The alternating DNA sequence d(CA/GTI) is known to adopt the left-handed Z-form in negatively supercoiled DNA in vitro. This element represents a significant ...
The EMBO Journal vol.5 no.7 pp. 1727 -1734, 1986
DNA conformation and chromatin organization of a d(CA/GT)30 sequence cloned in SV40 minichromosomes
A.Rodriguez-Campos, M.J.Ellison1, L.Nrez-Grau2 and F.Azorin Unidad de Quimica Macromolecular del C.S.I.C., Universidad Politecnica de Catalunya, 08028 Barcelona, Spain, 'Department of Biology, Massachussets Institute of Technology, Cambridge, MA 02139, USA, and 2Departamento de Gendtica Molecular, Centro de Investigaci6n y Desarrollo del C.S.I.C., 08034 Barcelona, Spain Communicated by J.A.Subirana
The alternating DNA sequence d(CA/GTI) is known to adopt the left-handed Z-form in negatively supercoiled DNA in vitro. This element represents a significant fraction of the highly repeated DNA sequences found in eukaryotic genomes. We have cloned an alternating d(CA/GT)30 sequence into SV40 minichromosomes at the unique Hpall restriction site which occurs in the transcriptional leader region of the viral late genes. By comparing the linkage differences of topoisomers obtained from viral DNA with or without the d(CA/GT)30 insert, at various stages of the lytic cycle, we conclude that this sequence does not predominate in the Z-form in vivo. Furthermore, we find that the d(CA/GT)30 sequence is packaged into nucleosomal core particles and the region of the minichromosomes which contain the d(CA/GT)30 sequence is organized as nucleosomes. Key words: d(CA/GT)n elements/Z-DNA/SV40 minichromosomes/DNA supercoiling/nucleosomes
DNA in vivo exists as chromatin. Here we have cloned a d(CA/GT)30 repetitive sequence into the Simian virus 40 (SV40), which proliferated in the form of a minichromosome in lytically infected cells. We then addressed the question of whether the cloned d(CA/GT)30 element adopts the left-handed Z-DNA conformation in SV40 minichromosomes. Formation of Z-DNA within the d(CA/GT)30 sequence will change the linkage of the two DNA strands that will ultimately affect the degree of supercoiling of the purified SV40 DNA. Since we did not find any change in superhelicity associated to the presence of the d(CA/GT)30 sequence, we conclude that this sequence is not stable in the Z-conformation in SV40 minichromosomes. Furthermore, we show that the sequence itself is packed into nucleosomes.
Results 'In vivo' DNA conformation of the d(CA/G7)30 sequence An alternating d(CA/GT)30 sequence was cloned at the unique HpaH site present in the SV40 genome (see Materials and methods). Conservation of the repeated sequence was determined by restriction endonuclease mapping and DNA sequencing of the recombinant virus SV40/CA30 (data not shown). The unique HpaII site in SV40 is localized in the leader region of the Sma I
SV40
BanmHl
Xba I
SV40
4 m * CGCCGACGGATCCCGGGATCTCTAGAGTCGGCT ...
SV40/O
...
Introduction DNA in solution can adopt several structural conformations (Zimmerman, 1982). Z-DNA is a left-handed conformation of doublestranded DNA which is favoured in sequences containing alternating purine and pyrimidine residues (Wang et al., 1979). Under physiological ionic conditions, left-handed Z-DNA is in a higher free energy state than right-handed B-DNA. Several factors are known to stabilize Z-DNA, including negative supercoiling and protein binding (Rich et al., 1984). In particular, d(CG/GC)n and d(CA/GT)n repetitive sequences have been shown to form Z-DNA when in a closed circular negatively supercoiled molecule (Klysik et al., 1981; Peck et al., 1982; Haniford and Pulleyblank 1983a; Nordheim and Rich, 1983). Non-repetitive sequences can also form Z-DNA through negative supercoiling as judged by anti-Z-DNA antibody recognition (Nordheim et al., 1982b; Azorin et al., 1983; DiCapua et al., 1983; Stockton et al., 1983). Anti-Z-DNA antibody binding can stabilize Z-DNA (Revet et al. , 1984; Lafer et al., 1985b). It is likely that other specific ZDNA binding proteins would have a similar stabilizing effect. Alternating d(CA/GT)n sequences are widely found in eukaryotic genomes (Hamada and Kakunaga, 1982). The copy number of this repeat element ranges from 100 in yeast to several thousands in higher eukaryotes (Hamada et al., 1982). These repeated elements have the potential of forming left-handed Z-DNA and they actually do when cloned into negatively supercoiled plasmids (Hamada et al., 1984). However, eukaryotic
-T SV4/A
Taq I
6)
Acc I (1628)
-
© IRL Press Limited, Oxford, England
Fig. 1. General organization of the recombinant viruses SV40/O and SV40/CA30. The DNA sequence of the region surrounding the d(CA/GT)30 insert is shown.
