Proc NatL Acad. Sd. USA
Vol. 78, No. 1, pp. 69-73. January 1981
Biochemistry

Unique primed start of phage 4x174 DNA replication and mobility of
the primosome in a direction opposite chain synthesis

(origin of complementary strand replication/protein n'/mobile replication promotor/antielongation direction/DNA-dependent ATPase)

KEN-ICHI AMI* AND ARTHUR KORNBERG

Department of Biochemistry. Stanford University School of Medicine, Stanford, California 94305
Contributed by Arthur Kornberg, Septemberl. 1980

ABSTRACT         ent of the 4x174 viral circle sus-
tains the primed start of compEentary DNA strand synthesis in
oitro, even though the intact circlepermits primed starts at many
sites. The 3Wnucleotide fragment rom restriction nuclease di es
tion contains the recognition site for protein n', a DNA-depentfen;
ATPase essential for priming 4x174 DNA replication. This n' rec-
ognition site contains within it a 44-nucleotide se uence with a po-
tential hairpin structure and maybe regarded as ftre startin signal
for replication [Shlomai, J. & Kornberg, A. (198O)Proc Natf Acod
Sci USA 77, 799-803]. After initiation on the 3' side of this se-
quence, the priming system (primosome) repeatedly generates
primers by moving processively on the DNA template in a direc-
tion opposite to chain elongation. This primosome mobility is an
attractive model for the discontinuous phase of Escherichia coli
chromosome replication, in which processive primosome move-
ment with the replicating fork is proposed for repeated initiations
of nascent replication fragments.

A specific fra

Conversion of single-stranded DNA (ssDNA) of phage 4x174
(4X) to duplex replicative form (RF) is a model system (1-3) for
the discontinuous phase of Escherichia coli chromosomal repli-
cation in which the nascent replication (Okazald) bagments are
initiated. In this conversion, 4X DNA coated by ssDNA-bind-
ing protein (SSB) must first be activated in a prepriming stage
for subsequent priming by primase (2, 4-6). In this prepriming
reaction, E. coli proteins n, n', n", i, dnaC, and dmB interact to
form a prepriming replication intermediate (4, 5, 7). Unlike
phages M13 (8,9) and G4 (10-12), which possess unique origins
for priming complementary strand replication, 4X permits mul-
tiple starts in DNA replication, as indicated by both in oiuo (12)
and in oitro studies (5, 7). However, the strict specificity of the
prepriming system for 4X DNA (7, 13) and the specific recog-
nition by protein n' of a 55-nucleotide sequence in c$X DNA,
located in the same intergenic region as the G4 origin (14),
strongly suggest a unique origin for 4X complementary strand
replication.
This present work shows that complementary strand replica-
tion is in fact initiated at or near the protein n' recognition locus
by a multiprotein-DNA complex (a mobile replication promoter
or "primosome"), which then migrates with a unique polarity on
the ssDNA template in a direction opposite to primer and DNA
chain synthesis. Subsequent reports will describe participation
of ATP in the processivity of primosome movement and conser-
vation of the primosome in successive stages of @X DNA repli-
cation. These studies ofprimosome structure and function have
implications for events taking place at the E. coli chromosome
replication fork.

       MATERIALS AND METHODS
Nucleic Acids and Enzymes. DNAs, E. coli replication pro-
teins, and other materials were as described (13-16). Restriction

The publication costs of this article were defrayed in part by page charge
payment. This article must therefore be hereby marked "duertipe-
ment" in accordance with 18 U. S. C. 11734 solely to indicate this fact.

