.1.. gen. Virol. (I975), 26, 59-69
59
Printed in Great Britain
Unidirectional Replication of a Minority of Polyoma Virus
and SV40 D N A s
By D. L. R O B B E R S O N * , L.V. C R A W F O R D, C H R I S T I N E S Y R E T T
AND A N N E L I E W I L D E J A M E S
Imperial Cancer Research Fund Laboratories, Lincoln's Inn Fields,
London WC2A3PX, U.K.
(Accepted 12 September 1974)
SUMMARY
Polyoma DNA replication is initiated predominantly at a site which is 29 ~o from
the EcoRI cleavage site. Molecules replicating from this site, after digestion with
EcoRI, appear as linear structures with a double stranded loop centred at the origin
of replication. These forms constitute 9o % of all replicating intermediates.
Approx. IO % of the replicating intermediates ofpolyoma and SV4o DNAs occur
as Y-forms after treatment with EcoRI. These structures have probably resulted
from unidirectional replication initiated at an additional origin of DNA replication
which is located near the EcoRI cleavage site on the genomes of these viruses.
INTRODUCTION
Both polyoma and SV4o DNAs can be cleaved at specific sites with the restriction endonuclease EcoRI (Morrow & Berg, 1972; Folk & Wang, I974; Robberson & Fried, I974).
EcoRI cleavage in one unique region of replicative intermediates of polyoma and SV4o
provides a basis for the physical mapping of the origin and direction of DNA replication.
The predominant mode of DNA replication that has been detected in these viruses corresponds to bidirectional synthesis initiated at sites which are 29 and 33 % of the distance
from the EcoRI cleavage sites on polyoma (Crawford, Syrett & Wilde, I973) and SV4o
(Fareed, Garon & Salzman, I972) DNAs, respectively. The EcoRI cleavage sites in the two
DNAs are probably not in functionally equivalent regions so that the similarity of the two
values, 29 °/o and 33 %, may be fortuitous. In both cases, the two forks in the bidirectional
replicating molecules grow at approximately equal rates with termination of DNA synthesis
at a site which is diametrically opposite the initiation site. In addition to these major populations of Cairns forms, the pool of replicative intermediates contains minor species of
replicating molecules which have the formal appearance of the letter Y, after cleavage with
EcoRI. Y-forms were previously reported to constitute approx. IO % of the replicating intermediates of SV4o DNA (Fareed, et al. I972). In this earlier study these structures were
interpreted as replicative forms cleaved by EcoRI which had also been broken at one of
the replication forks. An alternative explanation for this type of replication intermediate
would be that it arises from unidirectional replication initiated at an origin near the EcoRI
cleavage site.
* Present address: Department of Biology, M. D. Anderson Hospital and Tumor Institute, Houston,
Texas 77025, U.S.A.
6o
D . L . R O B B E R S O N ~ L. V. C R A W F O R D ~ C. S Y R E T T A N D A. W I L D E J A M E S
METHODS
Preparation of replicating DNA. Stocks of small plaque and large plaque polyoma virus
were used to infect mouse embryo fibroblast secondary cells as previously described (Crawford et al. I973). A stock of SV4o virus (SVL) was used to infect BSC-I cells. These stocks
had all been plaque-purified and grown at low multiplicity. In all cases, the infection was
performed at a multiplicity of Io to 2o p.f.u./cell. At about 3o h after infection, intracellular DNA was extracted by the Hirt procedure (Hirt, I967) and banded in ethidium bromide
CsC1 gradients as described by Radloff, Bauer & Vinograd (I967). The gradients
were then fractionated into lower, intermediate and upper band portions. After removal of
ethidium bromide by extraction with isopropanol, the DNA collected from the intermediate
region of the ethidium bromide CsC1 gradients, enriched for replicative intermediates,
was digested with EcoRI as previously described (Crawford et al. I973). The endonuclease
was prepared by a modification (Folk & Wang, I974) of the method of Yoshimori (i971).
Some properties of this preparation of EcoRI have already been described (Robberson &
Fried, i974). Non-replicating samples of SV4o and polyoma DNAs were prepared in the
same manner and correspond to lower band samples isolated from cells 48 h after infection.
