Deoxyribonuclease

To probe sequence-specific chromatin structure of SV40 minichromosome, we further modified previously described hybridization mapping. Actually (i) the digestion patterns by two nudeases 1 micrococcal and DNAase II) were compared and (ii) the kinetics of nuclease digestion was analyzed from early time points when only a fraction of minichromosomes was cleaved once to longer digestions when oligo-and mono-nucleosomal bands appeared. DNAase II is shown to possess certain sequence specificity different from that of microcoo-cal nuclease. The major finding is that DNAase II preferentially cleaves the SV40 minichromosome at a distinct region of the genome known as 72 bp modulator element. Other hyper-sensitive sites are located near the replication origin and T-ag binding site II and also near BamHI site where termination of replication and "late" transcription occurs. Micro-coccal nuclease splits the Bgll-Hpall region in a different manner.

To probe sequence-specific chromatin structure of SV40 minichromosome, we further modified previously described hybridization mapping.Actually (i) the digestion patterns by two nudeases 1 micrococcal and DNAase II) were compared and (ii) the kinetics of nuclease digestion was analyzed from early time points when only a fraction of minichromosomes was cleaved once to longer digestions when oligo-and mononucleosomal bands appeared.DNAase II is shown to possess certain sequence specificity different from that of microcoocal nuclease.The major finding is that DNAase II preferentially cleaves the SV40 minichromosome at a distinct region of the genome known as 72 bp modulator element.Other hypersensitive sites are located near the replication origin and T-ag binding site II and also near BamHI site where termination of replication and "late" transcription occurs.Micrococcal nuclease splits the Bgll-Hpall region in a different manner.

UfTHODUCTION
Specific chromatin organization of certain genomic regions can be analyzed by hybridization techniques.The indirect end labeling protocol is applicable for studying both the nucleosome positioning and the distribution of hypersensitive sites in chromatin (see /1,2/ for reviews).
Micrococcal nuclease and DNA&BO I used in previous studies revealed a strong sequence specificity when analyzed on pure SNA /3,4/.Looking for a better biochemical probe, we have introduced DNAase II which can act in different ionic conditions.Although this enzyme also turned out to be sequence specific, together with micrococcal nuclease analysis it gave less ambiguous results.We have also extended the indirect end labeling procedure to different stages of nuclease digestion to discriminate between the most preferential sites and those which appear later in digestion.
In this paper, we describe the mode of DNAase II action at very early stages of digestion when several hypersensitive sites appear.The same sites are attacked on pure SV4O DNA with much lower probability.Other results relevant to nucleosome positioning on SV40 DNA will be publushed elsewhere.

ft AND METHODS
Preparative procedures.CV1 monkey kidney cells were grown and infected with SV40 virus (strain Rh 911) as described earlier /5, 6/.Viral mlnichromosomes were extracted from nuclei 36-40 hours postlnfection in a buffer containing 0.1 M NaCl, 10 mM TEA-HC1, pH 7.8, 0.1 mM EDTA, 0.5% NP40, 0.5 mM PMSF /7, 8/.The extraction was performed in a Blaessig homogenizer (2-3 strokes on ice).The overall isolation procedure usually did not exceed 1 hour.Minichromosomes were separated from other nucleoprotein complexes by velocity sedimentation /5/.The virion to mini chromosome ratio at these stages of infection ranged from 1.5 to 2.0 /5, 8/, implying that virion degradation is not occurring, as can be found when lower ratios are obtained.
Minichromosomes from assembled virions were prepared by disrupture of intranuclear viral particles /5-8/ by a procedure modified from Brady et al. /9/-SV4O DNA was prepared according to Hirt procedure /10/.
Micrococcal nuclease ( 15000 u/mg, Worthington) digestion was performed at 4°C or 37°C directly in the sucrose gradient buffer (see above) after Cad, addition to 1 mM.Purified SV4O DNA (~80% of form I) was digested and analyzed under identical conditions.
Mapping of nuclease cuts on SV40 genome.The procedure for hybridization mapping was described previously /11, 12/.In this study, the indirect end labeling was performed for a range of digestion kinetics.Briefly, each sample (corresponding to a different stage of nuclease digestion of minichromosomes) was deproteinlzed by a proteinase-phenol procedure, then purified DNA was redigested by one of the restriction enzymes which cleaved SV4O MA once.Bedigested DNA was fractionated by agarose electrophoresis, transferred onto nitrocellulose filters /13/ and hybridized to DNA probes adjacent to this restriction nuclease sit*.In particular, after EcoRI redigestion EcoRI-BamHI 752 bp fragment was used for clock-wiBe direction and Haelll fragments I, G or J for counterclockwise direction (see fig. 4).Thus, after two subsequent hybridizations of the same gel, mapping in both directions could be done.

