67
Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 838 Control of Chromosome and Plasmid Replication in Escherichia coli BY JAN OLSSON ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2003

Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

  • Upload
    others

  • View
    7

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 838

Control of Chromosome andPlasmid Replication in

Escherichia coli

BY

JAN OLSSON

ACTA UNIVERSITATIS UPSALIENSISUPPSALA 2003

Page 2: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Dissertation for the Degree of Doctor of Philosophy in Microbiology presented at Uppsala University in 2003 ABSTRACT Olsson, J. 2003. Control of chromosome and plasmid replication in Escherichia coli. Acta Universitatis Upsaliensis. Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 838. 65 pp. Uppsala. ISBN. 91-554-5627-8 Life is cellular. Cells grow and divide to give two new cells; this process is called the cell cycle. The chromosome in a bacterium is replicated into two identical copies before the cell divides. DNA replication is a fundamental process common to all forms of life.

In my thesis, I have studied control of chromosome and plasmid replication in Escherichia coli, a rod-shaped bacterium. Plasmids are extrachromosomal autonomously replicating DNA molecules.

I have combined the classical Meselson-Stahl density-shift and DNA hybridisation with theoretical analysis of DNA replication. The minimal time between two successive replications of the same molecule, the eclipse, was determined for both plasmid and chromosome.

The aim was to investigate the processes ensuring the precise timing of chromosome replication in the cell cycle. In wild-type strains, the chromosomal eclipse was long. Mutations affecting the so-called sequestration process, the superhelicity of the DNA, and the initiation protein, DnaA, reduced the eclipse.

Fast-growing E. coli has overlapping replicative phases with synchronous initiation from multiple initiation sites, oriC. I have investigated the complex interplay between different control processes by measuring the length of the eclipse and the degree of asynchronous initiation in various mutants.

I have measured the eclipse period of plasmid R1 during up- and downshifts in plasmid copy number. The length of the eclipse was found to be determined by structural events as well as by the properties of the copy-number-control system.

During downshift from very high copy numbers, the rate of plasmid replication started very slowly and gradually increased until the normal copy number was achieved, in accordance with the +n model.

The CopB system of plasmid R1 was shown to be a rescue system preventing cells with few plasmid copies from losing the plasmid in some of the daughter cells. Jan Olsson, Department of Cell and Molecular Biology, Box 596, Biomedical Centre, SE-751 24 Uppsala, Sweden © Jan Olsson 2003 ISSN 1104-232X ISBN 91-554-5627-8 Printed in Sweden by Eklundshofs Grafiska AB, Uppsala 2003

Page 3: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Till minnet av min Far, Rune Olsson

Page 4: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand
Page 5: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

MAIN REFERENCES

This thesis is based on the following papers, which will be referred to in the text by their Roman numerals. I Olsson, J., Dasgupta, S., Berg, O.G. and K. Nordström

Eclipse period without sequestration in Escherichia coli. Molecular Microbiology, 44 (2002): 1429-40.

II Olsson, J.A., Nordström, K., Hjort K. and S. Dasgupta

Eclipse Period and Synchrony of Initiation of Chromosome Replication in Escherichia coli: Roles of SeqA, Dam, DnaA, and Nucleoid Structure. Submitted.

III Olsson, J.A., Dasgupta, S. and K. Nordström

Eclipse period during replication of plasmid R1: Contributions from Structural events and from the Copy-Number-control system. under revision.

IV Olsson, J.A., Dasgupta, S., Berg, O.G., and K. Nordström

Replication of Plasmid R1 during Downshift in Copy number. Manuscript.

V Olsson, J.A., Paulsson, J., and K. Nordström

Effect of the CopB auxiliary Replication Control System on the Stability of Maintenance of Par+ Plasmid R1. Submitted

Reprints were made with the permission of the publisher.

