Volume 12 Number 7 1984
Nucleic Acids Research
Initiation signals for complementary strand DNA synthesis in the region of the replication origin of
the Escherichia coli chromosome
Antoine R.Stuitje, Peter J.Weisbeek* and Michiel Meijer
Department of Electron Microscopy and Molecular Cytology, University of Amsterdam, Plantage
Muidergracht 14, 1018 TV Amsterdam, and •Department of Molecular Cell Biology, State
University of Utrecht, Padualaan 8, 3584 CH Utrecht, The Netherlands
Received 10 January 1984; Revised and Accepted 15 March 1984
ABSTRACT
We have used an in vivo plasmid-cjiX174 packaging system to detect replication initiation signals in the region of the replication origin (orlC) of the
Escherichia coli chromosome. The results obtained are summarized as follows:
(i) Neither within nor close to oriC effective signals for initiating complementary strand synthesis could be detected. We conclude that initiation mechanisms for leading and lagging strand synthesis at oriC are not identical to
any known priming mechanism of DNA synthesis, (ii) At least five signals that
can function as complementary strand origins on ss~plasmid DNA are located in
a region about 2000-3300 base pairs away from orlC in the clockwise direction
on the chromosome. We suggest that these signals are protein n' like recognition sequences since they are dependent for their activity on dnaB protein and
show sequence similarities to other putative n' recognition sequences. Surprisingly, some of the signals are located on the template for leading strand
synthesis.
INTRODUCTION
Replication of the Escherichia coli chromosome initiates at a unique site,
the replication origin (oriC), and proceeds bidirectionally (1). oriC has
been obtained either as part of a phage or plasmid hybrid (2-4), or as a
minichromosome which contains oriC as the single replication origin (5-7). The
nucleotide sequence of the oriC region has been determined (8,9) and the
sequences required for autonomous replication were delimited to a specific
region of 245 base pairs (4). Initiation of replication on minichromosomes
seems to resemble initiation of chromosomal replication iji vivo and ill vitro
(7, 10-12).
For initiation of bidirectional replication at oriC distinct priming events
are required to ensure leading and lagging strand synthesis at both replication forks. Replication studies with plasmids and single stranded (ss) DNA
phages have revealed the existence of at least three different priming mechanisms (13, 14). (I) Synthesis of a specific primer by RNA polymerase on the ss
DNA template of filamentous phages and in plasmid Col El, (ii)
specific
priming on the ss DNA template of phage G4 by DNA primase encoded by the dnaG
© IRL Press Limited, Oxford, England.
3321
Nucleic Acids Research
gene, (ill) generation of short primers at multiple sites on the 0X174 ss
DNA template by a host encoded multi-enzyme complex, termed the primosome. The
involvement of two of these mechanisms in priming of leading strand synthesis
at the IS. coli replication origin have been suggested. First, it has been speculated that DNA primase is involved because of sequence similarities between
the G4 (-) origin and sequences in both strands within oriC (8). Second, the
sensitivity of initiation of replication to rifampicin (15,16) and the presence of two RNA polymerase promoters within oriC, in a back to back arrangement (17), may indicate a role of RNA polymerase in leading strand priming
events.
Apart from priming leading strand synthesis a system must be present that
can account for priming lagging strand synthesis. A multi-enzyme complex, the
primosome, is a very likely candidate for these multiple priming events. It
consists of at least seven different proteins (n, n', n 1 ', i, dnaB, dnaC, and
primase). The complex is assembled after the interaction of the prepriming
protein n', a ss DNA specific ATPase (19), with a specific sequence on the ss
0X174 DNA template (18). Following its assembly, the primosome moves in the 5'
— » - 3 ' direction of the template thereby enabling primase to synthesize short
primers at non-specific sequences in the opposite direction (18). Based upon
these observations Kornberg (20) has proposed a model for the initiation of
bidirectional replication at the E_. coli origin. According to this model the
initiation of leading and lagging strand synthesis at both replication forks,
is directed by the assembly of a primosome on either DNA strand at oriC (20).