endonuclease fragments of ssDNA (17, 18) for use as templates
for DNA synthesis were prepared by digestion of 4X or G4
DNA (30 pg) at 37°C for 15 hr with Hue 111 (400 units) or Hha I
endonuclease (320 units) in 400 pI of 50 mM Tris-HCI (pH 7.5)/
5 mM MgClJ0.5 mM dithiothreitol. Digestion products were
precipitated with ethanol from phenol-treated reaction mix-
tures and dissolved in 50 mM Tris-HC1 (PH 7.5)/1 mM EDTA.
Buffer A was 100 mM Tris-HCI (PH 7.5)/20% sucrose/40 mM
dithiothreito1/200 pg of bovine serum albumin per ml.
DNA Replication Assay. Components were added in order at
0°C and incubated 20 min at 30°C: 5 pI of buffer A, 1.2 nmol
each of [3H] or [cK-~'P]~CTP, dGTP, dATP, and dTTP (each at
2000 dpm/pmol), 2.5 nmol each of GTP, CTP, and UTP, 20 nmol
of ATP, 0.2 pmol of MgCl,, 450 pmol (as nucleotide) of 4X
ssDNA or its digestion products, 0.25 pg of rifampicin, SSB as
indicated, 0.2 pg of DNA polymerase 111 holoenzyme, 0.4 pg
of dmB protein, 0.1 pg each of dnaC, i, and n' proteins, 0.14
pg of protein mixture n + n", 0.1 pg of primase, and water to
25 pl. With G4 DNA or its digests as templates, only SSB, DNA
polymerase 111 holoenzyme, and primase were included.
Other Methods. Agarose gel electrophoresis of DNA was as
described (16, 19). The relative amount of 32P radioactivity in
each band was determined by densitometric tracing of the auto-
radiogram with a Quick Scan Jr. TLC, Helena Laboratories
(Beaumont, TX). Sizes of DNA products were determined after
heat denaturation in 98% (wt/vol) formamide by electrophore-
sis in a 7 M urea/2.5-7.5% gradient polyacrylamide gel (5,20).
Transfer of DNA to diazobenzyloxymethyl (DBM) papert and
DNA hybridizations were as described (21.22).

RESULTS

Unique primed start of DNA replication

Specific DNA Synthesis on a C4 DNA Fragment. Whether
a unique primed start of DNA replication can be sustained by a
fragment of a phage DNA circle was first determined with G4
DNA. A Hha I endonuclease digest (15 fragments) of G4 DNA
(23) supported DNA synthesis with SSB, primase, and DNA
polymerase 111 holoenzyme at 18% the level of untreated G4
DNA (Table 1). Omission of primase abolished almost 80% of
this activity. Agarose gel electrophoresis of the G4 DNA prod-
ucts showed that more than 90% ofthe DNA was synthesized on
the 1494-nucleotide Hha I fragment 1, which contains the origin

~

Abbreviations: +X, phage 4x174; ssDNA, single-stranded DNA; RF,
double-stranded DNA in the circular replicative form; SSB, ssDNA-
binding protein; DBM, diazobenzyloxymethyl; p[NH]ppA, J'denyl-
yl imidodiphosphate.
*Present address: Department of Chemistry, The Institute of Medical
Science, The University of Tokyo, 4-61. Shirokanedai. Minato-ku,
Tokyo 108, Japan.
t In some experiments, DNA fragments were transferred to aminothio-
phenol paper, a procedure suggested by Brian Seed of the California
Institute of Technology.

69

70 Biochemistry: Arai and Kornberg

Proc. NatL Acad. Sci. USA 78 (1981)

Table 1. Template activity ofHae III and Hha I endonuclease

digests of phage DNA

           DNA
        synthesis,
Riming pmol

Nuclease
treatment
oftemplate wmponent(s) G4 &X

None


Hae III


Hh I

+ 290   360
- 8    11
+ 4    39

        2 18
+ 51 18
       12 5

-

-

DNA synthesis was with 1.0 pg of SSB. The priming component for
the G4 templates was only primaee; the priming wmponents for &X in-
cluded the prepriming proteins n, n', n", i, dnaB, and dnaC, as well as
primaee.

of G4 complementary strand replication (10, 23) (Fig. 1). The
extent of synthesis was nearly what was expected from the loca-
tion of the origin within fragment l and its relative size.
No DNA synthesis was detected with a Hue 111 endonuclease
digest (14 fragments). Although the 442-nucleotide fragment
Z5a contains the G4 complementary strand origin, the cleavage
site is at the stem of the "downstream" hairpin located about
60 bases downstream from the start of the primer and presum-
ably required for recognition by primase (24). These results in-
dicate that for primase the circular form of DNA is not essential
and that primase can recognize the specific signal for the com-
plementary strand origin of G4 DNA even in a small DNA
fragment.
DNA Synthesis on a Unique Fragment of q5X DNA Requir-
ing the Prepriming Proteins. DNA synthesis with an unfrac-
tionated 4X DNA Hoe 111 digest (25) was 11% that with intact
+X viral circle (Table 1). Omission of prepriming proteins n, n',
n", i, and dnuC abolished only halfthe activity, suggesting some
nonspecific initiation on DNA fragments. The effects of SSB on
specific initiation dependent on prepriming proteins, and on

i1h.b I
c_:

!?IO. 1. Autoradiogram of DNA prcducta synthesized on phage
DNA Hae JII or Hha I endonuclease fragmenta. DNA synthesis was
with saDNA or ita Hue III or Hha I digest: G4 with 1.5 pg of SSB, &X
with 4 lrg of SSB. The DNA products were separated by electrophoreeis
in a 1.5% agarose gel and detected with ethidium bromide (arrows) and
autoradiography.