Page 6: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

TABLE OF CONTENTS

INTRODUCTION……………………………………………………………..7

Some general rules for life 7 Two DNA-molecules in the bacterium, the chromosome and the plasmid 7

It is crucial to control DNA replication 7 THE CELL CYCLE OF ESCHERICHIA COLI …………………………………………………9

Asynchrony index 10 Flow cytometry 11 The organisation of oriC 14 Initiation of chromosomal replication 14 Chromosome Replication and its co-ordination to the other cell cycle phase 15

DnaA a protein with many functions 15 The GATC sites in oriC remain hemimethylated for 1/3 of a cell generation 16 Sequestration control 17 Additional roles of SeqA in the cell cycle 17

THE R1-PLASMID…………………………………………………………………...19 Plasmid ecology in bacteria 19 The R1-plasmid 19 The basic replicon 20 The nature of copy number control (CNC) 22 +n mode of plasmid replication 23 Eclipse and the Multistep Control model. 23 The par-locus a true partition system 24

PRESENT INVESTIGATIONS ……………………………………………25 ECLIPSE PERIOD MEASURED BY MESELSON STAHL DENSITY SHIFT………...………25

The eclipse period 25 Previous measurements of the eclipse period 26 Meselson-Stahl Density shift combined with Southern blot 26 Theoretical analysis of DNA replication with different eclipse periods 27 Properties and advantage of the method 28

CONTROL OF CHROMOSOME REPLICATION (papers I and II)………………………..29 Our focus is on the initiation of the chromosome 29 Eclipse period at different growth rates 29 Eclipse period in the absence of sequestration 30 DnaA(ts) and SeqA have additive effect on the eclipse period 32 Superhelicity and the eclipse period 34 Asynchrony index a rough parameter 34 Correlation between eclipse and asynchrony 37

THE NATURE OF PLASMID R1 REPLICATION (PAPERS III AND IV)…………………..39 Theoretical prediction on the Multistep control system 39

R1 plasmid with a temperature-dependent copy number 40 Eclipse period during Upshift (paper III) 40 Copy-number increase at temperature shifts from 30°C to 37, 38, 39 or 42°C 44 Down shift experiment gave surprising results (paper IV) 45 The +n mode of replication fits with our data from the down shift experiments46

Page 7: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

COPB-CONTROL SYSTEM , A RESCUE SYSTEM? (paper V)………………………..…48 Previous experiments on the CopB-control system 48

CopB-control system has a larger effect on the stability of parA+-plasmids 48 DISCUSSION (INITIATION OF CHROMOSOME REPLICATION)…………………...51

Requirements for an eclipse period 52 Correlation asynchrony index against eclipse period 53 Future projects: 53

DISCUSSION (PLASMID)……………………………………………………55 The upshift supports Multistep control 55 Heterogeneous plasmid population 55 Excess CopB reduce the plasmid stability 55

ACKNOWLEDGEMENT………………………………………………… 58 REFERENCES……… ……………………………………………………..…60

Page 8: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

6 Jan Olsson

ABBREVIATIONS AND GENE DESIGNATIONS

∆t the eclipse period A Asynchrony index AAA ATPases Associated with various cellular Activities ADP Adenine Di Phosphate ATP Adenine Tri Phosphate bp base pair CAA Casamino acids CCC Covalently Closed Circular CNC Copy Number Control CopA Copy number factor A CopB Copy number protein B CopT Copy number Target CsCl Cesium cloride Dam DeoxyAdenosine Methyltransferase DNA Deoxy riboNucleic Acid DnaA DNA replication protein A dsDNA double stranded DNA E. coli Escherichia coli et al. Et alii EtBr Ethidium bromide EtOH Ethanol FIS Factor for Inversion Stimulation GFP Green Fluorescent Protein HH Heavy-Heavy dsDNA HL Heavy-Light dsDNA HU Histone like protein? IHF Integration Host Factor IncFII Incompatibility family II kbp kilo base pair LL Light-Light dsDNA Mbp Mega base pair mRNA messenger RNA MukB Cell phenotype (Mukaku, Japanese for seedless) ODXXX Optical Density at the wave length XXXnm ORF Open Reading Frame oriC Origin of replication of the E. coli Chromosome oriR Origin of replication of the plasmid R1 pgm Phosphoglucomutase Pol Polymerase RepA Replication initiator A for plasmid R1 RBS Ribosome Binding Site RNA Ribo Nucleic Acid Rpm Revolutions per minute SeqA Sequestration protein A SMC Structural Maintenance of Chromosomes SSB Single Strand Binding protein ssDNA single stranded DNA Tap Translational activator peptide terC Termination site of the E. coli Chromosome ts Temperature sensitive wt Wild-type

Page 9: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 7

INTRODUCTION

Some general rules for life All organisms are cellular either as a single cell or multicellular. All cells grow in size and then divide in two daughter cells. We call this repetitive process the cell cycle. The cell cycle contains three key processes, genome duplication (replication), genome separation and cell division.