In this study we have tried to obtain some insight into the mechanism that
is used by 15. coli to initiate DNA synthesis at oriC. Recently, we described a
system suitable for the detection of signals for the complementary strand
DNA synthesis In vivo (21). In this system, plasmids containing the 0X174
origin for (+) strand synthesis are packaged as single stranded circles Into
0X phage coats, upon infection of plasmid containing cells by 0X174 phage.
These plasmid containing particles are capable of transducing <pX sensitive
cells to antibiotic resistance and this results in cells with intact ds
plasmid DNA. It was found that the transduction efficiency of the plasmid containing particles strongly depended on the presence of complementary strand
initiation signals in the packaged ss DNA strand. The complementary strand
origins of phages G4, Ml3 and 0X174 all were fully active in this transduction
system (21) .
We have used this system to identify and localize complementary strand initiation signals In a region of 4012 base pairs surrounding the E. coll repli-
3322
Nucleic Acids Research
cation origin. Several dnaB dependent replication signals were found on both
strands in a region 2000-3300 base pairs away from oriC in the clockwise
direction on the chromosome. No such signals were detected within or close to
oriC. The implications of these results will be discussed.
MATERIALS AND METHODS
Bacterial strains, phages and plasmids
The following E_. coli strains were used: E_. coli C recA171 $X S , £. coli
HF4712 sup A U G <JiXs and E_. coli LD311 uvrA end I thyA dnaB ts ((.Xs (22). The phage
<)iXam3 is a lysis defective amber mutant in gene E of $X174 (23). Plasmid
pPRlll or its deletion derivative pPR111.6 were used for the detection of initiation signals (21). Minichromosome pCM959 served as source of DNA fragments
of the £. coli oriC region (7,8).
Enzymes
Restriction endonucleases, calf intestinal phosphatase and T4 DNA ligase
were from New England Biolabs, Boehringer Mannheim or Amersham. Nuclease SI
was from Sigma. The enzymes were used according to the manufacturer's recommendations.
DNA techniques
Preparative isolation of plasmid DNA was by the alkaline lysis method as
described by Birnboim and Doly (24). Restriction fragments were isolated from
gel slices by electro elution using the Isco sample concentrator model 1750.
Blunt ended restriction fragments were cloned into the Hindll site of pPR111.6
using T4 DNA ligase. Prior to transformation the DNA was cleaved with HindII
to select for recombinant plasmids. Cohesive end ligation was carried out with
calf intestinal phophatase-treated vector DNA to select for recombinant
plasmids. E_. coli C recA171 was transformed by the freeze-shock procedure of
Maniatis ej^ al^ (25). Transformant colonies were screened for the presence of
recombinant plasmids using the boiling procedure of Quigley and Holmes (26).
The nature and orientation of the inserts were determined by restriction
enzyme analysis. Several deletion mutants were obtained by SI nuclease treatment to remove incompatible 5' or 3' ends following digestion with different
restriction enzymes.
Packaging and transduction assay
A detailed description of the packaging and transduction assay has been
published previously (21). Essentially, E_. coli C cells harbouring plasmids
with the $X (+) origin were infected with $X am 3 phages at a m.o.i. of 3.
The infected culture was incubated for 3 hours and the phage particles were
3323
Nucleic Acids Research
isolated by lysozyme and chloroform treatment of a concentrated cell suspension. The number of plaque forming units (p.f.u.) in the total phage progeny
was determined by infecting the $X 8 amber suppressor strain HF4712. The number
of transducing particles was determined after infection of the suppressor-less
strain E^. coli C by determining the number of kanamycin-resistant transductants. The transduction efficiency is expressed as the ratio of the number of
transductant colonies and the number of plaque forming units. The assay for
the requirement of dnaB protein has been described previously (21). The dependency on dnaB protein is expressed as the ratio of transductants obtained at
30°C by infecting E_. coli LD 311 dnaB ts at 42°C and 30°C.