1727
A.Rodriguez-Campos et al. B
.*
0~~~
DNA
SITION
T=o
U
Io
_
*
~0
4i *
0
*' D-O
T = + 107 \DEPRMATN
/ DEPROTEINZATION
T = -24
T= i-24 +1-10.71 -34.7
1ez
Fig. 2. Schematic representation of the experimental rationale designed to determine the DNA conformation of the d(CA/GT)30 is the linking difference.
text for details.
sequence in
SV40/CA30. See
r
late mRNAs (Tooze, 1981). Cloning a repeated d(CA/GT)30 sethis site does not seem to have any strong effect on the viability of the virus. The HpaII site occurs at the border of the SV40 control region which is known to be nucleosome free in a sub-population of SV40 minichromosomes (Saragosti et al., 1980; Jakobovits et al., 1980). In addition to the
quence at
d(CA/GT)30 sequence, SV40/CA30 also carries a few additional base pairs originating from the plasmid vector used during the cloning steps. These sequences are also present in SV40/O, which does not contain the d(CA/GT)30 element. The sequence of these regions, in addition to the general features of these constructs, are shown in Figure 1. In all our experiments we used SV40/O as control rather than wild-type SV40. Figure 2 outlines schematically the strategy we used here to determine the in vivo conformational state of the d(CA/GT)30 sequence in SV40/CA30. In general, SV40 minichromosomes are not under any torsional stress (Germond et al., 1975). The linking difference of SV40 minichromosomes 'in vivo' is predominantly determined by the number of supercoils which are constrained by nucleosomes. Contributions to changes in linkage relative to the relaxed state from other sources appear to be negligible, owing largely to the action of topoisomerases within the cell. DNA isolated from SV40 minichromosomes is negatively supercoiled, showing a rather broad distribution of topoisomers, which reflects its nucleosomal organization (Germond et al., 1975). Each nucleosome contributes about 1.0 turns to the linking difference of the deproteinized SV40 DNA (Keller, 1975). Since SV40 minichromosomes are estimated to contain about 24 nucleosomes, deproteinization yields closed circular DNA with an average linking difference of approximately
1
2
-
-24 turns. Transition to the Z-conformation within the d(CA/GT)30 sequence would alter the helical twist of this segment. The total winding change that would be expected if the entire 60 bp of the d(CA/GT)30 sequence were to be converted from a right handed 10.5 bp-per-turn helix to a left-handed 12.0 bp-per-turn helix is 10.7 superhelical turns. In the absence of strand scission, this change in helical winding would necessarily alter the linking difference of the minichromosome in the positive direction by 10.7 1728
Fig. 3. Distribution of topoisomers generated by mature SV40/CA30 (lane 1) and SV40/0 minichromosomes (lane 2). Purified SV40/CA30 DNA and SV40/0 DNA were loaded on a 1% agarose -TBE gel containing 1.25 ,ug/mJ of chloroquine. Electrophoresis was performed at 50 V for 14- 16 h.
turns. Superhelical deviations of this magnitude relative to the
normal superhelical density of the minichromosome are expected to be rapidly eliminated by the endogenous presence of DNA
topoisomerases. Upon deproteinization and examination of the distribution of topoisomers under conditions where Z-DNA is not favoured, the median of the topoisomers distribution would be found at a lower linking number than that observed had tran-
DNA conformation of a d(CA/GT)30 sequence in SV40 minichromosomes
1
2
3
4
5 SV4O/O ,.'44, ki.....
.
.
j
SV4O/CA30
Ael
Fig. 5. Selected electron micrographs of SV40/O and SV40/CA30 minichromosomes. The bar corresponds to 250 nm.
Fig. 4. Distribution of topoisomers generated by topoisomerase-I-treated SV40/O (lane 2) and SV40/CA30 (lane 4) minichromosomes. Lanes 1 and 3 show the topoisomer distributions of untreated SV40/O and SV40/CA30 minichromosomes respectively. Electrophoresis was carried out on a 1 % agarose-TBE gel containing 1.25 Agg/ml of chloroquine. The bands present at the bottom of lanes 2 and 4 correspond to pDPL6 DNA (Haniford and Pulleyblank, 1983b) that was added to the same assay as an internal control and relaxed by the action of the enzyme. Under these electrophoretic conditions relaxed DNA run as positively supercoiled. Lane 5 shows negatively supercoiled pDPL6 DNA.