Anatomy, traits, behaviour, metabolism, cell growth and cell division follow instructions specific for each kind of organism. Closely related organisms have similar instructions. Some inventions during the history of evolution are that good so they have been conserved over billion of years; this is reflected by similarities of those instructions even between very diverse organisms. The carriers of these instructions are the chromosomes, giant DNA molecules. The building blocks of DNA are deoxyribonucleotides with four different bases: adenine (A), thymine (T), cytosine (C) and guanine (G). These building blocks function like letters in a book. The numbers and sizes of the chromosomes in different organisms vary among organisms, humans have 46 chromosomes, while most bacteria have only one. However the general rule for all life on earth is that the information manages to be faithfully inherited to the next generations.

It is very important to control the cell cycle properly. Uncontrolled cell cycle can cause cancer and other degenerative diseases. Two DNA-molecules in the bacterium, the chromosome and the plasmid. Most bacteria have only one circular chromosome suspended in the cytoplasm without a surrounding nuclear membrane. With an electron microscope, the chromosome can be seen as an amorphous structure within the bacteria. The chromosome size in a typical bacteria can range from 0.6-6 Mbp (Neidhardt et al., 1990) and is 4.64 Mbp in E. coli (Blattner et al., 1997). In addition to the chromosome, bacteria often contain extrachromosomal DNA-molecules called plasmids. Most of the plasmids like the bacterial chromosome, are circular but much smaller, ca 1kbp-Mbp. The plasmids are not required for normal growth but might become essential under special situation and provide selective/competitive advantage to the plasmid-containing host, e g. many of the plasmids are known to carry genes giving antibiotic resistance, special metabolic enzymes, tolerance to heavy metals, etc. It is crucial to control DNA replication The DNA molecules can be described as selfish molecules that have created diverse systems in order to survive, proliferate, and be stably inherited to the daughter cells at each cell division. DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand from the old DNA-molecule and one newly synthesised strand. Meselson and Stahl showed this in 1958 with an experiment that has been called "The most beautiful experiment in Biology". It is crucial for the DNA-molecules that at least one copy is inherited into each daughter cell. This requires that the DNA-replication is co-ordinated to cell divisions. The frequency of replication should not be higher than that of the host cells, since a too

Page 10: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

8 Jan Olsson

high frequency of replication would lead to too high burden for the cell and with a too low frequency, the DNA-molecule would fail to be stably inherited.

The rate of DNA synthesis is regulated mainly at its initiation of replication. The replicon model was proposed in 1963 according to a replicating unit consists of; requirement of a cis-element "Replicator" (initiation site of DNA-replication) and a trans-acting "Initiator" (a protein) that interacts with the replicator to initiate replication (Jacob et al. 1963). The regulation of initiation of replication has been shown to vary a lot among plasmids and chromosomes. I have focused on the E. coli chromosome and the R1-plasmid. Initiation of replication of the chromosome is dependent on many different factors and proteins (for review see Messer et al., 2001). The initiation frequency changes depending on the growth rate of the host. The growth rate is dependent on the nutrients, temperature, pH, cell density, etc. Therefore if the host changes its growth rate, the DNA-molecule has to change its rate of initiation of replication.

I am interested in control of the cell cycle and particularly in chromosome replication, a fundamental process for all life.

I have used Escherichia coli as the model organism in my studies of control of chromosome replication. I have also studied the replication control of smaller DNA molecules (plasmids) coexisting with the chromosome in the cells. E. coli is a rod-shaped gram-negative bacterium. The plasmid studied in my thesis is the low-copy-number plasmid R1.

Page 11: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 9

THE CELL CYCLE OF ESCHERICHIA COLI

The definition of a cell cycle is the time span from one cell division to the following. Figure 1 describes the three phases of the bacterial cell cycle: pre-replicative (B), replicative (C), and post-replicative (D).