RESULTS
DHAase II splits SV40 minichromosoaes in a highly specific Banner SV4O miaichromosomes purified from the nuclear extract late in the Infection were mildly digested with DNAaae II In the conditions favouring its intemucleosomal action (see Materials and Methods).As shown in Fig. 1A, only a part of form I DNA was converted into form III (see legend to Fig. 1).The absence of significant background below form I DNA showed that a fraction of the minlchromosomes cut with DMAase II more than once was negligible.
The double-stranded cuts were mapped after EcoRI re digestion of the purified DNA by hybridization with certain radioactive probes (Fig. 1 B-E).When total SV4O IMA (SV4O probe) was used, two prominent hypersensitive bands appeared which corresponded to 1.57 kb and~3»6 kb long fragments.(Pig. 1 B, indicated by the arrows).In addition, 3-4-less Intensive, but also preferential sites could be observed.
To map these sites, we employed an Indirect end labeling  Some previous studies on nuclease hypersensitivity were performed on mlnlchromosome preparations partially derived from disrupted viral particles.In experimental protocol used here free minichromosomes and assembled virions were separated by velocity sedimentation /5, 7, 8/.We nave probed intravirion mlnichromosomes prepared by a procedure modified from Brady et al« /9/ with DNAase II.In this case, much less prominent sensitivity in ?2 bp repeat can be observed (unpublished observation).
DNAase II cutting of naked SV40 DMA la also non-random Specific cleavage of SV4O UNA in viral minichromosomes could be due to DNAase II sequence specificity.In this case, essentially the same sites would be attacked on naked DNA whereas nudeosomal organization of SV4O minichromosome would only modulate the cleavage probability found on pure DNA.Thus, identical digestion and analysis procedures were applied to SV4O DNA preparations.
We analyzed the location of DNAase II cleavages on naked DNA in the vicinity of the genome region where the major hypersensitive site on the minichromosomes was found.DNA samples at similar stages of the digestion (judged by DNA III to DNA I ratio) were redigested with EcoRI and indirectly end labeled with I or G probe, hybridization pattern (Fig. 2A) did not reveal strong hypersensitivity.All bands labeled by .I-probe are of comparable intensity in contrast to the results obtained on the minichromosomes (Fig. 2 B).Thus, the same region of SV40 genome in the minichromosomes was differently attacked by DNAase II depending apparently on the peculiarities in chromatin structure.
In general, the sites attacked by DNAase II in the minichromosome can be also detected among those attacked on free DNA.However, either some "potential" DNA sites were not observed in the minichromosome digestions or the corresponding bands were much less intensive.
Thus, DNAase II is a sequence specific nuclease with cleavage sites differing from those fa& micrococcal nuclease (see below).

Mlcrococcal nucleaae cleaves the same region in SV40 mlnlchromosomes In a different
We have compared patterns of DNAase II cleavage of the SV40 mlnichromosome and naked SV4O DNA with those obtained after mlcrococcal nuclease digestion.It was shown previously that mlcrococcal nudease cleaved SV4O ENA In both cases In a specific manner.Mapping of the most preferential sites (corresponding to a primary double-stranded cut) showed their distribution through the whole SV4O genome (/10/, see also /14/X The preferential cleavage in the 300-400 bp long segment adjacent to the Bgl1 site was not so evident as in the case of DNAase II.On the other hand, structural alterations in this particular region of SV40 chromatin have been established by a number of biochemical and physical methods /15-20/.Therefore, it was interesting to detect some preferential sensitivity of this "open" region against a background of prominent sequence specificity of mlcrococcal nuclease. It turned out that altered structure was reflected in the cleavage pattern after extensive digestion when each minichromosome had been split several times (Fig. 3).Hybridilation mapping from the Bcofil site revealed 5-6 multiple cleavages through the whole BglI~EpaII segment, most of them being also found on naked DMA (data not shown).Three sites were mapped Inside or immediately adjacent to 72 bp repeat, but in contrast to DKAase II digestion there was no single prominent cleavage.In fact, a site ~1 kb counterclockwise to the EcoBI site was attacked more frequently than those located between the Bgll and the Hpall sites (see Fig. 3).