RESULTS
Cloning of restriction fragments into vectors suitable to detect initiation
signals
Previously, we have described an in vivo assay to detect sequences that can
function as complementary strand initiation signals (21). In general, recombinant plasmids that contain the $X174 (+) origin are a substrate in vivo for
gene A protein, produced by infecting $X174 phage. This results in progeny
particle yields that consist of 95% normal <f>X phage particles and about 5%
plasmid particles (27). The plasmid particles consist of circular singlestranded plasmid DNA encapsulated in normal phage coats. With the help of a
kanamycin resistance marker on the plasmid genome it was demonstrated that
such plasmid particles can transduce $X sensitive cells (27). The transduction
efficiency of the plasmid particles was found to be independent of the length
of the plasmid DNA within a size range of about 4400-5600 nucleotides (28).
Moreover, within this size range of plasmid molecules the transduction efficiency strongly increases if complementary strand initiation signals are present on the packaged ss DNA strand (21). It was demonstrated that the n'
protein recognition sites of $X174 and pBR322 and the complementary strand
origin of phages G4 and M13 are fully active in this system. (21).
Plasmid pPRlll carries the $X (+) origin and consists of 4410 base pairs.
Since it lacks a proper complementary strand initiation signal on the packaged
DNA strand pPRlll-phage particles have a low transduction efficiency (about
5 x 10~^; 21). Therefore, this plasmid can be used to detect those sequences of
the oriC region that can function as complementary strand initiation origins.
A library of recombinant plasmids was constructed that contains restriction
fragments from the orlC region, cloned in both orientations into pPRlll or its
deletion derivative pPRlll.6 (21; Fig. 1). Minichromosome pCM959 served as a
3324
Nucleic Acids Research
» X l.lOM
H>nd II
Iz9
Pstl
1
Amp"
Kan"
'4
IR
IR
Fig. 1. Schematic structure of the $X (+) origin containing recombinant
plasmids pPRlll and pPR111.6. Large open and filled boxes represent <fiX Haelll
fragments. Arrows indicate direction of transcription of the $X genome and the
genes conferring resistance to ampicillln and kanamycin. Hatched box represents the DNA fragment that is absent in pPR111.6. IR, inverted repeat.
source of restriction fragments. This minichromosome consists exclusively of
chromosomal DNA and its sequence has been determined (7,8, 29). The plasmids
that were constructed for this study are listed in Table I. Since only the DNA
strand that is linked to the strand of the vector plasmid, which contains the
$X (+) origin is assayed in the transduction system, we have indicated the
orientation of the cloned fragments in Table I. (L) denotes that the strand
with 3'-»-5' polarity in the clockwise direction on the chromosome is linked to
the $X (+) origin and (H) indicates that this is the case with the strand with
5'*-3' polarity in the clockwise direction on the chromosome. Since the transduction assay is only indicative for the presence or absence of complementary
strand initiation signals in the case of plasmids within a size range of
4400-5600 base pairs the size of the constructed recombinant plasmids is also
indicated in Table I.
Localization of sequences that function as complementary strand initiation
signals
In a previous report it was shown that deletion derivatives of plasmid
pPRlll have drastically reduced transduction efficiencies (21). In one of
these plasmids, pPRlll.6, a deletion of 450 base pairs was generated at the
PstI site (Fig. 1 ) . The transduction frequency of pPRlll.6 particles is about
10 3
times lower compared to pPRlll plasmid particles. It was shown that the
efficiency
of transduction is restored to the level of pPRlll by reintro-
ducing DNA fragments of 450 to 1600 base pairs. This property allows the isolation of recombinant plasmids, since plasmids with an increased transduction
efficiency are enriched by a number of cycles of packaging and reinfection.