SSB, M?

FIG. 2. Effecte of SSB on specific and nonspecific DNA synthesis
with +X DNA (A) and ita Hae UI endonuclease fragments (B1 as tem-
plates. Reaction mixtures included only dnaB protein, primase, and
DNA polymeraae III holoenzyme (o), or also prepriming proteins n, n' ,
n", i, dnaC. and dnaB (0).

nonspecific initiation obtained with only dnaB protein and pri-
mase (13), were examined with intact 4X circular DNA and its
Hoe 111 fragments. Nonspecific initiation on the intact circle was
inhibited almost completely by SSB (Fig. M), whereas that on
the fragments was more resistant to SSB; an 80% inhibition re-
quired a 4-fold excess of SSB beyond that needed to coat the
ssDNA (26) (Fig. 28). At this high level of SSB, 85% of the DNA
synthesis was dependent on prepriming proteins; the nonspe-
cific DNA synthesis required only DNA polymerase 111 holoen-
zyme (data not shown).
Of the DNA produced with only dnaB protein, primase,
DNA polymerase 111 holoenzyme, and 1.5 pg of SSB, 19%,
%%, and 53% was synthesized on the Z1,23, and 24 fragments,
respectively (Fig. 3, trace a). When prepriming proteins n, n',
n", i, and dnaC were included, these ratios changed to 53%.
338, and 14% (Fig. 3, trace b), suggesting that prepriming pro-
teins stimulate specific initiation on the Z1 fragment and sup-
press nonspecific initiation on the Z4 fragment. With a 4-fold
excess of SSB (3.6 pg) over that needed to coat the ssDNA (26).
DNA synthesis on the 23 and ZA fragments was preferentially
inhibited; 822, 12%, and 6% of the DNA was synthesized on
the Z1, 23, and Z4 fragments, respectively (Fig. 1 and Fig. 3,
trace c). Other Hue 111 fragments were completely inert. These
results demonstrate that SSB inhibits nonspecific DNA synthe-
sis and that the 1353-nucleotide Z1 fragment is specifically ac-
tivated as a template for DNA replication by prepriming pro-
teins and primase.

24

Y "i "1"

ad
bL
L Electrophoresis

FIG. 3. ['%']DNA products synthesized on Hoe III endonuclease
fragmenta of &X DNA. DNA synthesis wan with 450 pmol (as nuclm
tide) of Hae III digests of &X ssDNA, including in trace a only dnaB
protein, pnmaae, DNA polymeraae XII holoenzyme, and 1.5 pg of SSB
hce b, prepriming proteins n, n', n", i, and dnaC, as well as the pro-
teins in trace a; and trace c, the proteins in trace b except that 4 pg of
SSB wan used. DNA producte were fractionated by electrophoresis in
1.5% agarose gel and quantitated by densitometric tracings of '9 au-
toradiograms. Arrows indicate the Hoe III fragments.

Biochemistry: Arai and Kornberg

Proc. Natl Acad. Sci. USA 78 (1981)

71

. 1 r'<?L : ,1
. .it. -58,. iiir

- ,t*i

FIG 4. Phymcal map of I#X DNA. The Hae JII and Hha I reetriction
. endonudeage cleavage   and the origin of viral DNA strand reg
lication at nucleotide 4306 tpIr I site = 0) on the hnpr map are taken
from refs. 25 and 27. The protein n' recopitim site ia located at nu-
cleotides 2301-2351 (14). The open mw indicates the DNA elonga-
tion direction.