Nucleates as probes to specific DNA-protein interactions in chromatin
The specificity of chromatin structure at defined regions of the eukaryotic genome has been under intensive investigation during the last several years (see /1,2»21/ for reviews).Two main problems are of particular importance: (i) the peculiar chromatin structure involving the elements which control gene activity (promoters, terminators, modulators, Mapping of micrococcal nuclease cleavages In f th BllHll i p the vTcinTEy of the Bgll-Hpall region.A. Kinetics of micrococcal nuclease digestion (at 37 C) of TEe SV40 minicnromosomes (from the right to the left) after EcoEI redigestlon of the purified DNA.End labeling by I-probe.J. Low resolution gel indicating ENA fragments obtained after micrococcal nuclease digestion of pure SV4O DNA and redigestion with EcoHI.End labeling with F-probe.etc.), and (ii) the specificity of nucleosome location with respect to DNA sequence with its possible (if any) relation to replication and transcription.In both cases, nucleases are usually exploited as primary biochemical probes to chromatia structure.After introducing secondary cleavages into purified DNA by restriction nucleases, these primary cuts could be unambiguously mapped on genomic DNA by the simple hybridization procedure termed as Indirect end labeling /11, 12/.In ideal situation the primary nuclease cleavages would be solely determined by the specificity of the chromatin structure.However, micrococcal nuclease widely used in the study of chromatin has been shown to be highly sequence specific /3,4,11,29/.As shown here, DNAase II which in certain conditions cuts MA between nucleosomes and gives rise to a 200 bp ladder characteristic of micrococcal nuclease /37-39/ is also sequence specific.These two nucleases cut naked SV40 SNA mostly at different sites of the genome (not shown).One could suggest that the cbromatin structure only modulated the probabilities of cutting in those potential sites which are cleaved on naked MA.
This specificity of nucleases complicates the analysis of the nucleosome positioning along certain DMA sequences (see/1-4, 22-29/).On the other band, the use of several nucleases differing in their sequence specificity can give less ambiguous results and is being used for nucleosome mapping on the S740 DNA /36/.
Another type of specific chromatin organization is reflected in hypersensitivity to nucleases of certain genomic regions, in particular, of the upstream sequences for some genes /12,15,21,30,31/.For example, the pattern of hypersensitivity to DNAase I sites has been shown to be specifically changed during differentiation and correlates remarkably well with transcriptional switch and changes in DNA methylation /32,33,3V.
The existence of a DHAase I hypersensitive region in SV40 mini chromosomes had been reported /15,18/.The same region mapped between the Bgll and the flpall sites on the SV40 genome has also been shown to be preferentially cleaved by other nucleases /1b,17/.In this paper, we analyze the cleavage pattern of this genome region in detail, using indirect end labeling of EcoRI redigested DNA.
On the chromatin structure around Bftll-Hpall region as probed by nucleases Previous studies have shown that the region of the SV40 genome comprising approximately 400 bp clockwise to the replication origin has an altered chromatin structure in the viral minichromosome /15-2O/.Among other possibilities, absence of nucleosomes In this region has been suggested /18,19/.Recently we have shown, that labeled H Haelll fragment (mapped for Rh 911 strain ineide the BglI-%aII segment, see legend to Fig. 4) hybridises to oil go-and mononucleosomal bands of nucleate digested SV4O chromatin /5O/.These data, however, do not exclude that a certain (probably, functionally distinct) subfraction of the SV4O mlnichromosomes isolated late in the infection lack nucleosomes in this particular region.In this respect, our most recent data are hardly consistent with the existence of the naked DNA stretch even in a subfraction of SV40 minichromosomes.In fact, there are at least several potential DNAase II specific "sites" inside the Bgll-Hpall segment (Fig. 2).However, only one of them is hypersensitive in the SV40 minichromosome.
From the data presented in Figures 1 and 2, the most prominent cleavage site for DNAaee II can be mapped 1560-1570 bp counterclockwise to the EcoRI site (which is located for Eh 911 strain used here 1711 nucleotides from the Bgll site).Thus, the cleavage occurs ~140-150 nucleotideB clockwise to the Bgll site just inside a single copy of the 72 bp repeat.Another preferentially attacked site can be mapped inside T-antigen binding site II /49/ (see Fig. 4) just near the origin of DNA replication.In this respect, it should be noted that T-antigen binding could Influence the chromatin structure of the whole Bgll-^pall segment.Our preliminary results with strain ??S minichromosomes imply that both copies of 72 bp repeat are equally cut by DNAase II.The cleavage near T-ag binding site II (located in this case 1.8 kb from the BcoRI site) is also observed (data not shown).
The 72 bp sequence has been shown to be an obligatory cis-actinK element for SV40 "early" gene expression in vivo /40,41/ and probably similar to modulator elements found in other genes /42-47A Similar element from polyoma virus has been reported to control replication of the viral DNA /43/.Interestingly, this 72 bp element can be replaced by a non--homologous sequence of 73 bp repeat from the long terminal repeat (LTE) of murine leukemia virus (P.Gruss, personal commun ,and/54/).0nthe other hand, nuclease hypersensitivity in-indicated above 72 bp element.
One could imagine that 72 bp repeat and Immediately adjacent DNA sequence are involved in chromatin attachment to the nuclear matrix to provide proper structural stete of the template (P.Gruss, W.Schaffner and R.Kamen, personal communications).Another interesting possibility has been suggested by Moreau et al. /46/ and postulates the role of 72 bp element as bidirectional entry site for RNA polymerase or some other component of the transcription machinery.We believe that both of these mechanisms would be reflected at chromatin level as some alterations in a regular nucleosome structure.In addition, our most recent results indicate against a direct contribution of 72 bp sequence to DNA superhelicity (in preparation).
We have found that DNAase II action at early stages of the digestion is different to that of micrococcal nuclease or DNAase I.In fact, although multiple micrococcal nuclease cleavages between the Bgll and the Hpall sites could imply enhanced level of sensitivity of the whole 300 bp segment, there is no hypersensitivity inside and around 72 bp element (Fig. 3, 4).As shown recently, DNAase I preferentially cleaves the Bgll-Hpall segment on the SV40 minichromosome at several distinct points, one of them being mapped Inside or immediately adjacent to 72 bp element (/5V, and C.Cremisi cited in /33/)-In this case, however, there are several other sites which are attacked with the same probability.As shown here, DNAase II in contrast to micrococcal nuclease and DNAase I, recognizes 72 bp modulator as a ma.ior hypersensitive site on the SV40 genome.This fact could be due to a very local alterations in chromatin structure which involves only DNAase II potential site.The role of ionic conditions specific for DNAase II digestion is also not excluded.In any case, this remarkable feature of DNAase II could be helpful for studying the peculiarities of chromatin structure in specific regions.We are now applying the same procedure to probe integrated SV40 DNA in transformed cells and other modulator sequences of eukaryotic genome.