Moreover, this procedure strongly selects for recombinant plasmids carrying
complementary strand origin signals on the packaged DNA strand. Following this
3325
Nucleic Acids Research
Table I . List of orlC-vector hybrid plasmids"
Plasmid
Construction* 1
pEM20
pEM63
Haelll fragment (-43 to +973) pCM959 : : Hindi!
pPRlll.6
as pEM20
Bglll fragment (-471 to +22) pCM959 : : BamHl
pPR111.6
as pEM61
BamHl-BgUI fragment (+92 to +2006) pCM959 : :
BamHl pPR111.6
as pEH95
A Ec£RV-BamHl fragment pEM95
a EcoRV-BamHl fragment pEM96
BglH fragment (+2006 to +3335 and -677 to -471)
pCM959 : : BamHl pPRlll
as pEM51
pEH107
a PvuII fragment pEM51
pEM97
A PvuII fragment pEM63
pE«28
PvuII fragment (+2389
: : H i n d u pPR111.6
as pEM28
BamHl fragment (+2191
-105) pCM959 : : BamHl
a Pstl-PvuII fragment
pEM21
pEM61
pEM62
pEM95
pEH96
pEM92
pEM93
pEM51
pEM7 5
pEM80
pE«91
pEM81
E. c o l i chromosomal
to +3224) pCM959
to +3335 and -677 to
pPR111.6
pEM80
pEM53
P s t I fragment (-677 to +488 and +2566 to
+3335) pCH959 : : PstI pPRlll
a Bgll fragment pEM81
Pstl-Hinfl fragment (+2566 to +2987)
pCM959 : : Hindu pPRlll.6
ABamHl-PsU fragment pEMSl
pEK74
A Bgll-PvuII fragment pEH53
pEM6O
ABamHl-Bgn fragment pEMSl
pEMlOO
pEH26
pEH2O9
pEM223
a
L
5.0
H
L
5.0
4.5
-471 to +22
+92 to +2006
H
L
4.5
5.9
+92 to +2006
+727 to +2006
+727 to +2006
+2006 to +3335 and -677
to -471
+2006 to +3335 and -677 to
-471
+2006 to +2389, +3224 to
+3335 and -677 to -471
+2006 to +2389, +3224 to
+3335 and -677 to -471
+2389 to +3224
H
L
H
L
5.9
5.2
5.2
5.9
H
5.9
L
5.1
H
5.L
I.
4.8
+2389 to +3224
+2191 to +3335 and -677 to
-105
+2191 to +2566, +3226 to
+3335 and -677 to -105
-677 to +488 and +2566 to
+3335
+2566 to +2752
+2566 to +2987
H
H
4.8
5.7
H
5.0
L
6.3
L
H
4.8
4.4
+2006 to +2191, +2566 to
+3335 and -677 to -471
+2006 to +2191, +2566 to
+2752, +3226 t o +3335 and
-677 to-471
+2006 to +2191, +2922 to
+3335 and -677 to-471
L
5.6
L
5.1
L
4.8
L
4.8
f. „
_L1G££
+180 tO +/5bO
6.5
+488 to +727 and +2389 to
+2566
The term orlC-vector hybrid plasmids Indicates the cloning of fragments of pCM959 Into pPRlll
or pPRlll.6. It does not refer to the actual presence of a functional orIC sequence.
Restriction sites and nucleotide numbering are shown in Fig. 2A. SymbolsAand
deletion and insertion, respectively.
c
Size
(Kb)
-43 to +973
-471 to +22
.lag
: : PstI pPRlll
a EcoRV-PvuII fragment pEM209
Orientation 0
:: mean a
L and H denote the DNA strand that is assayed in the transduction system. For a definition
see text and Fig. 2C and 2E.
procedure plasmid pPRlll.63 was obtained by cloning a partial Haelll digest of
pCM959 Into pPRlll.6 (Fig. 2; 21). The chromosomal sequences present on
pPRlll.63 are located outside orIC in a region between position 2319 and 2921
on the pCM959 map (Fig. 2B). In the study described in this paper we did not
use this selection procedure to isolate recombinant plasmids, and, therefore,
plasmids with relatively weak complementary strand initiation signals are not
selected against. In Fig. 2 we have compiled the results of the transduction
analysis performed with relevant recombinant plasmids. Fig. 2A shows a physical map of the 12. coli chromosomal DNA that is present in pCM959. Note that at
the point of circularization of chromosomal DNA in pCM959 (indicated by_f ) ,
3326
Nucleic Acids Research
$ S
A
I >
chromosome
! iiiii
(70kD)
Transduction °
efficiency x 10
pEM
20
008
pEM
61
0 01
!002
pEM
92
001.
