissociation of the histone octamer structure (31). Circular dichroism measurements indicated that thepremelting DNA was altered in conformation about 10C below the temperature at which its strand separation occurred (31). Using hydrodynamic methods, we couldshow a change in particle shape occurring 10C below the temperature of the absorbance melting transition for core particles containing poly(dAdT) (32). Studies of these particles, labeled with 32P the5 at ends, using S1 nuclease, provided evidence that the region of DNA involved in the fist, reversible transition was 20 to 25 bp at each end of the core particle (32). We have investigated the effects of the variant H2A, and H2B, on this conformational change by melting chicken erythrocyte and S. purpuratus sperm core particles in 1m~ sodium cacodylate, pH 7.1 (Fig. 5). The thermal denaturation profile for the erythrocyte particle is quite similar to that previously reported by Weischet et al. (31).A broad transition, equivalent to melting of about 40 to 45 bp of core particles DNA, is centered at about 62C. The remaining melt occurs at about 76C; the magnitude of this phase is 70% of the total hyperchromicity, consistent with melting of 100 bp of core particle DNA. Comparison of this melting profile with that for sperm core particles (Fig. 5) reveals two major differences. First, the major melting temperature for the sperm core particle is about 2C higher thanthat for the erythrocyte particle, being 78.5C. Second, the reversible, first transition is either absent or shifted to higher temperatures for the sperm core particle. Thus, a significantportion of the total melt for the erythrocyte core particle has occurred at 60C, while no increase in absorbance is observed at that temperature for the sperm core particle. We have further analyzed the melting data for the sperm core particle by computing a derivative melting curve (Fig. 6) 2 from the datashown in Fig. 5. The derivative curve is mostly simply resolvedas the sum of four separate transitions (Table 11). The fist is small, occurring at 63.4C and of a magnitude corresponding to melting of less than 2 bp of DNA per core particle. We assume that this is due to short tails of DNA on some of the core particles in our sample and w ignore it in i l

60
Temperaiure

70

80 Downloaded from www.jbc.org at HINARI, on April 23, 2010

FIG. 5. Thermal denaturation profilesof chicken erthrocyte and S. purpumtus sperm core particles. Changes in absorbance with temperature were measured as described under Experimental M Procedures for samples in 1 m sodium cacodylate, pH 7.1 Chicken erythrocyte, - - -; S. purpuratus sperm, .

60

70 Temperature

80

FIG. 6. Derivative plot of the melting profile of the S. purpumtus sperm core particle. The melting profileinFig. 5 was digitized at 0.2Cintervals and derivatives calculated by linear least is square ft to the data using 11-point (2C) data sets around each experimental point.

TABLE I1 Resolved transitions in the meltingprofile of the sperm core Darticle

~~

Transition

Melting temperature

Fraction of total

Base pairs/core oarticle

0.01

I I1
111

IV Calculated assumingthat the fraction of the total melting reflects denaturation of a given portion of the 145-bp core particleDNA in a set of totally homogeneous core particles.