4X Complementary Strand Replication Origin and the Pro-
tein n' Recognition Locus. Hha I nuclease cleaves @ DNA into
17 fragments (s), of which the 300-nucleotide fragment 6 con-
tains the 55-nucleotide recognition locus of protein n' (Fig. 4).
Hha I digests of #X DNA supported DNA synthesis in the pres-
ence of prepriming proteins at 5% the level with untreated
DNA (Table 1). Fragment 6 accounted for 80% of the DNA in-
corporation as judged by agarose gel electrophoresis (Fig. 1)
and by hybridization to various regions of +X DNA (see below);
about 20% was related to a fragment from incomplete digestion
of ~#IX DNA that migrated between Hha I fragments 1 and 2 and
contained sequences offragments 3 and 6 (see below). Omission
of the prepriming proteins almost abolished DNA synthesis on
the complete Hha I digest ("able l), on fragment 6, and on hg-
ments 3 plus 6 (data not shown). These results indicate that SSB-
coated #X DNA contains a unique origin for complementary
strand synthesis within a 300-nucleotide fragment at or near the
protein n' recognition site.

Mobility of priming system in a direction opposite
chain synthesis
Models for Polarity of Primosome Migration. Multiple
primers are synthesized on almost every region of the chromo-
some of SSBcoated 4X DNA by prepriming proteins (n. n', n",
i, dmC, and dnaB) and primase when uncoupled hom DNA
synthesis (5, 7). Because SSB-coated 4X DNA has a single ori-
gin for a complementary strand start at or near the protein n'
recognition site, the primosome, known to contain at least pro-
teins dmB, n', and primase (refs. 5 and 28; unpublished re-
sults), must migrate progressively along the viral DNA strand to
achieve primer synthesis at multiple sites. Two possible models
for the polarity of primosome migration are considered (Fig. 5)
One model assumes that the primosome migrates in a direction
opposite to DNA and primer synthesis (antielongation), namely
the 5'43' polarity of the ssDNA template; the other assumes
migration in the same direction as priming and DNA elonga-
tion. Our results fit best with the antielongation direction
model.
Sizes of DNAs Synthesized on Hae III Fragment Z1 and Hha
I Fragment 6 as an Indicator of Polarity of Primosome Move-
ment. The 5' end of the protein n' recognition locus at position
2301 on the Sanger map (14,s) is separated from the 5' end of
the Hue 111 fragment Z1 by 526 residues and from the Hha I
hgment 6 by 248 residues. Were primosome movement in the
elongation direction (3'+5' polarity of the template), DNA
products shorter than 600 nucleotides on fragment Z1 and
shorter than 250 on fragment 6 would be expected. Instead, the
heterogeneous products synthesized on fragment Z1 were
longer than 600 nucleotides, the most abundant length being
1200-1300 (data not shown), suggesting de nmo chain initia-
tions predominate on the 3' side of the protein n' recognition
site. In fact, Hha I digestion of [=P]DNA products synthesized
on the Hue 111 fiagment Z1 yielded three main hgments that
comigrated with Hha I kagments 3, 6, and Sa on agarose gel
electrophoresis (data not shown), the 614-nucleotide Hha I hg-
ment 3 being located at the 3' side of the protein n' recognition
site (Fig. 4). A similar result was obtained with the DNA syn-
thesized on Hha I fragment 6. The products were homogeneous
in size and essentially the full length of the template (data not
shown).
Map Positions of DNA Synthesized on Hae III Fragment Z1
and Hh I Fragments 3 plus 6 as an Indicator of Polarity of Pri-
mmme Movement. Molecular hybridization can be used to
map DNA synthesized on DNA fragments. Products of #X RF
I digestion by endonuclease Hue 111 or Hha I were separated on

Elongation citwctmn

n' rscagnirion

minus strand
svnthesrs

Fro. 5.  Poesible models for the polarity of primosome migration on +X DNA.

72 Biochemistry: Arai and Kornberg

Proc. Nutl. Acud. Sci. USA 78 (1981)

Hhu I

Origin 1 2 3 4 5.6 7 89

If 11 \I 11

-4

Electrophoresis

FIG. 6.  Hybridization to Hha I fragments of DNA synthesized on
the Hae III fragment Z1 of +X ssDNA. +X RF I DNA (2.5 pg) was in-
cubated 2 hr at 37°C with Hho I(15 units) in 50 mM TriwHC1 (pH 7.W
5 mM MgCl,/l mM dithiothreitol. DNA fragments (arrows) were sep
arated by electrophoresis in 1.5% agarose gel, stained with ethidium
bromide, and transferred to DBM paper (21). '?-Labeled DNA synthe-
sized on Hae III fragment Z1 was extracted from the agarose gel by
electrophoresis and hybridized in 10% dextran sulfate 500 (Pharmacia)
(22) at 42°C for 24 hr tothe Hha I DNA fragments bound to DBM paper.
The paper was washed and autoradiographed at -80°C, using an in-
tensifying screen (Du Pont Cronex). The densitometric tracing of the
'9 autoradiogram is shown.