• 003
pEM
74
PEM223
pEM 107
pEH
» +
+ 1
•*(
• 002
t03
08
= 03
09
'05
pPRt1U63
08
= 03
28
1 1
= 03
pEM 100
006
-•003
I Ii
s
i
L
pEM
21
007
pEM
62
006
= 003
• 002
pEM
93
006
'003
pEM
91
007
•002
pEM
26
005
• 002
pEM
97
08
!02
pEM
75
1 3
!02
i
E
006
07
60
pEM
C
:003
H
I
IV
Fig. 2. Signals for Initiating complementary strand DNA synthesis in the
oriC-region. A. Linear map of pCM959 opened at the point of circularization
(29) of the E_. coll chromosome. Arrows indicate direction of transcription of
the respective genes. B, D. Fragments located on the L and H-strand, respectively, assayed for the presence of complementary strand initiation signals in
the $X174 directed transduction system of plasmid DNA, (
- ) , no signals
present; (Mil) signals present. The transduction efficiencies are mean values
of at least two experiments with their standard deviations. The value for
plasmid pPRlll is about 5 x 10~*. C, E. Location of complementary strand initiation tiation signals on the L and H-strand, respectively.
sequences that are located to the left and right of oriC on the E^. coll chromosome are joint. Therefore, cloning of restriction fragments that overlap
this circularization point result in plasmids that contain both sequences to
the right and left of oriC. In fig. 2B and 2D the data of the transduction
experiments are arranged according to the DNA strand that is assayed. Fig. 2B
and 2D show the results obtained with fragments of the L and H-strand of the
chromosome, respectively. The chromosomal fragments present in the different
plasmids are indicated by Mil or
, which refers to the presence or
3327
Nucleic Acids Research
absence of a complementary strand initiation signal on each fragment. The
actual data of the transduction efficiency are also indicated in Fig. 2. From
these results we conclude that sequences within and close to the minimal
replication origin cannot function as complementary strand origins in this
transduction system (Fig. 2; pEM20 and pEM 21). The possibility existed that
the failure to detect such replication signals in this region of the chromosome is caused by the presence of a functional E^. coli replication origin on
pEM20 and pEM21. However, similar plasmids in which the minimal oriC sequence
was destroyed by a 16 base pairs deletion at its left boundary did not show
any increase in transduction efficiency (not shown). We found several such
signals on both strands in a region about 2000 to 3300 base pairs to the right
of oriC. Fine mapping of these sequences was hampered because several initiation signals turned out to be present in this region of the chromosome. In
fig. 2 we have presented the results obtained with the relevant plasmids that
permit the mapping of the initiation signals. For example, the initiation
signal located between BamHl and PvuII sites on the L-strand (designated I;
Fig. 2C) is identified by comparing pEM74 and pEM107. pEM107 contains an initiation signal and such a signal is absent on pEM74. Therefore, the signal of
pEM107 must reside in the sequences that are not present on pEM74. Consequently, these sequences map between BamHl and PvuII sites as Indicated in
Fig. 2C. The other signals (II and III; Fig. 2C) present on the L-strand were
identified in a similar way (compare pEM223 with pEM20 and pEM60 with pEM74).
The initiation signals on the H-strand are identified by comparing pEM91 with
pEM97 (IV; Fig. 2E) and pEM75 with pEM91 and pEM26 (V; Fig. 2E).
The nature of the initiation signals present on the fragments I to V is
indicated by the dependency of the transduction on functional dnaB protein.
This assay was performed essentially as described previously (21). Table II
Table II. Effect of dnaB protein on transduction frequency of plasmid DNA
a
b
Plasmida
Ratio of transductants obtained after incubating
LD311 dnaB ts at 42°C and 30°C.
pEM107
pEM28
pEM97
pEM75
pDG61b
0.006
0.005
0.008
0.003
1.3
The structure of the plasmids is shown in Fig. 2
The structure of pDG61 is described elsewhere (21).