63.4 68.5 73.0 78.5

0.10 0.19 0.69

15 28 100

10706

Sperm Core Particle

SPERM ERYTHROCYTE

The highest temperature transition occurs a t 78.5"C, somewhat higher than that for the erythrocyte particle. As in the case of the erythrocytecore particle, this transitionfor sperm includes about 70%of the total hyperchromicity. It seems likely that this highest temperature transition reflects disruption of the structure the histone octamer and of melting of the 100 bp of DNA in the center of the core particle segment. A higher temperature for the melt in sperm uersus erythrocyte signals greater stability of either histone-histone or histoneDNA interactions in the spermcore particle; we would expect histone-histone interactions tobe similar in the two particles due to the closely similar protein sequences for the two species in the regions of thehistonesthoughtto be involved in protein-protein interaction (10). Interpretation of the second and third phases of the sperm coreparticlemelt is more difficult. Tentatively, we would suggest that these transitions reflect the same process that occurs in the premeltof other core particles, namelydenaturation of about 20 to 25 bp of DNA a t each end of the core particle. In the sperm core particle, the much more highly basic nature of the H2B, might lead to a greater stabilization of these DNA regions; more energy is required to disrupt their interactions with the histone octamer.Part of the attractiveness of this hypothesis arises from the presence in S. purpuratus of two forms of H2B. (Fig. 1).These are present about in a 2:l ratio of the larger or more basic form to the smaller or slightly less basic form. This is the same ratio as the two phases of the premelting transition for the spermcore particle (Fig. 6, Table 11). The two transitions a t 68.5OC and 73C for the spermparticle might result from two types of core particle end segment interactionswith the two types of H2B,, leading to different stabilizations of the end regions of core particle DNA. Overall, the addedstabilization of core particle structure by the variant histones in the sperm core particle is quite significant. It is very nearly equivalent in both melting temperatures and profiles to a 10-fold increase in ionic strength for the erythrocyte particle (cf. Fig. 6 with Fig. 5 of Ref. 31). The two types of investigation of core particle structure detailed above are physical in nature and dynamic terms of in conformational alterations of core particlestructure. The third type of study of the spermcore particle we describe differs in kind from the previous two. Here, we addressthestatic interactions of histones and DNA by mapping the locations and relative susceptibilities of cutting sitesfor DNase I in the nucleosome core particle. Fig. 7 compares the autoradiograms resulting from such amapping experiment for sperm and erythrocyte core particles. Fig. 8 is a scan of the autoradiograms a t 2-min digestion time. General features the cutting of site susceptibilities for typical core particles previously reported (4, 19,23,33-38) are reproduced here for the erythrocyte sample. There is low frequency of cutting a t sites about 30, 60, 80, and 110 bases from the 5' end of the DNA and a general symmetry in the cutting pattern(for convenience,we will denote distancesfrom the 5' end as multiples of 10; other studies have shown that the actualaverage number is nearer 10.4 and that there is an apparent variation in cutting site spacing in two domains of the core particle (34, 37, 38)).The map for the sperm core particle also shows cutting at sites spaced a t multiples of about 10 bases from the 5' end of the DNA. Again, low frequency of cutting is observed for the sites 30, 80, and 110 bases from the end. The site in the middle of the core particle,70 bases fromthe 5' end, which is moderately frequently cut in other core particles, is similar in its rate of cleavage in the sperm core particle. The sites 20 and 40 bases from theend, highfrequency cuttingsitesinother core particles, are relatively infrequently cut in the sperm core

mm

120
90

1 00

70-

50 40

30 -

20 -

Downloaded from www.jbc.org at HINARI, on April 23, 2010

FIG. 7. DNase I mapping of cutting sites in S. purpumtus sperm and chicken erythrocyte core particles. 5'-32P-Labeled core particles were digested with 200 units/ml of DNase I for the indicated times, DNA purified, denatured, and electrophoresed on 12% polyacrylamide gels containing 7 M urea. DNA sizes indicated are approximate distance in bases from the 5 end, assuming 10 bases ' between bands. The figure is a negative of the autoradiogram.

