Volume 5 Number 6 June 1978
NllCleiC Acids Research
Preparation and isolation of covalently closed circular rDNA molecules from DNA of Xenopus laevis
Lino C. Polito and Silvio Spadari*
International Institute of Genetics and Biophysics, CNR, Naples, and *Laboratorio di Genetica
Biochimica e Evoluzionistica, CNR, Pavia, Italy
Received 17 January 1978
ABSTRACT
We describe a method leading to the formation of closed circles of
rDNA starting from total DNA of Xenopus laevis. Linear DNA molecules were
digested with exonuclease 3 and self-annealed. Open circles were enriched
and covalently closed by the simultaneous use of polynucleotide kinase,
DNA polymerase and polynucleotide lijase. Closed circles of rDNA^ were
shown to be alkali-resistant, to have higher density than linear molecules
in cesium chloride density gradients containing ethydium bromide, and to
have the sedimentation constant expected for a single repeat unit of rDNA
comprehensive of its spacer.
INTRODUCTION
Tandemly arranged repeated DNA sequences can be revealed by their ability to form circles or compound figures after exonuclease 3 digestion
2 3
and self-annealing ' . Given a certain length of the repeated DNA sequence, the extent of single circles and the kind of circle formed, depends
upon the molecular weight of the DNA used and the extent of exonuclease 3
3
digestion, as seen by electron microscopy . These methods are potentially
useful for the isolation of repeated DNA sequences as single, covalently
closed circles '
' . We describe here the isolation of genes for rRNA
7-9
which are known to be repeated and tandemly arranged
. The data present-
ed in this paper indicate that genes for rRNA were isolated as closed c i r cles. Our efforts were concerned with the development of techniques potent i a l l y capable of leading to the isolation as closed c i r c l e s , of any tandemly repeated sequence existing in the chromosome.
© Information Retrieval Limited 1 Falconberg Court London W1V5FG England
2211
Nucleic Acids Research
MATERIALS AND METHODS
Chemicals: Deoxyguanosine-5'-triphosphate, deoxycytidine-5'triphosphate,
deoxyadenosine-5'triphosphate, thymidine-51triphosphate were obtained from
SERVA (Heidelberg); adenosine-5'triphophate from SIGMA (Missouri), 5',-amylase
and pancreatic ribonuclease were obtained from Worthington (N.J.). ribonuand pronase from Calbiochem (Calif.)- Bacterial E. coli Str. B
3
11303 was obtained from GBI (Ohio). Uridine-5- H (25 curie/mM) and deoxy-
clease T
adenosine
H(G)-5'triphosphate (5 curie/mM) was obtained from New England
Nuclear (Frankfurt, Germany);
chemical
P carrier free (60 curie/mM) from Radio-
Centre (Amersham, England); SSC
is 0.15 M NaCl, 0.015 M sodium
citrate.
Exonuclease 3: was prepared from E. coli as described by Richardson
and Kornberg
and assayed according to Richardson, Lehman and Kornberg
DNA polymerase: was prepared from E. coli as described by Richardson
12
et al. . Enzyme activity was measured on activated DNA also in the final
steps of purification.
Polynucleotide kinase: was obtained from T, infected E. coli cells as
*
4
described by Richardson
Polynucleotide ligase: was also obtained from T
14
as described by Weiss and Richardson
jf -
infected E. coli cells
P ATP: necessary to assay polynucleotide kinase and ligase during
purification, was prepared as described by Glynn and Chappel
DNA: was extracted from blood of Xenopus laevis as described by Birnstiel et al.
RNA: was labelled with uridine-5- H (50 -piCi/ml) from a stabile Xenopus
laevis line of kidney cells , kindly supplied by M. Birnstiel, mainly according to Loening et al.
. In short, after three days of incubation,
cells were collected, RNA extracted and ethanol precipitated. The precipitate was dissolved in 0.4% SDS, 10 mM sodium acetate, 10 mM EDTA buffer,
pH 5.0, and the different RNA components were fractionated on a 5-20%
linear sucrose density gradient in the same buffer. The 28S and 18S ribosomal fractions were pooled and dialyzed against 2 x SSC for 2-3 hours at
room temperature to eliminate the majority of SDS and after against fresh
2212
Nucleic Acids Research
2 x SSC at 4°C overnight. In different preparations the specific activity
of rRNA was in the range 5 x 10
- 1 x 10
cpm/jag.