Prior to EcoRI cleavage, these latter DNA samples contained less than 1% of the molecules
as Cairns forms of the type previously described (Hirt, I969).
Electron microscopy. Samples for electron microscopy were prepared by a modification
of the Kleinschmidt basic protein film technique (Davis, Simon & Davidson, 1971). Grids
were rotary shadowed with platinum:palladium and examined in a Siemens IOI electron
microscope. Micrographs were taken at an instrumental magnification of ~o ooo and enlarged
approx. 2o times. Length measurements on tracings of the molecules were performed with
a map measure.
RESULTS
The predominant topological form apparent in the replication of polyoma and S¥4o
DNAs is a duplex circular molecule containing a single duplex circular loop (Fig. I a and 2).
As replication proceeds, the size of this loop increases with each of the two replication
forks moving at approx, equal rates. This loop is centred approx, one-third of the molecule
from the region within which the EcoRI cleavage occurs, as measured on EcoRI treated
SV4o and polyoma DNAs. These forms constitute approx. 9o % of the replicative intermediates of SV4o and polyoma DNAs (Table 1). The remainder of the replicative forms
after cleavage by EcoRI consist largely of single-fork molecules with a formal appearance
that resembles a Y (Fig. 3 and Table 1). The Y-forms that occur in SV4o and polyoma
DNAs possess two duplex arms with approximately equal lengths (Fig 4a) which appear to
extend throughout the length of the virus genome (Fig. 4 b) These structures may result from
the EcoRI digestion of unidirectionally replicating molecules initiated near the cleavage
sites (Figs. r e, d). Y-forms are rarely seen in preparations of non-replicating DNA (see
Methods). Those that are seen have presumably arisen through fortuitous sticking of linear
fragments in the course of EcoRI digestion and specimen preparation for electron microscopy. These non-replicating structures occur as Y-forms with two arms of unequal lengths
(Fig. 4c).
The results of a linear least squares analysis of measurements for the segments of SV4o
and polyoma Y-forms in Fig. 4 b is presented in Table 2. If Y-forms have arisen through
unidirectional replication events initiated at origins within the region that is subsequently
cleaved by our preparation of EcoRI, then we would expect length measurements on the
Unidirectional replication o f virus D N A s
6i
II
0 /
d
0 /
Fig. I. Diagrammatic drawings of replicative forms before (on the left) and after (on the right)
cleavage with EcoRI. The region within which each molecule is cleaved by the EcoRI preparation
is indicated by the two vertical lines in (a). (a) Bidirectional replication initiated approx, one-third
of the molecule from the EcoRI site. (b) Bidirectional replication initiated near the EcoRI site. (c)
Unidirectional replication initiated near the EcoRI site that proceeds to the right. (d) Unidirectional
replication initiated near the EcoRI site that proceeds to the left. (e) Bidirectional replication initiated
as in (a) plus unidirectional replication initiated as in (c). (f) Bidirectional replication initiated as in
(a) plus unidirectional replication initiated as in (d). (g) Bidirectional replication initiated at both
origins O and O'.
5
VIR
26
62
D.L. ROBBERSON, L. V. CRAWFORD, C. SYRETT AND A. W I L D E JAMES
Fig 2. Examples of bidirectional replicating molecules of polyoma before (on the left)
and after(on the right) cleavage by EcoR1. Magnification: u~m = 3q cm.
Table I. Frequencies of polyoma and SV40 replicative forms after treatment with EcoRI
% of replicating
intermediates as
Bidirectional
forms*
Unidirectional
forms?
Double initiation
forms:~
Molecules classified
Polyoma
(large plaque)
86 (IO)
Polyoma
(small plaque)
87 (I2)
SV4o
89 (8)
13
12
I0
I
I
I
252
143
203
* The predominant structure classified here contains a loop located approx, one-third of the genome from
the EcoRI site (Fig. 2). The numbers of terminal loop molecules (Fig. 5) are given in parentheses. These
have been included in the totals for bidirectional replication together with a small number of double Ymolecules (Fig. 6).
t Presumed unidirectional replicative forms initiated near the EcoRI site.