particle, about as infrequently as the site 30 bases from the end (Fig. 8). The basic wrapping of DNA in the core particle, leading to DNase I nicking sites a t 10 base intervalsalong the strands, appears to be common for sperm and other adult tissues previously studied. However, the detailsof the DNase I cutting siteavailability within the core particle suggest that the presence of H2A, and/or H2B, in the spermcore particle core leads to a modulation of this structure compared to particles from other tissues. Comparison of the autoradiograms for sperm and erythrocyte core particles (Fig. 7) makes apparent the fact that the rate of disappearance of the 145-base DNA is slower and that there is less total radioactivity in the resolved bands for sperm than erythrocyte at any of the time points in the digestion. This indicates a general decreased susceptibility to the nuclease of the sperm core particle uersus the erythrocyte core particle, but might also indicate an increased frequency of cutting at the site10 bases from the 5' end; fragments of this size are only marginally insoluble ethanol precipitation and on hence are underrepresented on electrophoresis of extracted samples. T o examine this question, we have measured cutting rates at the site bases from the 5' end of core particle DNA 10 for sperm and erythrocyte particles by determining the rate of solubilization of the label after digestion with DNase I and precipitationwithperchloric acid/sodium chloride. Fig. 9 shows the results of the experiment. Initially, hydrolysis of the end labelto acid solubility occursabout 2-fold more slowly for the spermparticle than for that from erythrocyte. Laterin the digestion, rates become more closely similar (not shown). Thus, in addition tospecific sites within the central region of the core particlebeing less susceptibleto hydrolysis by DNase 10 I for sperm uersus erythrocyte, the site bases from the end of nucleosomal DNA is also less accessible to this nuclease.

Sperm Core Particle

10707

FIG. 8. DNase I map of cutting sites in S. purpumtus sperm and

chicken erythrocyte core particles. Densitometric scans of the 2-min digested autoradiograms i Fig. 7 are disn played. The ordinate i linear with optis cal density.

n
Downloaded from www.jbc.org at HINARI, on April 23, 2010

12

0 9

6 5 4 BAND NUMBER

the 20- and 40-base sites (Fig. 7 and 8), cuts which lead to acid-soluble DNA when combined with cuts at the highly susceptible sites 10 or 50 bases from the end. Together, these alterations in cutting may explain the DNase I resistance of total DNA in the sperm core particle.

In attempts to address the role played by different inner histones in organization of the structure of the core particle, S P SPam most studies have involved reassembly of nucleoproteins containing partial complements of the histones, followed byanal0 ysis by physical or chemical means (26,27,39,40). While the 0 1 2 3 4 0 1 2 3 4 Trne .min TIME, mm information obtained in such studies must eventually be acFIG. 9 (left).Kinetics of DNase I cutting at the site 10 bases commodated by modelsfor nucleosome structure, they suffer from the ends of core particles. 5-3*P-Labeled core particles from from the obviously artificial nature of the complex under chicken erythrocyte (- - -) and S. purpuratus sperm () were scrutiny. Study of the core particle from sea urchin sperm has digested with200 uNts/ml of DNase I for the indicated times and the advantages in assessing features of the role of H2A and H2B fraction of the radiolabel madesoluble in 5%CCbCOOH determined. ul FIG. 10 (right). Kinetics of DNase I digestion of total DNA of in nucleosome structure; the particle contains a f l complecore particles. Core particles from chicken erythrocyte(- - -) or S. ment of the inner histones, has H3 and H4 which differfrom purpuratus sperm () were digested for the indicated times with their counterparts in chicken erythrocyte by only a single 200 units/ml of DNase I and the fraction of the total A m made amino acid residue each (41, 42), and yet contains H2A and soluble i 0.5 M HClO.,, 0.5 M NaCl determined. n H2B which differ from their peers in adult urchin tissues or other species. Since H3 and H4 are essentially identical in sequence for the Since H3 + H4 are thought to form the overall structural two types of particles, we assume that these differences result framework for the nucleosome (26, 27, 39, 40), it is not surfrom differences in histones H2A and/or H2B between the prising that a number of features of the sperm core particle are similar to those of other tissues, specifically chicken erythtwo types of core particles. Fig. 10 details the rates of digestion of bulk sperm and rocyte. Superimposed on this framework of similarity with erythrocyte core particle DNA to acid solubility. At early other core particles, some features of the structure of sperm phases of digestion,the sperm particle total DNA is solubilized core particles apparently are modulated by the presence in hs about 3- to 4-fold slower than that from erythrocyte. Ti is the histone octamer of the variant hiatones H2& and H2B,. expected in view of the decreased frequency of cutting at the One of the most striking differences is the rateof digestion by site 10 bases from the end (Fig. 9), a cut which leads alone to DNase I of sperm core particles relative to that for erythrocyte acid-soluble DNA, and the decreased frequency of cutting at particles. The initial rate of digestion of the sperm particle
4