Experimental procedure: The schedule of a typical large scale experiment is described here.
A solution containing 20 mg of high molecular weight DNA of Xenopus
laevis in 20 ml 1/10 x SSC was syringed twice through a N°18 needle to
a mean length of 6-8 microns as determined by E.M. The DNA solution was
then dialyzed against 40 volumes of 0.06 M Tris buffer pH 8.0.
The incubation mixture for exonuclease 3 digestion was the following:
MgCl
0.006 M, fi-mercaptoethanol 0.001 M, DNA 170 micro grams /ml, Exo 3
60 units/ml in a total volume of 90 ml. After 75 minutes at 37°C, further
22 units/ml of Exo 3 were added and digestion was continued for 45 min.
25% of the DNA was acid soluble.
The enzyme was then removed by a chloroform-isoanyl alcohol extraction
and the DNA solution was dialyzed against 40 volunes of l/loo x SSC for
5 hours. The solution was then adjusted to 14 micrograms DNA/ml in 0.9 M
NaCl and 0.01 M Na citrate and kept for 12 hours at 55°C to anneal the
single strand ends. These conditions are stringent to get a complete renaturation of the "steady ends"
'
and were chosen to minimize the possi20
bility of concatenation. The Cot 1/2 of Xenopus rDNA has been estimated
to be 0.002 under comparable conditions. In our hybridization mixture the
—8
concentration of single stranded rDNA is 1.2 10
moles/liter (in nucleotides). Under these conditions, considering also the single stranded rDNA
not contemporarely engaged in the faster reaction of self annealing, the
extent of concatenation estimated to occur is not more than 10%.
The annealed DNA solution was then loaded onto a 300 ml nitrocellulose
column previously washed with 5 liters of 0.9 M NaCl that can trap up to
9 21
20 mg of single stranded DNA '
. After loading, the column was washed
with 800 ml 0.09 M NaCl, and the wash was added to the loading solution.
Then the column was washed sequentially with 800 ml each of 0.3 M, 0.15 M
and 0.075 M NaCl. Each of these fractions was kept separate. The column
was finally washed with 800 ml 0.1 M NaOH. The last fraction was neutralized
and all the fractions were adi-sted to at least 0.3 M NaCl. The DNA of each
2213
Nucleic Acids Research
fraction was precipitated by the addition of two volumes of ethanol in
the cold, overnight. The precipitate was collected by centrifugation for
30 minutes at 10.000 x g and each DNA fraction was dissolved in 5-10 ml of
Tris 0.05 M pH 7.5 and dialyzed against the same buffer overnight,. A sample
• 22
of each DNA fraction was denatured, adsorbed to nitrocellulose filters
and
hybridized with
H labelled Xenopus rRNA. As a control,4.25 mg DNA were soni-
cated to a mean length of 0.5-1 micron and treated exactly as described above.
The 0.3, 0.15 and 0.075 M NaCl fractions of non-sonicated DNA were pooled.
170 micrograms were used for a small scale experiment. Then the remaining
590 micrograms (which gave a saturation value of 0.2% after hybridization
with rRNA) were also treated for ring closure.The incubation mixture used
for ring closure contained 0.02 M Tris pH 7.8, 0.007 M fcgCl,, 0.013 M DDT,
0.6 umoles/ml dXTP, 1.2 umoles/ml ATP, DNA polymerase <• U/ml, polynucleotide kinase 50 U/ml and polynucleotide ligase 1.7 U/ml in a total volume of
25 ml. The incubation was at 15°C for 210 minutes. The enzymes were then removed by chloroform extraction and the DNA dialyzed overnight against 40 volumes of 1 x SSC. The DNA solution was warmed to 40°C under a flow of nitrogen to remove traces of chloroform, made 0.2 M NaOH and kept at room temperature for 30 minutes. After neutralization the NaCl concentration was adjusted to 0.9 M and the solution loaded onto a 10 ml nitrocellulose column.