:~Replicative forms which appear to have been initiated at the predominant origin for bidirectional
synthesis and at a second origin positioned near the EcoRI site (Fig. 7 to 9).
presumed replicated segments of these single-fork molecules to conform to a linear relationship that intercepts within approx. 5 % of the measured position of the EcoRI cleavage site.
In replicating samples of polyoma and SV4o DNA, the lengths of the replicated arms of the
Y-forms are fitted to linear relationships with slopes and intercepts that are close to those
expected for unidirectional replication initiated near the EcoRI site. In contrast to this
rather close agreement found for replicating Y-forms, measurements on the non-replicating
Y-forms do not fit a unique linear relationship. This is evident in the large error in the
slopes determined for lines that do not intercept the expected position for EcoRI cleavage
( a t a = I . o ) (Fig. 4c and Table I).
In addition to Y-forms we have noted the occurrence of linear replicative forms which
terminate in a duplex circular loop or bubble (Fig. 5). These have been included in our
scorings as part of the population of bidirectional replicating molecules initiated at the predominant origin, although they may equally well have come from unidirectional replication
(Table I). The contour length of the loop in these terminal loop molecules was seldom
equal to one genome. About half of these forms had total lengths close to one genome,
as though they were non-replicating D N A in which one end of the unit length molecule
had become looped back. The other half were probably replicating molecules since, in these
molecules, half the length of the loop plus the length of the tail equalled one genome.
These could have come from unidirectional or bidirectional replicating forms in which one
fork was near the EcoRI site.
Bidirectional replicative forms initiated near the EcoRI site will, after cleavage by EcoRI,
give rise to double Y-molecules in which all the replicated arms are the same length (Fig.
I b and 6). Only two such replicative forms were detected in the pool of replicative molecules
63
Unidirectional replication o f virus D N A s
(a)
(b)
Fig. 3. (a) Y-forms of replicating small plaque polyoma virus D N A treated with EcoRI. Magnification: ~ # m = 3"4 cm. (b) Y-forms of replicating SV4o D N A treated with EcoRI. Magnification:
I ]~m = 4 ' 2 cm.
in polyoma and SV4o examined ;n this study. These forms are to be contrasted with the
double Y-molecules derived by cleavage of bidirectional replicating molecules initiated at the
predominant origin of replication described for polyoma and SV4o DNAs in which the
bidirectional replication event has proceeded past the EcoRI cleavage site (Fareed et al. ~972;
Crawford et al. 1973) (Fig. 6).
If origins for DNA replication do in fact exist at two positions on the virus genomes,
one might also expect to find examples of replication forms in which initiation had
occurred at both origins of the same molecule. The topological form of these replicative
intermediates expected after cleavage with the EcoRI preparation is shown in Fig. ~e, f.
Examples of replicative forms with evidence of this type of double initiation were detected
among the replicative intermediates of both SV4o and polyoma, and comprised r % of all
replicative forms (Fig. 7, 8 and 9). Unidirectional replication apparently proceeds to either
the left or right of the minor origin on these molecules. Replicative forms containing evidence
of bidirectional synthesis initiated at both origins of the same molecule (Fig. I g) were not
detected in this study.