DISCUSSION

lo

10708

Sperm Core Particle

Derivative analysis of the sperm melt reveals that threemain transitions are present (Fig. 6, Table 11); the largest corresponds to melting of about 100 bp of DNA per particle, similar tothe major transition for other core particles. The two smaller are in the proportion of 2:1, similar to the ratio the of two forms of H2B,in S. purpuratus sperm (Fig. 1). The simplest interpretation of these data is that the temperature for the Fist phase of the melt, unwinding and denaturation of the 20- to 25-bp tails at the ends of the core particle DNA, is increased in the sperm core particle to a temperature nearly sufficient to disrupt the structure of the entire core particle. The two portions of the first unwinding transition might then correspond to two types of stabilization of these end segments by the two different H2B, species. Certainly, the increased stability of histone-DNA interactions near the end regions of the core particle in sperm (versus erythrocyte) detected by melting analysis is very consistent with the differences in DNase I digestion rates of this region for the two particle types (Fig. 9). There are two interesting parallels between certain properties of the sperm core particle and the chromatosome, a particle containing about 160 bp of DNA and, in addition to the inner histone octamer, a molecule of H1 orH5 (19). Both are more slowly degraded by DNase I than erythrocyte core particles. The chromatosome appears to totally lack the first transition conformational change on heating (19);in the sperm core particle, stable histone-DNA interactions lead to a significant increase in the energy required to disrupt the association of the ends of core particle DNA with the histone octamer. We have previously described features of a plausible mechanism for transcription of DNA bound to histones (19, 32,44); one essential element of the proposal was unwinding of the ends of core particle DNA from the octamer to allow the beginning of transcription into a nucleosomal DNA segment. Such unwinding is apparently precluded in the chromatosome by the presence of H1 or H5; evidence suggesting that chromatin containing H1 may be transcriptionally inactive has been summarized previously (19). We now !ind that the unwinding conformational transition is much less easily achieved in sperm core particles than in others; sperm is essentially inactive as a template for RNA polymerase (45). While only correlative, the data demonstrate two ways through which the unwinding conformational transition can beblocked or inhibited and suggest further studies which might elucidate the role of this change in core particle structure in the transcription of chromatin DNA.
Acknowledgments-We are grateful to Mrs. Bonnie Richards for preparation of the manuscript and Dr. David Wright for performing to the amino acid analysis. REFERENCES 1. Kornberg, R.D. (1977) Annu. Reu. Biochem. 46,931-954 2. Felsenfeld, G. (1978) Nature (L0nd.j 271,115-122 3. McGhee, J. D., and Felsenfeld, G. (1980) Annu. Rev. Biochem. 49, 1115-1156 4 Simpson, R. T.(1978) Cell 13,691-699 . 5. Lacy, E., and Axel, R. (1975) Proc. Natl. Acad. Sci. U. S. A. 72, 3978-3982 6. Weintraub, H., and Groudine, M.(1976) Science 193,848-856 7. Garel, A., and Axel, R. (1976) Proc. Natl. Acad. Sci. U.S. A. 73, 3966-3970 8. Wekbrod, S., and Weintraub, H.(1979) Proc. Natl. Acad.Sci. U. S. A. 76,631-635 9. Weisbrod. S.. Groudine, M.. and Weintraub, H.(1980) Cell 19, 289-301 10. Isenberg, I. (1979) Annu. Reu. Biochem. 48, 159-191 11. Newrock, K. M., Alfageme, C. R., Nardi, R. V., and Cohen, L.M. (1978) Cold Spring Harbor Symp. Quant. Biol. 42,421-431 12. Easton, D., and Chalkley, R. (1972) Exp. Cell Res. 72,502-506
'