1.5% agarose gels, denatured, and transferred to DBM paper
(22). With the Hh I digest (10 bands), [32P]DNA synthesized
on Hue I11 fragment Z1 hybridized mainly to Hhu I fragments 3,
6, 8a, and 9a (Fig. 6). With the Hue 111 digest (nine bands),
[32P]DNA synthesized on Hh I fragments 6 (Fig. 7A) and 3 plus
6 (Fig. 7B) hybridized only to fragment Zl as expected. The
DNA synthesized on the latter as template hybridizes not
only to Hhu I fragment 6 but also to fragment 3. Inasmuch as the
Hha I fragment 3 region is inert for initiating replication (Fig.
l), primosome movement on fragment Z1 and Hha I frag-
ments 3 plus 6 in the 5'43' direction of the template may be
inferred (Fig. 4).
Synchronized Initiation on Circular DNA as an Indicator of
Polarity of Primosome Movement. Formation of the preprim-
ing replication intermediate is rate limiting in the 4X ssDNA
+ RF reaction (4). DNA replication is accelerated by preincu-
bation of SSB-mated 4X DNA with prepriming proteins in ATP

- .I

-- $

FIG. 7. DNA synthesized on rpX saDNA Hha I fragments 6 and 3
plus 6 hybridization to Hae III or Hha I fiagmenb. +X RF I DNA (3 pg)
was incubated 2 hr at 37°C with Hue III (15 units) or Hha I(15 units).
Afterelectrophoresis in 1.5% agaroee gel, DNA fragments were visu-
alized by staining with ethidium bromide (A) and transferred to DBM
paper. The "P-labeled DNAs synthesized on a Hha I digest of +X
ssDNA located in fragment 6 (B) or fragments 3 plus 6 (0 were ex-
tracted separately from the aganme gel by electrophoresis and hybrid-
ized at 42°C for 40 hr to the DNA Hae III and Hha I fragments bound to
DBM paper. Other procedures were as in Fig. 6.

Table 2. Distribution of s1P4abeled ribonucleotide primera on

restriction fragment9 of &X RF Il

F'reiincubated with

Ratio

Haem si

ha- hi- ATP ~NH~DDA dNHIuuA+

-

-~ ~__ -- ~__

ment due Ratio* A B C BIA CIA

Z1 1353  1.00   100*   100*     loo*    1.00 1.00
22 1078  0.80   96    68      56     0.71 0.58
23 872  0.64    74    49      26    0.66 0.35
Z4 603  0.45   60    35      19    0.58 0.32

The reaction was in two stages. In the first stage, components were
added in order at O'C and incubated 20 min at 30°C: 10 pl of buffer A,
0.4 mol of MgClZ, 60 nmol of ATP or 50 nmol of dNHIppA, 1.5 nmol
(as nucleotide) of +X DNA, 4.5 pg of SSB, 0.6 pg of protein n', 0.6 pg of
protein mixture n + n", 0.7 pg of protein i, 0.6 pg of dmC protein, 1.6
pg of dnaB protein, and water to 50 pl. Components were added to a
second-stage mixture in order at 0°C: 10 pl of buffer A, 0.4 mol of
MgCl,, 60 nmol of [U-~~PIATP (2000 dpm/pmol), 50 nmol each of
[u-~*P]G'I", UTP, and CTP (each at 10,OOO dpm/pmol), 6 nmol each of
['HIdATP, dGW, d"P, and dCTP (each at 200 dpm/pmol), 1.5 pg of
p-, 10 pg of DNA polymeraee III holoenzyme, and water to 50 pl.
After incubation of the second-stage mixture at 30°C for 4 min, the
fi&stage mixture was added and the incubation was continued for 30
min. After addition of 10 pl each of 0.4 M EDTA and 10% sodium do-
deql adfate and +X RF I (5 pg as carrier), the RF products, filtered
through Bio-Gel A-5m (Bio-Rad), were incubated2 hr at 30°C with Hue
IIIendonueleaee(100units)in250~lof50mMTriwHCl(pH7.5)/5mM
MgCldO.5 mM dithiothreitol. The "P-labeled digestion prcducta were
electrophoresed in 1.5% agarose gel and autoradiographed.
* Size relative to Z1.
+ After 3 min, the second-stage reaction was supplemented with 1 mM

each of the four unlabeled rNTPs.