sition to the Z-conformation not occurred. In this case the linking difference of the isolated DNA will be the sum of the contribution due to the nucleosomes plus the contribution due to the conversion to the B-form of the d(CA/GT)30 sequence. In summary, if transition to the Z-form had occurred the average linking difference of the purified SV40/CA30 DNA would be 10.7 superhelical turns lower than the population average expected for SV40/O DNA. Minichromosomes from SV40/CA30- and SV40/O-infected CV 1 cells were prepared 36 h post-infection and deproteinized as described in Materials and methods. The distribution of topoisomers was visualized by electrophoresis on a 1 % agarose - TBE gel containing 1.25 jig/ml of the intercalator chloroquine (Figure 3). Deproteinization introduces negative supercoiling (Germond et al., 1975). The specific linking difference of the deproteinized SV40/CA30 DNA is high enough to induce the formation of Z-DNA within the d(CA/GT)30 sequence (Haniford and Pulleyblank, 1983a). In fact we found that
anti-Z-DNA antibodies specifically bind to purified SV40/CA30 DNA but not to SV40/O DNA, indicating that formation of ZDNA at the insert had occurred in deproteinized SV40/CA30 DNA. Upon intercalation, chloroquine relaxes negatively supercoiled DNA, so that under these electrophoretic conditions all topoisomers are in the B-form since they have a negative specific linking difference which is below the threshold required to flip or maintain the d(CA/GT)30 sequence in the Z-conformation. Under these conditions no significant difference was observed between the topoisomer distributions generated from SV40/CA30 or SV40/O minichromosomes (Figure 3), to be expected if transition to the Z-conformation of the d(CA/GT)30 sequence in the minichromosome had not occurred. An alternative explanation for the results observed in Figure 3 is that supercoils introduced into SV40/CA30 minichromosome as a consequence of Z-DNA formation in d(CA/GT)30 were not relaxed by topoisomerase action in vivo. If SV40/CA30 minichromosomes were positively supercoiled, treatment with topoisomerase-I (Topo-I) should remove any positive supercoils so that we expect the linking difference of Topo-I-treated and untreated minichromosomes to differ in -10.7 superhelical turns. The topoisomer distribution obtained when SV40/CA30 minichromosomes are treated with Topo-I (Figure 4, lane 4) turns out to be identical to that generated by untreated SV40/CA30 minichromosomes (Figure 4, lane 3). This distribution of topoisomer is also undistinguishable from that obtained with either Topo-Itreated (Figure 4, lane 2) or untreated (Figure 4, lane 1) SV40/O minichromosomes. Topo-I was working properly under our experimental conditions since it was able to relax negatively supercoiled DNA that was added to the same assay as an internal control (Figure 4). Therefore we conclude that SV40/CA30 minichromosomes are not under torsional stress. A group of faint bands migrating further than the predominant distribution can
1729
A.Rodriguez-Campos et al.
1
2
3
4
5
Fig. 6. Topoisomer distributions of deproteinized SV40/CA30 minichromosomes obtained at various times postinfection: 12 h (lane 1); 16 h (lane 2); 24 h (lane 3); 36 h (lane 4); and 48 h (lane 5). SV40/CA30 DNA obtained at each time was analysed in a 1 % agarose - TBE gel containing 1.25 tg/ml of chloroquine.
be observed in Figure 4, lanes 2 and 4. These bands correspond to the subset of minichromosomes that are relaxable by Topo-I (Luchnik et al., 1982). Under these electrophoretic conditions they run as positively supercoiled species. It is also conceivable that the identical topoisomer distributions of SV40/O and SV40/CA30 in Figure 3 may have arisen from a situation in which the insert of SV40/CA30 formed Z-DNA, but where the linkage difference produced by this change was balanced precisely by a loss of nucleosomes from SV40/CA30 minichromosomes. Figure 5 shows a few selected electron micrographs obtained from SV40/CA30 and SV40/O minichromosomes. The average number of nucleosomes per minichromosome obtained from 24 independent measurements was 22 2 for SV40/O minichromosomes and 21 + 2 for SV40/CA30 minichromosomes, which is in good agreement with the results obtained by others (Griffith, 1975; Germond et al., 1975). A loss of about 10-11 nucleosomes in SV40/CA30 minichromosomes relative to SV40/O would be required to balance the linkage change produced if the entire d(CA/GT)30 insert existed in the Z-form. In view of the above arguments, we conclude that Z-DNA does not exist to any appreciable extent within the d(CA/GT)30 insert of SV40/CA30 minichromosomes in vivo. The DNA conformation of the d(CA/G7)30 sequence is the same throughout the viral lytic cycle The experiments described above were performed with minichromosomes obtained 36 h post-infection. Even though the d(CA/GT)30 sequence apparently does not form left-handed ZDNA in such minichromosomes, the question still remains of whether it forms Z-DNA at any time during the viral lytic cycle. Figure 6 shows the distribution of topoisomers originating from minichromosomes harvested at different times after infection. Harvesting time points were selected to coincide with the major developmental stages of SV40 morphogenesis (Tooze, 1981): (i) early gene expression (12 h); (ii) viral replication (16 h); (iii) late gene expression (24 h); (iv) mature minichromosomes (36 h); and (v) viral assembly (48 h). It is apparent from Figure 6 that the population averages of topoisomers obtained at the different time points across the lytic cycle vary by no more than one turn, indicating that the d(CA/GT)30 sequence does not change its DNA conformation throughout the 1730