appears to be nearly 4-fold slower than that for erythrocyte. Digestion of core particle DNA to acid solubility by this nuclease requires nicking at two adjacent sites 10 bases apart, except at the site 10 bases from the ends. The site at 10 bases is one of the most frequently cut in studies of detailed susceptibilities of core particle DNA to DNase I (35,36). The rateof digestion of end-labeled sperm particles by DNase I is 1.5 to 2 times less than the rate erythrocyte particles. Hence, the for regions near the ends of the core particle DNA are less accessible to the nuclease in the sperm particle than in the erythrocyte nucleosome. Since the rate differences are notthe same for endlabel and total DNA, other siteswithin the core particle must also be less readily digested by this nuclease in the case of the sperm nucleosome. Inspection of the cutting site map (Figs. 7 and 8) confirms this supposition. Except for central region, 60 to 70 bp from the 5' end, all sites in the sperm particle are cut at lower frequency than in other core particles. Particularly, the sites 20 and 40 bp from the ends, highly accessible in most core particles, are much less frequently cut in the sperm particle. This is readily seen by comparison of the relative intensities of bands 2, 4, and 5 for sperm and erythrocyte; they are nearly equal in intensity for erythrocyte and differ markedly for sperm. While it is tempting to equate directly these differences in cutting site frequency withthe differencesin H2Aand H2B for the two types of particles, the effects in the sperm particle are so global that direct interaction of the histone variants with the affected cutting sites may be a naive interpretation. A more general alteration in histone-DNA interactions or unknown features of nucleosome structure may underlie the observed differences. The differences in cutting site frequencies for sperm versus erythrocyte apparently do not arise from differences in the DNA for the two preparations, since maps obtained for complexes of these two sets of inner histones with poly(dAdT) .poly(dA-dT) are closely similar to those for the native particles? Whatever the distinctive features of the sperm core particle that lead to differences in DNase I digestion from other core particles, they likely are due to the NHz-terminal regions of H2A, and H2B,. For the known sea urchin sperm histone sequences the NHz-terminal regions differ from the corresponding calf histones; the COOH-terminal two-thirds of the molecules are highly conserved (10, 14, 15). Since these COOH-terminal segments are thought to be involved in the interactions which lead to theproper quaternary structure of the histone octamer (lo), one expects the protein structure of the globular portion of the nucleosome core to be similar for sperm and other nucleosomes. Support for this expectation derives from the expansion of the sperm core particle at low ionic strength. Similar histone-histone interactions in the sperm and erythrocyte core particles may allow themto expand in similar fashion at low ionic strengths. In contrast to the above conformational alteration in core particle structure, a second conformational transition for the nucleosome is markedly dissimilar for sperm and erythrocyte core particles. A reversible melting of 30% of the DNA occurs in heating core particles; at ionic strength about 1 m~ it is centered at about 60C (31). Studies of core particles formed from erythrocyte inner histones and poly(dA-dT).poly(dAdT) (34)have allowed us to postulate a mechanism for this conformational change (32,43). The differences in melting between erythrocyte and sperm core particles are quitestriking (Fig. 5). The first transition of the erythrocyte particle is not present for the sperm nucleosome; the melting of the sperm core particle is also stabilized by about 2-3OC in the major, irreversible thermal transition.

Downloaded from www.jbc.org at HINARI, on April 23, 2010

* L. W. Bergman and R. T. Simpson, unpublished observations.