12 /
Fig. 1. Specific DNAase II cleavage of isolated SV40 chromosomes.SV40 minichromosomes were purified and mildly digested with DNAase II as Indicated in the Materials and Methods.Isolated SV40 DNA was redigested with EcoRI (B,C,D,E), fractionated in a 1.6J6 agarose 15 cm gel, blotted and hybridized with 32p_i a _ beled DNA probes.A. Mild digestion of minichromosomes with DNAase II.The arrow Indicates the direction from "zero" point (no enzyme added) to later stages of digestion, hybridization with total SV40 MfA (SV4O-probe).The positions of superhelical (I) and linear (HI) forms of SV40 DNA are indicated (form II DNA is not shown).(Note that form I DNA is underrepresented 3-4 fold due to losses during the blotting procedureB .The same DHA samples shown in A, but redigested with BcoEI and hybridized with the SV40 probe.The positions of Haelll fragments and their lengths are shown on the left slot.C. The same as in B, but indirectly labeled by BnmHI-BcoHI 751 t>P fragment.D. The same as C, but labeled by ,1-probe (I fragment of SV40 Eaelll digest, see Fig. 4)."B .The same as D, but after longer exposure.~" See Fig. 4 for the summary of the results.

. 2 .
Mapping of DNAase II cleavages in the vicinity of the Bgll-Hpall genome region for SV4O minichromosomes and bare DNA. A. Mild digestion of naked SV4O MA by DNAase II at rC (from the left to the right).Cleavages revealed after EcoRI redigestion and indirect end labeling with I probe (electro-phoreBis in a 2%, 25 cm long gel).The arrow indicates cleavage inside 72 bp element.B. Mapping of the major cleavage site on the SV4O genome Telectrophoresis in a 1.6J6, 30 cm gel).(1) Naked SV40 INA mildly digested by DHAase II (form III to form I ratio 2.0) and redigested with EcoHI.Labeling with G probe.(2) The same, but for SV40 minichromosomes extracted la£e in the infection.(3) The same as (2), but after longer exposure.early In the digestion with DNAase II Is not excluded.