The column was washed first with 3 bed volumes of 0.075 M NaCl, and then
with 3 bed volumes of 0.1 M NaOH. 510 micrograms of DNA were found in the
NaOH fraction. The loading fraction and the 0.075 M NaCl fractions were pooled and the DNA precipitated overnight with 2 volumes of ethanol. The precipitate was recovered by centrifugation and dissolved in 1 ml of Tris 0.01 M
pH 7.8. This DNA fraction was used to test for the presence of closed rings
of rDNA (test fraction). To avoid DNA breakage, no tests were made of the
fraction.
Sucrose density gradients: 5-20% linear sucrose density gradients were
prepared in Tris 0.01 M pH 7.8, total 28 ml. 0.25 ml test fraction was
layered on top, together with 20 micrograms mouse 28S rRNA as a marker. Centrifugation was in a Spinco SW 25.1 rotor at 23.000 rpm for 20 hours at 5°C.
One ml fractions were collected after puncturing the bottom of the tube
2214
Nucleic Acids Research
with a N°14 needle. After an OD reading, each fraction was divided in two.
One of these was sonicated for 10 seconds in a MSE sonicator. The sonicated
and the non-sonicated fractions were then made 0.2 M NaOH and kept at room
temperature for 45 minutes. After neutralization, each fraction was adjusted to 6 x SSC and passed through a nitrocellulose filt;er. The filters were
air dried and then kept for 4 hours at 80°C. They were then incubated in
3
2 x SSC containing 2 micrograms/ml H labelled rRNA of Xenopus (970.000
cpm/ug) or of 28S alone for 12 hours at 60°C. Unhybridized RNA was removed
by digestion with a 2 x SSC solution containing 50 micrograms/ml pancreatic
ribonuclease and 20 U/ml T
ribonuclease for one hour at 30°C. After exhau-
stive washing, the filters were dried and counted in a Mark 1 liquid scintillation counter.
Cesium chloride density gradients: The DNA to be analyzed was mixed with
cesium chloride solution in 0.01 M Tris pH 7.8 to a mean density varying
from 1.58 to 1.67. The solution also contained 100 micrograms/ml ethidium
23
bromide . Centrifugation was for 50 hours at 33.000 rpm at 25°C in a Spinco
SW39 rotor. One or two drop fractions were collected after puncturing the
bottom of the tube with a N°14 needle.The refractive index of each fraction
was determind. After dilution with water, each fraction was sonicated, the
DNA denatured, adsorbed to nitrocellulose filters and hybridized with rRNA
as described for the sucrose density gradients.
RESULTS
We can formally divide this work in two steps: in the first step we get
a 5-6 fold enrichment of rDNA genes in form of open circles; in the second
one we purify practically to the homogeneity the rDNA genes in form of
closed circles.
First step: to enrich for the circles, we made use of the observation
that DNA can be fractionated according to its content of single stranded
regions on columns of nitrocellulose. Completely double stranded DNA is
not retained by nitrocellulose, but DNA digested to 20% or more by treatment with Exo 3 is almost completely retained. As the ionic strength is
decreased, fractions of DNA are eluted which contain increasing propor-
2215
Nucleic Acids Research
TABLE
1
Enrichment of circularized rDNA by nitrocellulose column
Fraction
micrograms
recovered
% of total DNA
recovery
0.9 M NaCl
13
0.085
0.3 M NaCl
82
0.58
0.27
0.15 M NaCl
104
0.74
0.32
0.075 M NaCl
570
4.0
0.21
95.0
0.03
0.1 M NaOH
13.360
Unfractionated DNA
rRNA/DNA
at saturation
-
0.055
tions of single stranded regions. The bulk of the DNA can only be removed
in alkali.
It is to be expected that DNA which has been treated with Exo 3 to
produce single stranded tails will contain a smaller proportion of single
strands in those molecules which have formed circles, and therefore it is
to be expected that these could be separated from the linear molecules
on nitrocellulose. If rDNA forms circles more readily than the bulk of
the DNA, we would expect to see an enrichment of rDNA in those fractions
which are eluted from nitrocellulose at high ionic strength, that is in
those fractions which contain a small portion of single strands. This indeed
was found to be the case (Table 1 ) ; fractions eluted with 0.3 to 0.15 M
NaCl showed a 5-6 fold enrichment for ribosomal genes over the total DNA.