3328
Nucleic Acids Research
shows that the transduction frequencies of pEM107, pEM28, pEM97 and pEM75
plasmid particles strongly depend on
the presence of a functional dnaB pro-
tein. On the contrary, transductions of pDG61 particles does not depend on
dnaB protein. pDG61 contains the phage G4 complementary strand origin, which
is dnaG dependent only (21, 30). These results suggest that the initiation
signals identified in this region of the chromosome are primosome dependent,
which implies that they possibly contain protein n1 like recognition sequen-
DISCUSSION
The mechanism of initiating leading and lagging strand synthesis at the
£.
coli replication origin has been the subject of considerable interest and speculation in the past few years. Several studies suggested the involvement of
dnaG gene product (primase) and RNA polymerase in priming of leading strand
synthesis at oriC (8, 17). Furthermore, extensive in vitro studies on the
mechanism of primosome directed priming on ss DNA of phage $X17A, has provided
an attractive model for initiation of lagging strand synthesis at oriC (20).
Recently, we have described an experimental system which can be used to
detect the three known initiation signals for priming DNA synthesis on ss DNA
templates, i.e. priming by dnaG protein (primase) on phage G4 templates,
priming by RNA polymerase on phage M13 templates and primosome-directed
priming on 0X174 templates (21). By using this system it was shown that protein n' like recognition sequences for primosome assembly are present near the
replication origins of plasmids pBR322, CloDF13, pI5A and F (21). These
results indicated that primosome priming is the general mechanism of priming
lagging strand synthesis. However, the data presented in this paper
demonstrate that no such n' type, dnaG protein or RNA polymerase initiation
signal that can function in this assay system, is found within or close to
oriC. Although we cannot exclude that particular structural constraints or
binding of specific proteins may suppress the usage of n' sites in our system,
at least our data indicate that priming of lagging strand synthesis at oriC
differs distinctly from the ones mentioned above. This seems to be supported
by the observation of Tabata e_t £l. (31) that the initiation of bidirectional
replication occurs at non-specific sequences outside the minimal oriC region.
Recently, one of us reported that the putative lagging strand origin of
pBR322 (n', dnaB and dnaC dependent) is not essential for the maintenance of
this plasmid (32). By deleting the n' protein recognition site, plasmids were
obtained with a 2 to 3—fold reduction in copy—number. This result suggests
3329
Nucleic Acids Research
that the presence of n' recognition sites near the leading strand replication
origin of relaxed replicating plasmids is not absolutely required, but only
ensures the fidelity of primosome assembly for initiation of lagging strand
DNA synthesis. Therefore, it is possible that on ds DNA templates a leading
strand Initiation signal is the only requisite for assembly of a replisome to
ensure both leading and lagging strand DNA synthesis at the replication fork.
The replisome is assumed to consist of at least the DNA polymerase III
holoenzyme, the primosome and helix destabilizing proteins (13,14). The second
replication fork that is needed for bidirectionally replicating DNA molecules
may arise from the first replication fork without any additional requirements:
the lagging strand synthesized at the first replication fork will become the
leading strand in the opposite direction and this will trigger replisome
assembly in a similar way to ensure leading and lagging strand synthesis at
the second replication fork. According to this model only one single specific
priming event may be required to initiate bidirectional replication at oriC,
since primosome directed priming apparently occurs at non-specific
sequences
(18). This model is supported by the data of Hirose £t al^. (33) which show
that specific RNA-DNA transition sites are only present on the L-strand within
oriC. We suggest that these RNA-DNA transition sites represent initiation
sites of leading strand synthesis. Our data suggest that the signal to initiate this primer RNA represents a new type of priming mechanism since it does
not function in the conversion of ss plasmid DNA to ds plasmid molecules in
»X174
(2306-2352)
5'ATAACCCC IAAGCGG RAAAAATTTTAATTTTTGCCGCTGAGGGGTTG
pBR322
(L)
(2371-2325)
5 TGAGCGAGBAAGCGG KAGAGCGCCTGATGCGGTATTTTCTCCTTACG
pBR322
(H)
(2136-2182)
5 CTTGTCTG FAAGCGG ITGCCGGGAGCAGACAAGCCCGTCAGGGCGCG
E. coll
(I)
(2288-2242)
5 ACCGCGCA XAGCGC ITGATGCCACTCCAGCTCCAGGCGATCTTCGT
E. coll
(II)
(2470-2424)
5'CGGGATGC IAACCCG XTCCTCAGACGCTGGCGGCGATAAATTATTA
E. coll
(III)
(3161-3115)
5 TGGTGCGC JAACCGG IGACGGTTCCTGAGCAGGTTGATGGTCTGCAA
E. coll
(IV)
(2035-2080)
5 TTGCCGGGPAAGCTGIGCGATGGTCATCGCCACGACGTGCGCGCACC
E. coll
(V)
(3130-3176)
5'CTGCTCAG ;AACCGT:GCCCGTTCGCGCACCATGGTGCGGAAGGTTT
CONSENSUS:
Fig. 3. Nucleotide sequence comparison of DNA fragments containing dnaB protein dependent i n i t i a t i o n signals. The nucleotide sequences enclosed within
the box represent regions with maximal homology as determined by computer anal y s i s . Arrows indicate regions with dyad symmetry.