* The "P radioactivity in each band was quantitated by densitometric
tracing of autoradiograms and values relative to Z1 (set at 100) were
calculated.

(4) or 5'-adenylyl imidodiphosphate (p[ NHIppA) (data not
shown). As will be described elsewhere, migration of the pri-
mosome on SSB-mated DNA requires energy furnished by ATP
hydrolysis. Initiation of priming of +X DNA replication was par-
tially synchronized by forming a prepriming replication inter-
mediate by preincubation with nonhydrolyzable p[NH]ppA. It
is assumed that the primosome in the presence of ATP moves
rapidly from its assembly point near the protein n' recognition
site to multiple locations around the circle before primer syn-
thesis begins; substitution of p[NH]ppA should minimize this
movement. After formation of the prepriming intermediate,
components for priming and DNA elongation could be added,
and the direction of primosome migration could be determined
by the [3PP]RNA primer distribution around the circular map.
Upon preincubation with ATP, [32P]RNA primers were located
relatively randomly (Table 2). However, upon preincubation
with p[NH]ppA, primers were located preferentially on the
Hue 111 fragment Z1 and the ratio of =P radioactivity (B/A and
CIA, Table 2) on each fragment decreased in the order: 22 > 23
> 24. These results indicate that the primosome migrates in the
direction Z1+ 22 + 23 + ZA, consistent with the prediction
of the antielongation direction model.

DISCUSSION

The prepriming system reconstituted in oitro from six E. coli
proteins is highly.specific for 4X DNA coated with SSB (13).
This specificity apparently derives from the capacity of protein
n' to recognize and bind a specific intergenic sequence. This
initiating event in complementary strand synthesis is obscured
by rapid movement of the priming system (primosome) around
the circular template, leading to the apparently random prim-
ing and DNA syntheses observed in uioo and in oitro. In the

Biochemistry: Arai and Kornberg

Proc. Natl. Acad. Sci. USA 78 (1981)

73

present studies, primosome movement was restrained by using
DNA hgments generated by endonucleases Hue I11 and Hha I
(17, 18). Only hgments containing the protein n' recognition
site were effective in sustaining the prepriming, priming, and
DNA synthetic actions. Elevated levels of SSB were needed to
suppress nonspecific priming and replication.
In view of the complexity of prepriming, it is remarkable that
the 300-nucleotide Hha I fragment 6 suffices for the specificity
for chain initiation on 4X DNA. The DNA product size, 250 res-
idues, suggests that primers are made at or near the haupin
within the 55-nucleotide protein n' recognition locus where,
presumably, primosome assembly takes place. The effect of
protein n' binding and primosome assembly at this site needs to
be clarified. Because p[NH]ppA sufTices far the formation of a
prepriming replication intermediate (unpublished result), hy-
drolysis of ATP and movement ofprotein n' or dnaB protein are
not prerequisites for initial primosome assembly. The remark-
able feature of the unique origin for 4X complementary strand
synthesis is that this site is not necessarily defined as the site
where a unique primer is made. Rather it may be defined as the
site for assembly of the primosome, which then may migrate
processively along the viral circle. As will be described else-
where, the energy of ATP and dATP hydrolysis utilized by pro-
tein n' is essential to promote priming at many regions. Presum-
ably protein n' not only recognizes the origin of complementary
strand replication but also drives the primosome along the DNA
strand.
In the present study, polarity of primosome migration was
shown to be uniquely in the ontielongation, the 5'+3',direc-
tion of the template (Fig. 5). This is consistent with the pro-
ped mobility of the primosome on the lagging strand at the
replication fork (3,s). movement of the replication fork and pri-
mosome taking place in the same direction. Thus Far, the polar-
ity of primosome migration has been determined only when
coupled to priming and DNA elongation. Approximately one
primer is made per circle despite a capacity to synthesize mul-
tiple primers when uncoupled from DNA elongation (ref. 5; un-
published results). This indicates that the primosome can mi-
grate in the 5'+3' direction before priming. The random
distribution of primers on 4X DNA when preincubated with
ATP to form the prepriming replication intermediate suggests
that the.primosome moves along the DNA template even with-
out primase action. Whether this migration of a primosome
lacking primase is unidirectional with a 5'43' polarity remains
to be clarified.