In order to show that this enrichment was associated with circle formation and not due to some other special property of rDNA, such as bias
in G+C content or resistance to Exo 3, we repeated the procedure on DNA
sheared by sonication to a size shorter than the repeated length of the
ribosomal gene. In this case, it is not possible for rDNA to circularize;
and although some circles are formed, these presumably arise from repeated
sequences whose repeated length is shorter than that of rDNA. With this
short DNA there was no enrichment of rDNA in any of the fractions obtained
from nitrocellulose (Table 2) . The enriched fractions of rDNA behave as
predicted for open circles on nitrocellulose.
2216
Nucleic Acids Research
Table 2
Control nitrocellulose column
Fraction
micrograms
recovered
% of total DNA
recovery
rRNA/DNA
at saturation
-
0.9 M NaCl
60
2.0
0.3 M NaCl
109
3.6
0.040
0.15 M NaCl
322
10.8
0.048
0.075 M NaCl
625
20.9
0.041
1870
63.3
0.050
0.1 M NaOH
Unfractionated DNA
0.053
Second step: The experiments described below were intended to test the
feasibility of isolating rDNA in the form of closed circles and, incidentally, to demonstrate that much of the rDNA enriched as described above,
is indeed in the form of open circles.
Open circles can be covalently closed by treatment with polynucleotide
kinase, DNA polymerase and polynucleotide ligase. The kinase is needed to
replace missing 5'phosphates
4
, DNA polymerase to repair the single strand14
ed regions , and ligase to join the 3'OH group to the 5'phosphate
. After
this treatment covalently closed circles can be separated from all other
kinds of DNA by making use of the fact that closed circles are not dena.24
tured in alkali . We treated fractions of DNA which had been enriched for
open circles and rDNA as described above, with the three enzymes necessary
for closing circles, and then after denaturing the Dt!A in alkali, separated the native DNA from the rest on nitrocellulose. The presence of rDNA in
the form of covalently closed circles in the alkali resistant DNA was
shown in the following set of experiments. A fraction of the alkali resistant DNA was applied to a sucrose gradient with mouse 28S rRNA as a marker
(Fig. 1). Each fraction was tested for rDNA by hybridization before and
after sonication. The highly significant result of this experiment is that
virtually no hybridization is seen in the fraction before sonication. In
the sonicated fractions, two main peaks of hybridization are seen. One peak
is at 31 to 34S, assuming a sedimentation coefficient of 29-32S for the
2217
Nucleic Acids Research
(j
1
I
1
i
a
'iA
0.1 _
1 V
\
\
t
i
Si
!
i
i
i
i
r
\
1
/• V
\
\ \
1
1
* i
i
1
*
1
1
1
I
i
i
i
i
i
1
r
\
\
\
I
I
i
/
\
\
i
V...V
15
10
1
i I
p
I
' /
/'
0.05 "
6000
i
I
20
I
i
I
I
1
I
I
1
1/
1'
•4000
1 I*
if
if"
2000
I
25
FRACTIONS
Fig. 1 - Sucrose density gradient of Xenopus laevis DNA, after circula) , Optical denrization of rDNA and ring closure (see METHODS). (
sity; ( 0 — 0 — 0 ) , 3 H rRNA/DNA hybrids after sonication; (•••••••),
3
H rRNA/DNA hybrids without sonication.
marker
25-27
and the other, about two-thirds of the total hybridization, in
the light part of the gradient. We got the same pattern also if hybridization was made with the 28S alone.
Another fraction of the alkali-resistant DNA was banded in a cesium chloride gradient in the presence of ethidium bromide. This intercalating dye
reduces the density of linear DNA and nicked circles of DNA much more than
23 24 28
that of closed circles
'
'
.The gradient was hybridized with rRNA as
described for the sucrose gradient and again it was found that hybridization occurred after sonicating the fractions (Fig. 2 ) . We interpret this
to mean that the rDNA present must be in a form that cannot be denatured
without first breaking the molecules, that is, in a covalently continuous
double stranded form. We can conclude that, coupling nitrocellulose column
and sucrose or CsCl density gradients, it is possible to get practically
homogeneous rDNA genes.
2218
Nucleic Acids Research
6000 -
V
n
•i
i'(
3000
; ;
E
a
2000
1000
••it.
„
'; •' H I \fy o / ' . / $ ,?,
10
20
B
ji
* • »
30
40
FRACTIONS-DENSITY
50
DECREASING
". P \ K
60
70
Fig. 2 - Cesium chloride density gradients containing ethidium bromide.