viral lytic cycle. Slight changes in the distribution of topoisomers can be observed in Figure 6. These small differences might be related to conformational changes occurring elsewhere in the viral minichromosome which would involve shorter sequences. Chromatin organization of the d(CA/G7)30 sequence Alternative d(CA/GT)n elements are quite abundant in eukaryotic genomes. It is of interest to ascertain whether these sequences introduce any distortion on the nucleosomal organization of chromatin regardless of their actual DNA conformation. The d(CA/GT)30 sequence in SV40/CA30 falls just outside the nucleosome-free gap which spans the SV40 control region. In order to determine whether or not the dimensions of this nucleosome free region are subject to change upon introduction of the d(CA/GT)30 sequence, we examined the sensitivity of SV40/O and SV40/CA30 minichromosomes to cleavage by three restriction endonucleases, Bgll, EcoRI and XbaI. The accesibility for each restriction site was determined from the amount of linearized SV40 DNA produced by the action of the restriction endonuclease as a function of time of digestion. Bgll restriction site is highly accessible since this enzyme cleaves SV40 DNA at the origin of replication in the nucleosome-free region (Varshavsky et al., 1978; Tack et al., 1981). On the other hand EcoRI restriction endonuclease cleaves SV40 DNA at position 1782 in a region of the minichromosome which is organized as nucleosomes. Therefore EcoRI restriction site is only slightly accessible (Varshavsky et al., 1978). During the cloning of the d(CA/GT)30 sequence a new and unique XbaI site was generated. This restriction site is localized immediately adjacent to the d(CA/GT)30 sequence in SV40/CA30 and it is also present in SV40/O (Figure 1). As can be seen in Figure 7, SV40/CA30 and SV40/O minichromosomes do not show any significant difference in the accessibilities of the XbaI, EcoRI or Bgll sites. It is clear from this experiment that while the BglI site of both SV40/O and SV40/CA30 is accessible to cleavage, the EcoRI and XbaI sites of both minichromosomes show similarly reduced sensitivities to cleavage. These results strongly suggest that the DNA sequences around the XbaI site, including the d(CA/GT)30 insert, fall within a region organized by nucleosomes and are not incorporated into the nucleosome-free gap. Evidence that a significant portion of the d(CA/GT)30 sequence is indeed packed into nucleosomes is provided by the micrococcal nuclease digestion experiments presented in Figure 8. A time course for micrococcal nuclease digestion was performed for both SV40/CA30 and SV40/O minichromosomes. Samples were taken at increasing digestion times, deproteinized and loaded on a 0.8% agarose - TBE gel (Figure 8, panel I). A typical nucleosomal ladder is obtained in both cases. The pattern of micrococcal nuclease digestion appears to be the same for both SV40/CA30 and SV40/O minichromosomes, in agreement with the fact that they contain a similar number of nucleosomes. As digestion proceeds, the bands corresponding to the lower mol. wt species become more prominent. After a 120-min incubation, most of the micrococcal nuclease-resistant fragments are found as monoand dinucleosomes. The DNA fragments in Figure 8, panel I were transferred to nitrocellulose paper and probed with [32P]poly d(CA/GT)n as described in Materials and methods. Figure 8, panel II shows an 8-h autoradiography of the hybridized nitrocellulose paper. No hybridization at all is observed for SV40/O nuclease-resistant fragments since SV40/O does not contain any sufficiently long repetitive d(CA/GT)n sequence to form a stable duplex with the 32P probe. However, a strong hybridization is detected with all nuclease-resistant fragments obtained from
DNA confonnation of a d(CA/GT)30 sequence in SV40 minichromosomes
SV40/0 I
wL
I
I
0 Bgl I Eco RI Xba I A B 1 2 3123 1 2 3
I
I_
Bgl
o 0
-J
0e 40_ Eco RI Xba I
201'/^ x-I
°l
I
a
I.
TIME , min.
SV40/CA30 Eco RI Xba I Bgl I A B 1 2 3 1 2 3 1 2 3
1-
I
I
I
Bgl I
l~~~~
ui -J
1l
401
o'p
Eco RI _
20-
__ A
o 1L
-L
0
15
30
60 TIME,min.
Fig. 7. Accessibility of the BglI, EcoRJ and XbaI sites in SV40/O and SV40/CA30 minichromosomes. The left panel shows 1% agarose-TBE gels of the DNA extracted from cleaved minichromosomes following incubation for varying time periods: 15 min (lane 1); 30 min (lane 2); and 60 min (lane 3). Lane A corresponds to undigested SV40 DNA and lane B corresponds to minichromosomes incubated for 60 min in the absence of enzyme. R: relaxed; L: linear; S: supercoiled. Quantification of these results are shown at the right. Percentage cleavage is expressed as the amount of linear DNA divided by total DNA. The slower migration of negatively supercoiled SV40/CA30 DNA reflects the formation of Z-DNA at the d(CA/GT)30 insert in deproteinized SV40/CA30 DNA.