This work W~LE supported in part by grants from the National Institutes
of Health and the National Science Foundation. K.A. is a Fellow of the
American Cancer Society.

1.


2.

3.

4.


5.

6.

7.

a.

Q.

10.


11.

12.

13.

14.

15.

16.

17.

18.

19.


20.

21.

22.


23.

24.

25.

26.

27.

28.

Schekman, R., Weiner, A. & Kornberg, A. (1974) Science 186.
987-993.
Wickner, S. & Hunvik, J. (1974) Proc. Natl. Acad. Sci. USA 71,
4120-4124.
Kornberg, A. (1980) UNA Replicotion(Freeman, San Francisco).
Weiner, J. H., McMaeken, R. & Kornberg, A. (1976) Proc. Natl
Ad. Sd. USA 73,752-756.
McMacken, R., Ueda, K. & Kornberg, A. (1977) Proc. Nutl. Acad.
Sui. USA 74,4190-4194.
McMacken, R., Bouchk, J.-P., Rowen, S. L., Weiner, J. H.,
Ueda, K., Thelander, L.. McHenry, C. & Kornberg, A. (1977) in
Nucleic Acid-Protein Recognition. ed. Vogel, H. J. (Academic,
New York), pp. 15-29.
McMacken, R. & Kornberg, A. (1978) 1. Biol. Chem. 253,
3313-3319.
Tabak, H. F., Griffith, J., Geider, K., Schaller, H. & Kornberg,
A. (1974)J Biol. Chem. '249,3049-3054.
Geider, K.;Beck, E. & Schaller, H. (1978) Proc. Natl. dcad. Sci.
USA 75,645-649.
Bouchg, J.-P., Rowen, L. & Kornberg, A. (1978) J. Bioi. Chem.
253.765-769.

Fiddes. J. C., Barrell, B. G. & Godson, G N. (1978) Proc. Natl.
Ad. Sci. USA 75,1081-1085.
Sims, J. & Dressler, D. (1978) Proc. Natl. Acad. Sci. USA 75,
3084-3098.
Arai, K. & Kornberg, A. (1979) Pm. Natl. Ad. Sci. USA 76,
4308-4312.
Shlomai, J. 81 Kornberg, A. (1980) Proc. Natl. Acad. Sci. USA 77,
799-803.
Eisenberg, S., Harben, B., Hours, C. & Denhardt, D. T. (1975)
J. Mol. Bwl. 99,107-123.
Arai, K.. Arai, N., Shlomai, J. .% Kornberg, A. (1980) Proc. Natl.
Ad. Sci. USA 77,3322-3326.
Blakesley, R. W. & Wells, R. D. (1975) Nature (London) 257,
421-422.
Horiuchi, K. & Zinder, N. D. (1975) Proc. Natl. Acud. Sci. USA 72,
w5-w8.
McDonell, M. W., Simon, M. N. 81 Studier, I. W. (19771. Mol.
*Bioi. 110,119-146.
Maniatis, T., Jefhey, A. & van de Sande, H. (1975) Biochnnisty

Ahvine, J. C., Kemp, D. J. & Stark, G. R. (1977) Proc. Nut!. Acad.

Wahl, C. M., Stem, M. & Stark, G. R. (1979) Proc. Natl. Acad. Sci.
USA 76,3683-3687.
Codson, G. N., Barrell, B. G., Staden. R. & Fiddes, J. C. (1978)
Nature (bndon) 276,236-247.
Sims, J. & Benz, E. M., Jr., (1980) Proc. Natl. Acad. Sn'. USA 77,
9004KM.
Sanger, F., Coulson, A. R., Friedmann, T., Air, G. M., Barrell,
B. C., Brown, N. L., Fiddes, J. C.. Hutchison, C. A., Slocombe,
P. M. %Smith, M. (1978)J Mol. Biol. 135,225-246.
Weiner, J. H., Bertsch, L. L. 81 Kornberg, A. (1975)J Bwl. Chem.

Langeveld, S. A., van Mansfeld, A. D. M., Baas, P. D., Jansz, H.
S., van Arkel, G. A. 6Weisbeek. P. J. (1978)Nature(London)271,
417-420.
Ueda, K., McMacken, R. & Kornberg, A. (1978)J. Bid. Chem.

14,3787-3794.

Sci. USA 74,5350-5354.

SO, 1972-1980.

253,261-270.