3330
Nucleic Acids Research
the transduction assay.
Although no initiation signals were found within or close to oriC at least
five signals that can function as complementary strand origins on ss plasmid
DNA, were found in a region about 2000-3300 base pairs away from oriC in the
clockwise direction on the chromosome. The data presented demonstrate the
involvement of dnaB protein in the complementary strand synthesis directed by
these sequences. Since this dnaB dependency indicates the involvement of the
primosome, we have searched for sequence homology within the regions I to V
with that of putative n' protein recognition sequences. As previously
suggested (21) the sequence GAAGCGG, in conjunction with potential hairpin
structures, may play an important role in the recognition of n' protein. This
sequence is not present in the chromosomal DNA analysed in this study.
However, Fig. 3 shows that the regions I to V do contain sequences that
resemble this putative consensus sequence, surrounded by regions with potential stem-loop structures. Although the identification of these sequences may
facilitate to stipulate the minimal requirements of n' protein recognition
sites, their function in the replication of the E^. coli chromosome is yet
unclear. For example, their location, more than 2000 base pairs away from
oriC, makes it very unlikely that they are functional in the early steps of
replication of the E_. coli chromosome. In fact, in vivo (10, 33) and in vitro
(11, 12, 31) studies implicate that the initiation of both leading and lagging
strand synthesis occurs either within or very close to oriC. According to
these data and the current model of elongation of DNA replication it seems
plausible that the initiation signals found in this study are not essential
for chromosomal DNA replication. On the analogy of the lagging strand initiation signals near the leading strand origin of pBR322, it is tempting to
speculate that the n' type initiation signals found in this study may function
as back-up signals to ensure the fidelity of replisome function during elongation. Further studies on the enyzmology at the DNA replication fork are
required to access whether primosome assembly sites on both the leading and
lagging strand template of the 12. coli chromosome may function in the process
of elongation of DNA replication.
ACKNOWLEDGEMENTS
We wish to thank H. Veerman, M.A. de Jong and R. Teertstra for constructing
some of the plasmids and performing some of the tests, J.H.D. Leutscher for
making the drawings and Ch.E.A. van Wijngaarden for typing the manuscript.
3331
Nucleic Acids Research
REFERENCES
[ I ] M a s t e r s , M., Broda, P. (1971) Nature (Lond.) New B i o l . 232, 137-140
[2] Meyenburg, K. v o n , Hansen, F . G . , N i e l s e n , L . D . , R i i s e , E. (1978) Mol.
Gen. Genet. 160, 287-295
[3] Hiraga, S. (1976) Proc. Natl. Acad. Sci. USA 73, 198-202
[4] Oka, A., Sugimoto, K., Takanami, M., Hirota, Y. (1980) Mol. Gen. Genet.