SV40/CA30 minichromosomes, and in particular with mononucleosomal DNA, indicating that the d(CA/GT)30 sequence is found packed into nucleosomal core particles. Discussion SV40 is an excellent model system for eukaryotic chromatin. Here we have used SV40 minichromosomes into which a stretch of d(CA/GT)30 has been introduced, as a model for the general chromosomal organization of this sequence. Repetitive d(CA/GT)n elements are very abundant in eukaryotic genomes (Hamada et al., 1982) and they have the potential of forming left-handed Z-DNA (Hamada et al., 1984). We have also described the methodology to determine whether the inserted sequence forms Z-DNA in SV40 minichromosomes in vivo. This methodology is based upon changes in DNA linkage that are
associated with the B to Z transition. Haniford and Pulleyblank (1983b) have used a similar approach to study the occurrence of Z-DNA inside Escherichia coli. The major advantage of this approach lies in the fact that linkage is determined by the helical nature of the DNA in vivo and is fixed immediately upon isolation. Techniques which attempt to detect the presence of an altered DNA conformation in isolated chromatin are subject to the problem that DNA nicking, protein depletion or rearrangement, resulting from the manner of isolation, may destabilize the conformation in question. Our results indicate that the d(CA/GT)30 sequence remains predominantly in the right-handed B-conformation in vivo. From the results in Figure 3, we cannot exclude the possibility that the d(CA/GT)30 sequence would adopt the Z-conformation in a small percentage of minichromosomes that, from the precision of our technique, we estimate to be less than 1 %. However, one 1731
A.Rodriguez-Campos et al.
I
A
SV40/0 SV40/CA30 A 1 23 1 2 3A
1SV4 0
SV40/CA30 A 12 3 1 2 3 A
=-.4w
4
Fig. 8. Micrococcal nuclease digestion of SV40/O and SV40/CA30 minichromosomes. (1) 0.8% agarose -TBE gel of the nuclease resistant fragments obtained following incubation for varying time periods: 5 min (lane 1); 30 min (lane 2); and 120 min (lane 3). Lane A corresponds to SV40/CA30 DNA digested with SphI and XbaI to release a 223-bp fragment containing the d(CA/GT)30 insert indicated by the arrow. (II) Autoradiography of the gel in (I). DNA fragments were transferred to nitrocellulose paper and probed with [32P]polyd(CA/GT),, as described in Materials and methods.
should expect that minichromosomes containing the d(CA/GT)30 sequence in the Z-conformation would correspond to a functional subpopulation which, being a minor component of mature minichromosomes (obtained 36 h postinfection), would be predominant at some stage during the viral lytic cycle. As shown in Figure 6, no significant changes in superhelicity are detected throughout the lytic cycle of the virus. We cannot exclude the possibility that Z-DNA is formed transiently and to a small degree within the insert of the majority of minichromosomes. However, the B to Z transition is in general highly cooperative (Peck et al., 1982; Peck and Wang, 1983; Haniford and Pulleyblank 1983a, 1983b). Others have similarly concluded that the vast majority of d(CA/GT)n sequences in eukaryotic chromatin exist in the Bform (Gross et al., 1985). These conclusions have been drawn from the lack of the SI -nuclease sensitivity pattern characteristic of Z-DNA (Singleton et al., 1982) produced at d(CA/GT)n sequences upon treatment of isolated nuclei or nucleosomes with SI -nuclease. DNA nicking is likely to occur during nuclei isolation. Since nicks or breaks in the DNA of topologically distinct chromatin domains would release torsional strain required for 1732
Z-DNA stabilization, such studies cannot be considered as definitive. Negative supercoiling and specific protein binding are likely to be the factors which can preferentially stabilize Z-DNA in chromatin. A d(CA/GT)30 sequence forms Z-DNA when the specific linking difference is about -0.047 (Haniford and Pulleyblank, 1983a). Therefore a linking difference of about -23 superhelical turns would be required to stabilize this sequence in the Z-conformation in SV40. Although it is uncertain whether a subset of SV40 minichromosomes are supercoiled (Luchnik et al., 1982; Chen and Hsu, 1984; Barsoum and Berg, 1985), it is unlikely that they would ever be under such torsional stress. Z-DNA binding proteins have been isolated from several systems, including SV40 minichromosomes (Nordheim et al., 1983; Azorin and Rich, 1985; Lafer et al., 1985a. These proteins are believed to be very important in the in vivo stabilization of ZDNA. Z-DNA binding proteins appear to show sequence specificity so that they have a target sequence to which they bind very efficiently, while their affinity for any other Z-sequence is much lower. In particular, Z-DNA binding proteins isolated from SV40 minichromosomes show a quite low affinity for either