The cell grows continuously during the entire cell cycle. It takes about 40 minutes for the replication forks to move from oriC to the TerC-site. This period is called the C-period. From the end of the C-period it takes about 20 minutes until cell division is completed, this period is called the D-period (Cooper and Helmstetter, 1968). The period between birth of the cell to initiation of replication is called the B-period (Skarstad et al., 1983; 1985). The B- and C-period correspond to the eucaryotic G1- and the S-phase, respectively. The D-period corresponds to the G2-phase, M-phase and the cytokinesis in the eucaryotic cell cycle.

For E. col,i the C- and D-period are relatively fixed in length independent of generation time. The B-period varies at generation times longer than 60 minute. E. coli is known to be able to grow with a generation time shorter than the sum of the duration of the C- and D-periods. This apparent paradox is solved by overlapping replication cycles, initiated already in the mother cell generation or even in the grand mother cell generation (see figure 2). At generation times between 30 and 60 minutes, chromosomal replication is initiated in the previous generation, thus the initiation in the present generation will terminate in the daughter cell. At generation times between 20 and 30 minutes chromosomal replication initiated in the grand mother cell will terminate in the present cell cycle, initiations in the mother cell will terminate in the daughter cell, and initiations in the present generation will terminate in the daughter's daughter cell. At generations longer than 60 minutes the initiation and termination will appear in the same; cell cycle, hence it is obvious that chromosome initiation appears before termination. Terminations occur before initiations for generation times 20-30 minutes and 40-60 minutes (see fig. 2.). At generation times longer than 60 minutes only one oriC copy exists in the new-born cell and will be used for initiation for replication. If initiations appear in the previous generation relative to the

B-period

C-period

D-period

Fig.1. The chromosome (grey or black)is located in the cytoplasm of the bacteria.The bacterial cell cycle is divided in threeperiods. The bacterium grows during thewhole cell cycle. Chromosome replication(black) takes place during the C-period.During the D-period the chromosomes separate and divide the cells

Page 12: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

10 Jan Olsson

termination (30-60 min) two oriCs are initiated. At generation times shorter than 30 minutes, four oriCs are initiated. However, there is always only one termination per cell cycle. The average DNA content increases with increasing growth rate, hence the relative DNA synthesis per cell cycle increase with increasing growth (Cooper and Helmstetter, 1968).

In fast-growing cells, two or four origins are initiated per cell cycle; and all are initiated at the same time; this is called the synchronous phenotype (Skarstad et al. 1986). Even minichromosomes in exponentially-growing cells initiate synchronously at the same time as chromosome replication is initiated (Leonard and Helmstetter, 1986; Koppes and von Meyenburg, 1987). Synchronously replicating cells have 2n (1,2,4 or 8) oriC/cell. In the asynchronous phenotype all origins fail to initiate simultaneously; this results in cells with any integer (1,2,3,4,5,6,7,8,9….) number of oriC in the cells (see figure 5).

Asynchrony index Synchronous initiations are obtained by a lot of control systems. Mutants affected in one or many of the control systems cannot initiate the origins synchronously. It is important to understand the relation between asynchrony and the mutants. For a

C-period

B-period D

Birth of cell

Birth of Mother cell

Birth Of Grand-Mother cell

Initiation of Chromosome replication

150 -50 -100 -150

Time (min)

100

50

20

Initiation of Chromosome replication

Replication fork

Termination of Chromosome replication

Gen

erat

ion

tim

e (m

in)

New born cell

Mother cell

60

40

30

Fig. 2. The cell cycle has three periods, B (black area), C (light grey area) and D (white area) atgeneration times >60 min. Initiation of replication (grey line) starts the C-period. The C-period ends with termination of replication (striped line). Cell division (black line) always occurs 20 min aftertermination, the D-period. The vertical axis indicates the generation time. The horizontal axisindicates time after cell division. At generation times 30-60 min, replications are initiated in the mother generation, and at generation times <30 min replications are initiated in the grandmothergeneration (dot-lines).

The right side of the figure illustrates how the number of replication forks increases withdecreasing generation time from, 1, 2, and finally 4 origins are fired synchronously. Terminationappears prior to initiations at generation times 20-30 min and 40-60 min.