178, 9-20
[5] Yasuda, S., Hirota, Y (1977) Proc. Natl. Acad. Sci. USA 74, 5458-5462
[6] Messer, W., Bergmans, H.E.N, Meijer, M., Womack, J.E., Hansen, F.G.,
Meyenburg, K. von (1978) Mol.. Gen. Genet. 162, 269-275
[7] Meyenburg, K. von, Hansen, F.G., Riise, E. Bergmans, H.E.N., Meijer, M.,
Messer, W. (1978) Cold Spring Harbor Symp. Quant. Biol. 43, 121-128
[8] Meijer, M., Beck, E., Hansen, F.G., Bergmans, H.E.N., Messer, W., Meyenburg, K. von, Schaller, H. (1979) Proc. Natl. Acad. Sci. USA 76, 580-584
[9] Sugimoto, K., Oka, A., Sugisaki, H. Takanami, M., Nishimura, A., Yasuda,
S., Hirota, Y. (1979) Proc. Natl. Acad. Sci. USA 76, 575-597
[10] Meijer, M., Messer, W. (1980) J. Bacteriol. 143, 1049-1053
[II] Fuller, R.S., Kaguni, J.M., Kornberg, A. (1981) Proc. Natl. Acad. Sci.
USA 78, 7370-7374
[12] Kaguni, J.M., Fuller, R.S., Kornberg, A. (1982) Nature 296, 623-627
[13] Kornberg, A. (1980) DNA replication, W.H. Freeman and Co., San Francisco
[14] Kornberg, A. (1982) Supplement to DNA replication, W.A. Freeman and Co.,
San Francisco
[15] Lark, K.G. (1972) J. Mol. Biol. 64, 47-60
[16] Messer, W. (1972) J. Bacteriol. 112, 7-12
[17] Lother, H., Messer, W. (1981) Nature 294, 376-378
[18] Arai, K., Kornberg, A. (1981) Proc. Natl. Acad. Sci USA 78, 69-73
[19] Schlomai, J., Kornberg, A. (1980) Proc. Natl. Acad. Sci. USA 77, 799-803
[20] Kornberg, A. (1978) Cold Spring Harbor Symp. Quant. Biol. 43, 1-9
[21] van der Ende, A., Teertstra, R. van der Avoort, H.G.A.M., Weisbeek, P.J.
(1983) Nucleic Acids Res. 11, 4957-4975
[22] Derstine, P.-L., Dumas, L.B., Miller, C.A. (1976) J. Virol. 19, 915-924
[23] Weisbeek, P.J., van Arkel, G.A. (1978) in: 'The Single-Stranded DNA
Phages' (Denhardt, D.T., Dressier, B., Ray, D.S. eds.), Cold Spring Harbor Lab, Cold Spring Harbor, N. Y., pp. 31-49
[24] Birnboim, H.C., Doly, J. (1979) Nucleic Acids Res. 7, 1513-1523
[25] Maniatls, T. Fritsch, E.F., Sambrook, J. (1982) Molecular cloning. Cold
Spring Harbor Lab, Cold Spring Harbor, N.Y., p. 254
[26] Holmes, D.S., Quigley, M. (1981) Analytical Biochemistry 114, 193-197
[27] Weisbeek, P. van der Ende, A., van der Avoort, H., Teertstra, R., van
Mansfeld, F., Langeveld, S. (1981) in: "The initiation of DNA
replication' (Ray, D.S., Fox, C.F., eds.), Academic Press, N. Y., pp.
211-232
[28] van der Ende, A., Teertstra, R., Weisbeek, P.J. (1982) Nucleic Acids Res.
10, 6849-6863
[29] Buhk, H.J., Messer, W. (1983) Gene 24, 265-279.
[30] Zechel, K., Bouchg, J.-P. , Kornberg, A. (1975) J. Biol. Chem. 250,
4684-4689
[31] Tabata, S., Oka, A., Sugimoto, K., Takanami, M., Yasuda, S. Hirota, Y.
(1983) Nucleic Acids Res. 11, 2617-2626
[32] van der Ende, A., Teertstra, R., Weisbeek, P.J. (1983) J. Mol. Biol. 167,
751-756
[33] Hirose, S., Hiraga, S., Okazaki, T. (1983) Mol. Gen. Genet. 189, 422-431
3332