DNA conformation of a d(CA/GT)30 sequence in SV40 minichromosomes
d(CA/GT)n or d(CG/GC)n sequences when compared to the affinity for their recognition sequence (Azorin and Rich, 1985). Therefore it is possible that they will not be able to stabilize the d(CA/GT)30 sequence in the Z-conformation, even though they might do so at their target sequence. In this paper the d(CA/GT)30 sequence has been cloned at the unique HpaH restriction site present in SV40. Our results indicate that the inserted sequence is packed into nucleosomes. Formation of nucleosomes within this sequence might make its conversion to the Z-conformation more difficult. It is controversial whether nucleosomes can accommodate DNA in the Zconformation (Nickol et al., 1982; Miller et al., 1985). However, the Hpall site which appears to be as accessible to endonucleolytic cleavage as the BglI site (Weiss et al., 1985) occurs just at the border of the nucleosome-free gap, so that the d(CA/GT)30 sequence could have been incorporated into it if required. Most likely, nucleosome formation reflects the fact that this sequence adopts a regular B-DNA conformation. The inability of the experiments presented here and elsewhere (Gross et al., 1985) to detect Z-DNA formation within d(CA/GT)n segments in chromatin should not be interpreted as meaning that these sequences are incapable of adopting the Zform in vivo. It can be reasonably expected that Z-DNA formation is a selective process which occurs in response to environmental stimuli, which may only occur at certain times of the cell cycle or along a developmental pathway. In this regard it has been suggested that d(CA/GT)n elements are hot spots for genetic recombination (Hamada et al., 1982; Haniford and Pulleyblank, 1983b). Notably recd, a protein isolated from Ustilago which plays a fundamental role in genetic recombination (Kmiec and Holloman, 1982), has been shown to be a ZDNA binding protein (Kmiec et al., 1985). It remains to be seen whether or not this protein utilizes Z-DNA pre-existing in the chromosome to signal recombination or alternatively induces the formation of Z-DNA as a low-energy intermediate for the stabilization of the paranemic joint (Kmiec and Holloman, 1984). Formation of this intermediate would be facilitated at d(CA/GT)n sequences since they can adopt the left-handed Z-conformation more easily.
Materials and methods Construction of SV40/CA30 and SV40/O Construction of the recombinant viruses SV40/CA30 and SV40/O will be described in detail elsewhere. They derived from plasmids pE.9 and pE. 10. Plasmid pE.9 is a 2.2-kb plasmid derived from pDPL6 (Haniford and Pulleyblank, 1983b) and pUC 12 (Viera and Messing, 1982), which contains a polylinker with the following restriction sites: EcoRI-SstI-SmaI-BamHI-XbaI-Sal-BamHI-SmaI-A7baI-HindllIClaI. pE. 10 derives from pDHfl4 (Haniford and Pulleyblank, 1983a, 1983b) and is like pE.9 but contains a 60-bp segment of alternating adenine and cytosine residues cloned into the second SnaI site of the polylinker described above. SV40 DNA linearized at the HpaII site was inserted at the unique Sall site (cut with AccI) of either pE.9 (SV40/O) or pE. 10 (SV40/CA30). These shuttle vectors were then restricted with XbaI, which releases the plasmidic sequences but leaves part of the polylinker (from Sall to Xbal) in the SV40 genome including the d(CA/GT)30 sequence. Ligation was performed at a low DNA concentration to maximize formation of self-ligated species. Closed circular SV40 DNA recombinants containing the d(CA/GT)30 sequence (SV40/CA30) or only the carried over plasmidic sequences (SV40/O) were purified in a 1% low-melting-point agarose -TBE gel containing 0.5 14g/mn of ethidium bromide. Purified DNA was then transfected to just confluent CV1 cells by the DEAE -dextran method (Lai, 1980). Individual plaques were picked 21 days after transfection and used to inoculate fresh plates of CV I cells to produce high titer stocks. SV40/CA30 DNA and SV40/O DNA were characterized by restriction mapping and sequencing of the viral region around the insertion (Figure 1). Preparation of minichromosomes Confluent CVI cells were infected with either SV40/CA30 or SV40/O viruses