Virus preparations, passed at high multiplicity, accumulate defective DNA molecules in
which sequence rearrangements have occurred (Tai et al. 1972; Brockman & Nathans, I974;
Folk & Wang, I974; Robberson & Fried, ~974). If these rearrangements brought an origin
of replication near to the EcoRI site this could lead to the production of Y-forms, from the
5-2
64
D.L. ROBBERS•N,
L. V. C R A W F O R D ,
C. S Y R E T T A N D A. W I L D E
A
•
i
~
0.8
(a)
(b)
•
" O ~--
JAMES
"4:~3.t:j O • . _
[]
o
B
g 0-8~-[] °
0-4
[]
o
C
•
to-dr
o
0.4
"a~'°"2
~'6''0°-"
°
• .o
0.8
r-
1 1
/
u
]
I
I
0-2
I
I
0.4
I
1~1
0.6
i
1'0
0.8
0
o
a
•
0.8
(c)
a
[]
[]
•
•
•
[]
B 0"4
0.4
D
[]
•
[]
0
0
0
0.8
•
[]
I
0
1
0.2
•
[]
I
I
[]
I
0-4
I
0.6
I
I
0,8
I
1.0
A
Fig. 4. Length measurements of replicating and non-replicating Y-forms. (a) Drawing of Y-form
with arms designated A, B, and C. (b) Selected Y-forms in which the sum of measured lengths for
either A plus B or A plus C was within I 1% of the average length measured for bidirectional replicative forms cleaved by EcoRI (Fig. z) that were present in the sample. Measurements were made
on Y-forms detected among the replicative intermediates of large plaque polyoma virus (O), small
plaque polyoma virus (@) and SV4o ( ~ ) . The dashed lines indicate the expected lengths for arms B
and C for unidirectionalreplication initiated at the EcoRI site. (c) Selected Y-forms in non-replicating
samples (see Methods) of large plaque polyoma virus (O), small plaque polyoma virus (©) and SV4o
([]). The sum of measured lengths for A and B or A and C on each Y-form was within ~1% of the
averagelength of the polyoma or SV4o EcoRI linear D N A molecules present in the sample. Molecules
with lengths which deviated by more than I z % from that of EcoRI linear molecules have not been
included here. These comprised of 8 molecules of large plaque polyoma, 4 molecules of small
plaque polyoma and 4 molecules of SV4o.
Unidirectional replication of virus DNAs
Table 2.
Linear least squares analysis of length measurements
on segments of single fork molecules*
Slope*-~
Sample
Replicating polyoma
(large plaque)
Replicatingpolyoma
(small plaque)
Replicating S V 4 o
Non-replicating
SV4o and polyoma
65
c B/A
-- C/A
Intercept
~Bo~
Co~
-BoA/B§
-CoA/C~
No. of
molecules
o'94_+o'I2 o-87+ o.I6
0.95+_0.08 o.9r +o.Io
l.oi
I.O5
I4
o'86_+0.22 o'93+o'I7
0"94+_0"09 0"95+0"07
I-O9
I.O2
13
o'96_oq6 o'92+o'I5
o'54±o'42 o'52 +o-39
I.o+- o'Io o'98 +-o'o9
o-91+o.z2 o-91+-0.20
1-o4
1-68
I-O6
I"75
I8
I5
* Calculated from data presented in Fig, 4b, c.
?The error in the slope indicates the interval that contains the true mean at a level of confidence of 90 %.
:[: Measured intercepts and standard deviations on vertical coordinates of Fig. 4b, c.
§ Calculated intercepts on horizontal coordinates of Fig. 4b, c.
Fig. 5. Replicating forms of EcoRI-treated small plaque polyoma virus DNA which terminate
in a duplex circular loop. Magnification: i/zm = 3"4 cm.
Fig. 6. Double Y-molecules with all arms of similar lengths (on the left) and with pairs of arms
of dissimilar lengths (on the right). The molecule on the left was observed in a sample of replicating
large plaque polyoma virus and the molecule on the right was observed in a sample of replicating
small plaque polyoma virus. Magnification: ~#m = 3"4 cm.
E c o R I digested replicating molecules. Defective D N A molecules were therefore considered
as a possible source of Y-forms. Since most defective p o l y o m a D N A are resistant to E c o R I
( F o l k & W a n g , I974; Fried, I974), the fraction of a D N A p r e p a r a t i o n which is resistant to
E c o R I is a good indication o f its c o n t e n t of defective D N A . Only 1 % of the non-replicating
large plaque p o l y o m a D N A p r e p a r a t i o n examined was E c o R I resistant (Table 3). Such
E c o R I resistant molecules are therefore unlikely to have contributed significantly to the
66
D.L. ROBBERSON,
L. V. C R A W F O R D ,
C. S Y R E T T
AND
A. W I L D E
JAMES
Fig. 7- Molecules o f replicating SV4o D N A treated with EcoR1 containing a loop c o r r e s p o n d i n g to bidirectional synthesis a n d a single-fork Y-structure c o r r e s p o n d i n g to unidirectional synthesis initiated n e a r
the E c o R I site. Magnification: I # m = 3"9 cm.