0

Termination ofChromosome replication

Page 13: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 11

practical estimate of the level of asynchrony culture, a J-value was defined as the time interval between the first and the last initiation in one single round of initiation represented as a fraction of the generation time τ (Skarstad et al., 1986). The J-value was calculated from flow-cytometry data on rifampicin-treated cells; rifampicin blocks initiation of replication (see next chapter). Two different asynchrony indices have been defined. Both are calculated from the distribution of origins in the cell population. The first asynchrony index A was defined as the ratio (f3+f5)/f4, were fi is the fraction of cells with i origins (Skarstad et al., 1988). The number of origins is determined from rifampicin-runouts (see next chapter). A new definition of the asynchrony index (A) was used (Boye and Løbner-Olesen, 1990): A=(f3+f5+f6+f7)/(f2+ f4 +f8) (1) Flow cytometry The flow cytometer is a fast and excellent tool in cell-cycle research. It was originally developed for eucaryotic cells but is now frequently used in microbiology laboratories (Skarstad et al., 1995a; Skarstad et al., 1995b). The flow cytometer works essentially like a cell counter in which the cells are transported in a water jet stream through a focal plane of a monochromatic light beam. The light scatter is approximately proportional to the cell size (see fig. 3.). The DNA can be stained with fluorescent dyes, e.g. a combination of mithramycin and ethidium bromide (EtBr) intercalate with the DNA. The monochromatic light has an excitative wavelength of the fluorescent dye and gives rise to pulses of fluorescent light. The fluorescence is a measure of the DNA content per cell and is a linear parameter (Bernander et al., 2001). The strength with the method is that the size (light scatter) and the fluorescence (DNA content) are measured simultaneously for each cell. Thousands of cells can be measured very fast. It is therefore possible to make 2D-plots with size and DNA content on the two axes. (see fig. 4.)

Lamp

Fluorescence (DNA-content)

Light scatter (Cell-size)

Cell containing water-jet

Dichromatic mirror

Filterblock

Fig. 3. Schematic picture of the flow cytometer

Page 14: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

12 Jan Olsson

We can manipulate the cell cycle by different chemical substances. Cephalexin

inhibits cell division and rifampicin stops transcription, which is essential for initiation of chromosome replication. However rifampicin does not stop existing active replication forks from proceeding to completion. In a rif-runout experiment, cephalexin and rifampicin are added. During a rif-runout experiment, all replication forks would finish their ongoing replications, but no further initiations will be allowed, and cell division is blocked, resulting in an integer number of chromosome equivalents, each corresponding to one copy of oriC. The flow-cytometer technique can now be applied to measure the number of chromosome equivalents very easily. This is frequently used to determine the number of oriCs in the cell. Figures 4 and 5 show fluorescence cytograms after rif-runout. The content of other molecules in cells can be detected by flow cytometry. An inducible GFP-gene (Green Fluorescent Protein) was cloned to different plasmids with different copy numbers. The expression was induced and the fluorescence was analysed in separate cells to study the copy-number distribution by flow cytometry (Løbner-Olesen et al., 1999).

1 2 3 4

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4

1 2 3 4

MG1655 wt MG1655 seqA

size

size

Histograms

DNA-content

expo

nent

ial

rif-r

unou

t st

atio

nary

size DNA-content

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8 1 2 3 4

1 2 3 4

1 2 3 4

2D-plot

DNA-content DNA-content

Histogram 2D-plot

Fig. 4. Flow-cytometry data on cell-size (columns 1 and 4) and DNA-content distributions (columns 2 and 5) presented as histograms. In columns 3 and 6 cell size and DNA content are presented as 2-D plots. The cells were grown in M9 medium supplemented with glucose. Flow-cytometry data are presented for the wild-type strain (MG1655) and SeqA-deficient strain (MG1655 seqA) with synchronous and asynchronous initiation of replication, respectively. The top row presents data at exponential growth, the middle row rif-runout cells, and the bottom row cells in stationary phase. The difference between synchronous and asynchronous replication is well visualised in the rif-runouts; the DNA of the wild-type strain ends up in two peaks corresponding to 2 and 4 chromosome equivalents, while the DNA of asynchronous seqA-strain ends up in 1, 2, 3, 4, 5, 6, 7 and even eight chromosome equivalents.

size

Page 15: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 13

.