and minichromosomes were prepared 36 h postinfection according to Su and DePamphilis (1978). Purified minichromosomes were kept at -20°C in 50% glycerol. Analysis of the distribution of topoisomers Purified minichromosomes were deproteinized by the SDS - chloroform isoamylic method (Zamengohf, 1957) and the distribution of topoisomers was resolved in a 1% agarose -TBE gel containing 1.25 Ag/ml of chloroquine (Peck et al., 1982). Electrophoresis was carried out at 50 V for 14- 16 h and gels were stained with ethidium bromide. Under these conditions approximately 15 SV40-topoisomers can be resolved. Chloroquine relaxes negatively supercoiled DNA so that under these conditions all topoisomers have a specific linking difference below the one required to flip a d(CA/GT)30 sequence to the Z-conformation (Haniford and Pulleyblank, 1983a). When the distribution of topoisomers was determined at different times postinfection, Hirt's extracts were prepared at each time (Hirt, 1967). Negatively supercoiled DNA was purified in a 1 % low melting point agarose -TBE gel and the distribution of topoisomers resolved as described above. Topoisomerase-I treatment About 0.25 tg of purified minichromosomes (100 il) were diluted to 1 mi with a buffer containing 200 mM NaCl, 10 mM Tris-HCI and 0.1 mM EDTA (pH = 8.0) (Peck et al., 1982), and treated overnight at room temperature with 10 units of calf thymus topoisomerase-I (BRL). Samples were deproteinized and the distribution of topoisomers were resolved as described above. Restriction endonuclease digestion experiments All digestions were carried out at 37°C in a buffer containing 100 mM NaCl, 6 mM Tris-HCI (pH = 8.0), 6 mM MgCl2, 6 mM DTT and 0.1 mM PMSF. Digestions were stopped by the addition of 200 mM EDTA to a final concentration of 10 mM. Samples were deproteinized and loaded on a 1% agarose -TBE gel. Electrophoresis was performed at 50 V for 14- 16 h and gels were stained with ethidium bromide. The percentage of linear DNA was determined from the densitometric tracings of the corresponding photographs. Restriction enzymes were obtained from New England Biolabs. Micrococcal nuclease digestion and hybridization experiments Purified minichromosomes were digested with micrococcal nuclease (Sigma) at an enzyme/DNA ratio of 1 unit of activity (Boehringer)/mg DNA. Digestions were carried out at 37°C in a buffer containing 150 mM NaCl, 1 mM CaC12, 1 mM MgCl2 and 100 mM Tris-HCI (pH = 8.0). Digestions were stopped by the addition of 200 mM EDTA to a final concentration of 10 mM. Samples were deproteinized and loaded on a 0.8 % agarose -TBE gel. Electrophoresis was performed as described above. Transfer to nitrocellulose paper was performed according to Maniatis et al. (1982). Transfer of DNA was allowed to proceed for 2 h. Filters were dryed overnight at 65°C and hybridized to [32P]polyd(CA/GT)n as described by Hamada et al. (1982). [32P]polyd(CA/GT)n was prepared by nick-translation of polyd(CA/GT), (P.L.Biochemicals) according to Rigby et al. (1977). The sp. act. was - 1.0 x 108 c.p.m./yg. Autoradiography was carried out at -800C for 8 h. Electron microscopy Purified minichromosomes were diluted 6-10 times with distilled water to dilute the glycerol present in the preparation and then fixed for electron microscopy with 0.2% formaldehyde for 15 min. Samples were adsorbed onto freshly prepared carbon films for 5 min, rinsed with re-distilled water, stained with 0.5% uranyl acetate and washed for 5 min with re-distilled water. Grids were dehydrated for 10- 15 s in ethanol, dried and rotary-shadowed with Pt:C evaporated from an electron gun at an angle of 7°. The specimens were observed in a Phillips EM 301 and micrographs taken at x25 000.
Acknowledgements We are thankful to Drs J.A.Subirana, A.Rich and L.Cornudella for critical comments, advice and support. The technical assistance of Virginia Stiefel is acknowledged. This research was supported by grants from the Comisi6n Asesora de Investigaci6n Cientifica y T6cnica and the Consejo Superior de Investigaciones Cientificas. A.R.C and L.P.G. were recipients of a postdoctoral fellowship from the C.S.I.C. M.J.E. was supported by a fellowship from the M.R.C of Canada. A.R.C. dedicates this work to Marta.
References Azorin,F. and Rich,A. (1985) Cell, 41, 365-374. Azorin,F., Nordheim,A. and Rich,A. (1983) EMBO J., 2, 649-655. Barsoum,J. and Berg,P. (1985) Mol. Cell. Biol., 5, 3048-3057. Chen,.S. and Hsu,M.-T. (1984) J. Virol., 51, 14- 19. DiCapua,E., Stasiak,A., Koller,T., Brahms,S., Thomae,R. and Pohl,F.M. (1983) EMBO J., 2, 1531-1535.