Fig. 8. T h e molecule at the top was observed in replicating large plaque p o l y o m a D N A treated with
E c o R I a n d contains a loop corresponding to bidirectional synthesis as well as a single-fork Y-structure c o r r e s p o n d i n g to unidirectional synthesis initiated near the E c o R I site. T h e molecule at the
b o t t o m was observed in replicating small plaque p o l y o m a virus a n d represents a Y - f o r m , one a r m
o f which contains a loop indicative o f a second initiation event in the replication o f this molecule.
Magnification: I # m = 3"9 cm.
Fig. 9, Molecules o f replicating SV4o D N A treated with E c o R I in which two initiation events appeal
to have occurred. T h e molecule o n the left contains a single fork Y-structure a n d a loop which is no
at a position c o r r e s p o n d i n g to initiation of bidirectional synthesis. T h e molecule o n the righL
contains a loop corresponding to bidirectional synthesis a n d a small single-fork Y-structure. Magnification at ]eft: I # m = 4'2 cm. Magnification at right: I # m = 3"4 cm.
Unidirectional replication o f virus DNAs
67
Table 3. Examination of large plaque polyoma virus DNA for defective genomes
%duplex molecules as
Before first
After first
After second
After
EcoRI
digestion
EcoRI
digestion
EcoRI
digestion
denaturationrenaturation
Circular forms
Closed
Open
85"7
I2.8
o-8
o'7?
Linear forms
Without inhomology region
With inhomology region
Molecules classified
I'5
943
98"5
xoo6
o'7 (I-I)*
I.I (o.I)
0"6
o'5-~
98'2 (98"8)
940 (820)
98"7
o.2§
1959
*Numbers in parentheses and in the last column on the right refer to frequencies of molecules determined
on samples prepared by the formamide modification of the Kleinschmidt technique (see Methods).
t The majority of these forms are hydrogen bonded circles of EcoRI cleaved DNA.
++These forms are either renatured defective DNAs that had been randomly nicked in the preparations
or defective DNAs nicked in the course of the renaturation itself.
§ The region of inhomology in these renatured molecules occurs at one terminus and represents less than
lO% of the genome length.
population of Y-forms described. Defective DNAs which are sensitive to cleavage by EcoRI
and in which a sequence rearrangement had brought an origin of D N A replication close
to the EcoRI site would be expected to display a corresponding region of inhomology at this
site in heteroduplexes with non-defective virus DNA. A search for such heteroduplexes after
denaturation-renaturation of the EcoRI-treated sample described above revealed the
presence of only o-2 % of the renatured molecules with any evidence of a region which had
not hybridized (Table 3).
DISCUSSION
Replication of polyoma and SV4o DNAs is normally bidirectional, starting from a site
about one-third of the genome from the EcoRI cleavage site. This is based on examination
of replicating molecules cleaved by EcoRI (Fareed et al. I972; Crawford et al. 1973) and
on pulse-labelling of the DNA, followed by digestion with restriction enzymes (Danna &
Nathans, 1972; Crawford, Robbins & Nicklin, 1974) The results of pulse labelling experiments
would not be significantly affected by a tow frequency of replication initiated elsewhere in the
DNA.