180 microliters of "test-fraction" were used. The initial density was
-vl.5S g/cm^. One drop fractions were collected. The ^ H labelled rRNA
used for hybridization had a specific activity of 970.000 cpm/microgram.
(0--0--0), % rRNA/DNA hybrids after sonication; (•••#•••), 3 H rRNA
hybrids without sonication.
DISCUSSION
Three independent lines of evidence show that DNA enriched for circles
and treated to close the circles does indeed contain closed circles of
rDNA. First, the DNA does not hybridize to rRNA unless it is treated first
by sonication to break the circles. Second, the rDNA sediments at a rate
expected for closed circles with a molecular weight between 8.4 - 10.8 x 10 .
29
This calculation
was based on the assumption that closed circles sedi-
ment about 1.24 times faster than nicked circles
sediment 1.15 times faster than linear molecules
sedimentation constant
, that nicked circles
'
, and that the true
25—27
of the large rRNA of mouse is 29-32
lar weight of 8.4 - 10.8 x 10
. A molecu-
is quite close to the value expected for
one repeating unit of rDNA including the spacer '
. Finally, in cesium
chloride gradients containing ethidium bromide, the closed circles of
rDNA band with a much higher density than linear rDNA
'
2219
Nucleic Acids Research
We are surprised to find that a fraction of the covalently closed rDNA
sedimented very slowly. The hybridization of this DNA was also dependent
on sonication and we believe that this rDNA is in the form of covalently
closed linear molecules. Probably the combined action of DNA polymerase
and ligase can form such molecules by the addition of poly-dAT at the ends
of the linear duplex
. This could also explain the absence of an extra
peak of hybridization in the ethidium bromide gradient, since such molecules would bind as much of the dye as open linear molecules and, in addition, might be lightened by the presence of poly-dAT. It is our experience that material floating on top of the gradient is lost.
There are a few ways in which genes can be isolated from the bulk of
the DNA. The methods used in prokaryotes
are not yet applicable to higher
organisms. The approach made in higher organisms is principally based on
centrifugation in density gradients of DNA after interaction with different
substances, such as metal ions, Actinomycin, RNA, etc. '
. The method
described here can, in principle, be used to enrich and eventually be a
step for the isolation of any gene present in a tandemly duplicated form.
The experiment in which low molecular weight DNA was used demonstrates how
the method might be used to exclude long repeated length. A discussion of
3
this aspect has already been presented .
It is likely that genes isolated in this way, as closed circles, will
be one of the much useful material in structural studies and especially in
in vitro studies of regulation, for their structure closely mimic that of
the repeated genes in the chromosome, at least in that each gene is preceded and followed by a similar copy. For this reason we think that this material could be the best template for transcriptional studies not only for
what concerns the initiation and termination signals in eukaryotes, but
also to study other protein interactions in transcription.
Finally, our experiments suggest a method to test whether the sequences
complementary to any given RNA molecule are tandemly repeated, namely,
after processing the DNA in the way described, and hybridizing before and
after sonication, an increase in hybridization will indicate that the sequences are repetitious.
2220
Nucleic Acids Research
REFERENCES
1 Abbreviations: SSC, standard saline citrate (0.15 M NaCl - 0.015 M
sodium citrate); rRNA, ribosomal RNA; rDNA, DNA sequences that code
for ribosomal RNA; SDS, sodium dodecyl sulfate.