Sperm Core Particle

10709

13. Osaki, H. (1971) Deu. Biol. 26,209-219 30. Wu, H-M., Dattagupta, N., Hogan, M., and Crothers, D. M. (1979) F., and von Holt, C. Biochemistry 18.3960-3965 14. Strickland, M., Strickland, W. N., Brandt, W. 31. Weischet, W. O., Tatchell, Van K., Holde,E., K. and Klump, H. (1977) Eur. J. Biochem. 77,263-275 15. Strickland, W. N., Strickland, M., Brandt, W. F., and von Holt, C.(1978) Nucleic Acids Res. 5,139-160 (1977) Eur. J. Biochem. 77,277-286 32. Simpson, R. T. (1979) J. Biol. Chem. 254, 10123-10127 33. Finch, J. T., Lutter, L. C., Rhodes, D.,Brown, R. S., Rushton, B., 16. Tyler, A. (1953) Biol.Bull. (Woods Hole) 104,224-239 17. Keichline, L. D., and Wassarman, P.M. (1977) Biochim. Biophys. Levitt, M., and Klug, A. (1977) Nature (Lord.)269,29-36 Acta 475, 139-151 34. Simpson, R. T., and Kunzler, P. (1979) Nucleic Acids Res. 6, 18. McCarty, K. S., Jr., Vollmer, R. T., and McCarty, K. S. (1974) 1387-1415 Anal. Biochem. 61,165-183 35. Noll, M. (1977) J.Mol. Biol. 116, 49-71 36. Lutter, L.C. (1978) J. Mol. Biol. 124,391-420 19. Simpson, R.T. (1978) Biochemistry 17,5524-5531 20. L&tourgeon, w., and R w h , H. (1973) Arch. Biochem. Bbphys. 37. Bryan, p. N., Wright, E. B., and o h , D. E.(1979) Nucleic Acids 155,144-158 Res. 6,1509-1520 21. Whitlock, J. P., Jr., and Simpson, R. T. (1976) Nucleic Acids Res. 38. Lutter~L. c. (1979) Res. 6, 41-56 39. 3,2255-2266 Soher-Webb, B., Camerini-Otero, R. D., and Felsenfeld, G. 22. Maniatis, T., Jeffrey, A., and van deSande, H.(1975) Biochemistry (1976) Cell 9, 179-193 14,3787-3794 40. Boseley, P. G., Bradbury, E. M., Butler-Browne, G. S., Carpenter, B. G., and Stephens, R. M. (1976) Eur. J. Biochem. 62,21-31 ' and Whitlock, J' ", Jr' (1976) ' 9 347-353 41. Strickland, M., Strickland, W, N., Brand& W. F., and Halt, C. 23' Simpson? R' T , 24. Simpson, R. T., and Sober, H. A. (1970) Biochemistry 9, 3103(1974) FEBS Lett. 40,346-348 3109 42. Brandt, W. F., Strickland, W. N., Morgan, M., and von Holt, C. 25. Spadafora, C., Bellard, M., Compton, J. L., and Chambon, P. (1974)FEBS Lett. 40, 167-172 (1976) FEBS Lett. 69,281-285 43. Simpson, R. T., Shindo, H. (1979) Nucleic Acids Res.7,481and 26. Camerini-Otero, R. D., Sollner-Webb, B., and Felsenfeld, G . 492 (1976) Cell 8, 333-347 44. Simpson, R. T., Stein, A., Bitter, G. A., and Kunzler, P. (1980) i n 27. Bina-Stein, M., and Simpson, R. T.(1977) Cell 11,609-618 Novel ADP-ribosylationa of Regulatory Enzymes and Proteins 28. Kunkle, M., Longo, F. J., and Magon, B. E. (1978) J. Exp. Zool. (Smulson, M., and Sugimura, T., eds) pp. 133-142, Elsevier/ 203, 371-380 North Holland, New York 29. Gordon, V. C., Knobier, C. M., Olins, D. E., and Schumaker, V. N. 45. Johnson, A. W., and Hnilica, L. S. (1970) Biochim. Biophys. Acta (1978) Proc. Natl. Acad. Sci. U. S. A .660-663 75, 224,518-530

Downloaded from www.jbc.org at HINARI, on April 23, 2010