A unidirectional mode of replication initiated near the EcoRI site is now suggested to
account for the appearance of a minority of the replicative intermediates of both SV4o and
polyoma. After treatment with EcoRI, these molecules appear as Y-forms in which the two
arms of the Y have approximately equal lengths. This type of Y-form is essentially absent
from samples of non-replicating DNA. Furthermore, initiation of D N A replication appears
to have occurred at both origins in a few of the replicating molecules examined. It should be
pointed out that since this study is based entirely on the appearance of the D N A in the
electron microscope there is no direct proof that these were replicating molecules. However,
the most likely explanation of their appearance is that these molecules were in the process
of replication when isolated. Under conditions similar to those used here, this preparation
of EcoRI cleaves large plaque polyma D N A to linear forms with expression of additional
nuclease activities (Robberson & Fried, i974). The additional nucleases appear to act
preferentially to one side of the EcoRI site on polyoma D N A and within a distance of
approx. 5 % of the EcoRI site. Thus, this postulated second origin of D N A replication is
68
D.L. ROBBERSON, L. V. CRAWFORD, C. SYRETT AND A. WILDE JAMES
expected to occur within a distance of approx. 5 % from the EcoRI site on polyoma DNA.
The occurrence of Y-forms in replicating SV4o D N A treated with EcoRI implies a similar
pattern of additional nuclease digestion as observed for polyoma D N A and perhaps something special about the EcoRI site itself.
Since unidirectional replication from a site near the EcoRI cleavage site has not previously
been suggested for these DNAs, it is important to consider alternative explanations for these
results. The explanation given by Fareed et al. (1972) for the Y-forms they observed was that
they were molecules broken at one of the replication forks and cleaved by EcoRI. This
would account for molecules which were less than full length but not for full length molecules.
Most of the molecules discussed here were within I r % of the full length and the two arms of
the Y were approx, equal (Fig. 4)- These could only be produced from replicating molecules
which were not cleaved by EcoRI but which were broken simultaneously on both sides of a
fork. It seems unlikely that this would occur often unless the EcoRI preparation contains
an enzyme with a preference for the fork itself. Breakage on only one side of the fork would
seem much more likely and molecules in which this had occurred were seldom seen, either
here or previously (Fareed et al. I972).
Some of the Y-forms could have arisen through cleavage of replicating circular dimeric
molecules of SV4o and polyoma DNAs if the EcoRI sites on the monomer genomes of
these dimers are arranged appropriately. Approx. 1% of the intracellular forms of both
SV4o (Rush, Eason & Vinograd, i97 I) and polyoma (our examination of 5oo molecules)
occur as circular dimers. Total replicative forms constitute about 1% of the intracellular
DNA. Thus, the frequency of circular dimers is considerably lower than the frequency of the
Y-forms reported here and they are not likely to be the source of this particular replicating
intermediate. Similarly, defective molecules either sensitive or resistant to EcoRI, are not
frequent enough to account for the number of Y-forms seen. None of these alternative
explanations seem satisfactory and we therefore conclude that the Y-forms arise from
unidirectional replication initiated near the EcoRI site. This unusual mode of replication
might, when abortive, lead to the formation of defective DNAs as suggested recently
(Robberson & Fried, i974).
Note added in proof.
An examination of replicating SV4o D N A molecules cleaved by endonuclease HpalI,
carried out with the help of' Paul Nicklin, has now provided further evidence for a unidirectional mode of replication. Endonuclease HpaII cleaves SV4o D N A once, 74 % from
the EcoRI site (Sharp, Sugden & Sambrook, 1973). This therefore avoids the complications
caused by the use of a cleavage site close to the apparent minor origin of replication. Molecules in which replication was initiated as the predominant origin (67 % from the EcoRI
site) are mostly cleaved to double Y forms. Measurement of thirty such molecules showed
that the origin was 7 -+2 % (74 minus 67) from the HpalI cleavage site, as expected.
Molecules in which replication had been initiated at a position away from the normal
origin may be converted to forms with a single duplex circular loop by cleavage with HpaII
(Figs. Ia and 2). Measurement of seventeen such molecules showed that in fourteen of
them one end of the loop was located at 26 + 4 % from the end of the molecule. This is
consistent with this fork being close to the EcoRI cleavage site (26 % from the HpaII site).
In eleven of the molecules replication appeared to be going away from the HpaII site i.e.
clockwise on the map as normally drawn, and in three molecules towards the HpalI site. In
the remaining three molecules the position of the internal loop was not consistent with
Unidirectional replication of virus DNAs
69
unidirectional replication as above or bidirectional replication from the normal origin.