2 Thomas, C.A. Jr., Hamkalo, B.A., Misra, D.N. and Lee, C.S. (1970)
J. Mol. Biol. 51, 621
3 Polito, L.C., Graziani, F., Boncinelli, E., Malva, C. and Ritossa, F.M.
(1971) Nature New Biol. 229, 84
4 Goulian, M. and Kornberg, A. (1967) Proc. Natl. Acad. Sci. US 58, 1723
5 Polito, L.C. (1973) Atti Ass. Genet. Ital. 17, 33
6 Shachat, F.H. and Hogness, D.S. (1973) Cold Spring Harb. Symp. Quant.
Biol. 38, 371
7 Brown, D.D. and Weber, C.S. (1968) J. Mol. Biol. 34, 661
8 Birnstiel, M.L., Chipchase, M. and Speirs, J. (1971) Prog, in Nucleic
Ac. Res. Mol. Biol. 11, 351
9 Quagliarotti, G. and Ritossa, F.M. (1968) J. Mol. Biol. 36, 57
10 Richardson, C.C. and Kornberg, A. (1964) J. Biol. Chem. 239, 242
11 Richardson, C C , Lehman, I.R. and Kornberg, A. (1964) J. Biol. Chem.
239, 251
12 Richardson, C . C , Schildkraut, C.L., Aposhian, H.V. and Kornberg, A.
(1964) J. Biol. Chem. 239, 222
13 Richardson, C.C. (1965) Proc. Natl. Acad. Sci. US 54, 158
14 Weiss, B. and Richardson, C.C. (1967) Proc. Natl. Acad. Sci. US 57, 1021
15 Glynn, I.M. and Chappel, J.B. (1964) Biochem. J. 90, 147
16 Birnstiel, M.L., Wallace, H., Sirlin, J. and Fishberg, M. (1966) Nat.
Cancer Inst. Monogr. 23, 431
17 Loening, U.K, Jone, K.W. and Birnstiel, M.L. (1969) J. Mol. Biol. 45, 353
18 Hershey, A.D., Burgi, E. and Ingraham, L. (1963) Proc. Natl. Acad. Sci.
US 49, 748
19 Ritchie, D.A. , Thomas, C.A. Jr., MacIIartie, L.A. and Wensink, P.C. (1967)
J. Mol. Biol. 23, 365
20 Birnstiel, M.L., Grunstein, M., Speirs, J. and Henning, W. (1969) Nature
223, 1285
21 Boezi.J.A. and Armstrong, R.L. (1968) Methods in Enzymology, Vol.XII, A,
684, ed. Grossman L. and Moldave K.
22 Gillespie, D. and Spiegelman, S. (1965) J. Mol. Biol. 12, 829
23 Bauer, W. and Vinograd, J. (1971) Prog, in Molecular and Subcellular Biol.
Springer-Verlag 2, 181
24 Weil, R. and Vinograd, J. (1963) Proc. Natl. Acad. Sci. US 50, 730
25 Noll, H. and Stutz, E. (1968) Methods in Enzymology, Vol.XII, B, 129,
ed. Grossman L. and Moldave K.
26 Gierer, A. (1958) Naturforschung 13 b, 778
27 Timashev, S.N., Brown, R.A., Colter, J.S. and Davies, M. (1958)
Biochim. Biophys. Acta 27, 662
28 Radloff, R., Bauer, W. and Vinograd, J. (1967) Proc. Natl. Acad. Sci.
US 57, 1514
29 Studier, F.W. (1965) J. Mol. Biol. 11, 373
30 Vinograd, J., Lebowitz, J., Radloss, R., Watson, R. and Laipis, P.
(1965) Proc. Natl. Acad. Sci. US 53, 1104
2221
Nucleic Acids Research
31
32
33
34
35
36
37
38
39
40
41
42
Burton, A. and Sinsheimer, R.L. (1965) J. Mol. Biol. 14, 327
Wang, J.C. (1969) J. Mol. Biol. 43, 25
Bloomfield, V. (1966) Proc. Natl. Acad. Sci. US 55, 717
Reeder, R.H. and Brown, D.D. (1970) in Polymerase and Transcription,
Silvestri, L. ed., pp.249, North Holland Publ., Amsterdam
Kornberg, A. (1967) in Regulation of Nucleic Acid and Protein Biosynthesis, pp. 22, Elsevier Publ. Corp., Amsterdam
Shapiro, J., MacHattie, L.A., Eron, L., Ihler, G. , Ippen, K. and Beckwith,
J. (1969) Nature 224, 768
Corneo, G., Ginelli, E. and Polli F. (1967) J. Mol. Biol. 23, 619
Brown, D.D. and Stern, R. (1974) Ann. Rev. Biochem. 43, 667
Glover, D.M. and Hogness, D.S. (1977) Cell 10, 167
White, R.L. and Hogness, D.S. (1977) Cell 10, 177
Wellauer, P.K. and Dawid, I.B. (1977) Cell 10, 193
Pellegrini, M. , Manning, J. and Davidson, N.H. (1977) Cell 10, 213
2222