1733
A.Rodriguez-Campos et al. Germond,J.E., Hirt,B., Oudet,P., Gross-Bellard,M. and Chambon,P. (1975) Proc. Natl. Acad. Sci. USA, 72, 1843- 1847. Griffith,J. (1975) Science, 187, 1202-1203. Gross,D.G., Huang,S.-Y. and Garrard,W.T. (1985) J. Mol. Biol., 183, 251-265. Hamada,H. and Kakunaga,T. (1982) Nature, 298, 396-398. Hamada,H., Petrino,M.G. and Kakunaga,T. (1982) Proc. Natl. Acad. Sci. USA, 79, 6465-6469. Hamada,H., Petrino,M.G., Kakunaga,T., Seidman,M. and Stollar,B.D. (1984) Mol. Cell. Biol., 4, 2610-2621; Haniford,D.B. and Pulleyblank,D.E. (1983a) Nature, 302, 632-634. Haniford,D.B. and Pulleyblank,D.E. (1983b) J. Biomol. Str. Dynamics, 1, 593 -609. Hirt,B. (1967) J. Mol. Biol., 26, 365-369. Jakobovits,E.B., Bratosin,S. and Atoni,Y. (1980) Nature, 285, 263 -265. Keller,W. (1975) Proc. Natl. Acad. Sci. USA, 72, 4876 -4880. Klysik,J., Stirdivant,S.M., Larson,J.E., Hart,P.A. and Wells,R.D. (1981) Nature, 290, 672-677. Kmiec,E. and Holloman,W.K. (1982) Cell, 29, 367-374. Kmiec,E. and Holloman,W.K. (1984) Cell, 36, 593-598. Kmiec,E., Angelides,K.J. and Holloman,W.K. (1985) Cell, 40, 139-145. Lai,C.J. (1980) Methods Enzymol., 65, 811-816. Lafer,E.M., Sousa,R., Rosen,B., Hsu,A. and Rich,A. (1985a) Biochemistry (Wash.), 24, 5070-5076. Lafer,E.M., Sousa,R. and Rich,A. (1985b) EMBO J., 4, 3655-3660. Luchnik,A.N., Bakarayev,V.V., Zbarsky,I.B. and Georgiev,G.P. (1982) EMBO J., 1, 1353-1358. Maniatis,T., Fritsch,E.F. and Sambrook,J. (1982) Molecular Cloning. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Miller,F.D., Dixon,G.H., Rattner,J.B. and van de Sande,J.H. (1985) Biochemistry (Wash.), 24, 102-109. Nickol,J., Behe,M. and Felsenfeld,G. (1982) Proc. NatI. Acad. Sci. USA, 79, 1771-1775. Nordheim,A. and Rich,A. (1983) Proc. Natl. Acad. Sci. USA, 80, 1821 - 1825. Nordheim,A., Tesser,P., Azorin,F., Kwon,Y.H., Moller,A. and Rich,A. (1982a) Proc. Natl. Acad. Sci. USA, 79, 7729-7733. Nordheim,A., Lafer,E.M., Peck,L.J., Wang,J. Stollar,B.D. and Rich,A. (1982b) Cell, 31, 309-318. Peck,L.J. and Wang,J. (1983) Proc. Natl. Acad. Sci. USA, 80, 6206-6210. Peck,L.J., Nordheim,A., Rich,A. and Wang,J. (1982) Proc. Natl. Acad. Sci. USA, 79, 4560-4564. Revet,B., Zarling,D.A., Jovin,T.M. and Delain,E. (1984) EMBO J., 3, 3353 -3358. Rich,A., Nordheim,A. and Wang,A.H.-J. (1984) Annu. Rev. Biochem., 53, 791 -846. Rigby,P.W.J., Dieckmann,M., Rhodes,C. and Berg,P. (1977) J. Mol. Biol., 113, 237 -245. Saragosti,S., Moyne,G. and Yaniv,M. (1980) Cell, 20, 65-73. Singleton,C.K., Klysik,J., Stirdivant,S.M. and Wells,R.D. (1982) Nature, 299, 312-316. Stockton,J.F., Miller,F.D., Jorgenson,K.F., Zarling,D.A., Morgan,A.R., Rattner,J.B. and van de Sande,J.H. (1983) EMBO J., 2, 2123-2128. Su,R.T. and DePamphilis,M.L. (1978) J. Virol., 28, 53-65. Tack.L.C., Wassarman,P.M. and DePamphiis,M.L. (1981) J. Biol. Chem., 256, 8821 -8823. Tooze,J. (1981) DNA Tumour Viruses. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Varshavsky,A., Sundin,O. and Bohn,M.J. (1978) Nucleic Acids Res., 5, 3469-3477. Vieira,J. and Messing,J. (1982) Gene, 19, 259-263. Wang,A.H.-Y., Quigley,G.J., Kolpack,F.J., Crawford,J.L., van Boom,J.H., van der Marel,G. and Rich,A. (1979) Nature, 287, 680-686. Weiss,E., Ghose,D., Schultz,P. and Oudet,P. (1985) Chromosoma, 92, 391 -400. Zamengohf,S. (1957) Methods Enzymol., 3, 696-712. Zimmernan,S.B. (1982) Annu. Rev. Biochem., 51, 395-427.
Received on 24 March 1986
1734