Thus the configuration of fourteen out of seventeen of these internal loop molecules was
consistent with their coming from unidirectional replication starting at a position 26 _+4 %
from the HpaII site, i.e. within 4 % of the EcoRI site.
We are indebted to Alan Robbins for his help with preparation of virus DNA. We would
a l s o like t o e x p r e s s o u r g r a t i t u d e t o D r J. S i n k o v i c s f o r t h e u s e o f h i s e l e c t r o n m i c r o s c o p e
l a b o r a t o r y , t h e g i f t o f M r a n d M r s L. L i p p m a n , i n t h e c o m p l e t i o n o f t h e s e s t u d i e s .
REFERENCES
BROCKMAN, W. W. & NATHANS, D. (I974). The isolation of simian virus 4o variants with specifically altered
genomes. Proceedings of the National Academy of Sciences of the United States of America 71, 942-946.
CRAWFORD, L. V., ROBBINS, A. K. & NICKLIN, P. M. (I974). Location of the origin and terminus of replication
in polyoma DNA. Journal of General Virology 25, I33-I42.
CRAWFORD, L. V., SYRETT, C. & WILDE, A. (I973). The replication of polyoma DNA. Journal of General Virology
2 I , 515--52I.
DANNA, K. J. & NATHANS, D. (I972). Bidirectional replication of simian virus 4o DNA. Proceedings of the
National Academy of Sciences of the United States of America 69, 3o97-3IOO.
DAVIS, R. W., SIMON, M. & DAVIDSON, N. (t97I). Electron microscope heteroduplex methods for mapping
regions of base sequence homology in nucleic acids. Methods in Enzymology 21,413-428.
FAREED, G. C., GARON, C. F. & SALZMAN, N. P. 0972). Origin and direction of simian virus 40 deoxyribonucleic
acid replication. Journal of Virology io, 484-49 I.
FOLK, W. R. & WANG, H-C. E. (1974). Closed circular DNAs with tandem repeats of a sequence from polyoma
virus. Virology 61, 14o-I55.
FRIED, M. (I974). Isolation and partial cbaracterisation of different defective D N A molecules derived from
polyoma virus. Journal of Virology i3, 939-946.
HIRT, B. (1967). Selective extraction of polyoma D N A from infected mouse cell cultures. Journal of Molecular
Biology 26, 365-369.
HIRT, B. (I969). Replicating molecules of polyoma virus DNA. Journal of Molecular Biology 40, I4I-I44.
MORROW, J. V. & BERG, P. (I972). Cleavage of simian virus 40 D N A at a unique site by a bacterial restriction
enzyme. Proceedings of the National Academy of Sciences of the United States of America 69, 33653369.
RADLOFF, R., BAUER,W. & v1NOGRAD, J. (1967). A dye-buoyant density method for the detection and isolation
of closed circular duplex D N A : the closed circular D N A in HeLa cells. Proceedings of the National
Academy of Sciences of the United States of America 57, I514-I52I.
ROBBERSON, D. & FRIED, M. (1974). Sequence arrangements in clonal isolates of polyoma defective DNA.
Proceedings of the National Academy of Sciences of the United States of America 7 I, 3497-35Ol.
RUSH, M. G., EASON, R. & VINOGRAD, J. 0971)- Identification and properties of complex forms of SV4o D N A
isolated from SV4o infected African green monkey (BSC-I) cells. Biochimica et Biophysica Acta 228,
585-594.
SHARP, P. A., SUGDEN, B. & SAMBROOK, J. (I973). Detection of two restriction endonuclease activities in
Haemophilus parainfluenzae using analytical agarose-ethidium bromide electrophoresis. Biochemistry
i2, 3055-3063.
TAI, H. T., SMITH, C. A., SHARP, P . A . & VINOGRAD, J. (1972). Sequence heterogeneity on closed SV4o
deoxyribonucleic acid Journal of Virology 9, 317-325.
VOSHIMORI, R. N. (I971). A genetic and biochemical analysis of the restriction and modification of D N A by
resistance transfer factors. Ph.D Thesis, University of California, San Francisco Medical Center.
(Received 2 4 June 1974)