1

2

2

4

4

2

8

4

1

8

2

3

3

4

5

6

7

9

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

ASY

NC

HR

ON

Y

SYN

CH

RO

NY

RIF-RUNOUT

A

B

C

D

Fig. 5. Rif-runout experiments with cells with synchronous (A-C) or asynchronous (D) initiation of chromosome replication. The branched lines to the left indicate chromosomes with ongoing replication moving to the right. Non-branched lines correspond to non-replicating chromosomes that are the result of a rif-runout. The right side visualises the number of chromosome equivalents after rif-runout as determined by flow cytometry. A-C show examples of chromosomes in cells at different stages of replication and at different growth rates (Generation time; A= >60 min; B=30-60 min; C=<30 min). In synchronously-replicating cells, the cells after rif-runout end up with 2n chromosome equivalents (n is an integer) e g. 1,2,4 and 8. Asynchronous cells, e g. mutants affected on the mechanisms assuring the synchrony, end up also with chromosome equivalents ≠2n (3,5,6,7,9...) see figure 5.D. In the asynchronous cells, the separate initiations are scattered all over the cell cycle.

Page 16: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

14 Jan Olsson

The organisation of oriC Replication in most bacterial chromosomes starts at one specific site, called oriC. The definition of an origin is the minimal DNA sequence capable of autonomous replication. The length of minimal oriC is 258 bp and was estimated by deletion analysis of minichromosomes (Messer et al., 1978 & 1979; Oka et al., 1980, Asai et al., 1990). The origins of six Enterobacteriaceae have similar organisation reflecting conservative features (Zyskind et al., 1986). The organisation of oriC in E. coli is: one AT-rich box (12 bp), 3 AT-rich 13-mers starting with a GATC-site and five 9-mer DnaA-boxes. In addition another nine GATC-sites and seven, 6-mer ATP-DnaA-boxes are scattered over the oriC-sequence (see fig. 6) (Roth and Messer, 1995; Schaper and Messer., 1995; Oka et al., 1980; Fuller et al., 1984; Messer and Weigel 1986; Matsui et al., 1987). Initiation of chromosomal replication Initiation of chromosome replication in E. coli is a very complex process, not yet fully understood. Many different proteins, protein complexes, and processes act sequential during the initiation. Replication is bi-directionally initiated from oriC (Bird et al., 1972; Marsh and Worcel, 1977)

A theory of replication was first proposed 1963, in which the cis-element "Replicator" and the trans-acting "Initiator" interact to initiate replication (Jacob et al., 1963). In E. coli is oriC the cis-acting "replicator" and the DnaA protein is the trans-acting "initiator". DnaA binds to the mini-oriC consists of three AT-rich 13-mer tandem repeats and five 9-mer DnaA-boxes, see figure 6 (Oka et al., 1980). About 20-40 ATP-bound DnaA monomers bind to four of the five DnaA-boxes and introduce a 40° bend at each DnaA box followed by an unwinding of the three 13-mer repeats (Fuller et al., 1984; Crooke et al., 1993; Schaper and Messer, 1995; Langer et al., 1996). The binding of DnaA and the bending of the DnaA-box forms the 'initiation complex', and the unwounded structure is called the 'Open Complex' stretching 28 bp without single-strand-binding protein (SSB) and 44 bp with SSB-protein bound. This process is catalysed by three accessory proteins FIS, IHF and HU (Messer et al., 2001). The DnaB-hexamer ring structure is guided by six DnaC proteins to the

oriC →→→→5'- cggcttgaga A+T-rich box L-13-mer M 13-mer R 13-mer aagacctggG ATCctgggta ttaaaaagaa GATCtattta ttt aga GATC tgttctatt g t GATCtctta ttag GATCgc R 1

actgccc tgt ggataa caag GATCcggctt ttaaGATCaa caacctggaa agGATCatta taagctggGA TCagaatgag

Me R2 R3

ac tgtgaatG A TCggtGATC ctggaccgta ggg ttataca ca actcaaaa actgaacaac agttgttc tt tggataa cta

R4

ccggttGATC caagcttcct ga ←←←←oriC cagag tta tccaca gtag -3'

= ATP-DnaA-boxes 6-mers VERSALES=GATC-sites One frame = AT-rich region

Double frame =AT-rich 13-mer region Triple frame = DnaA-box 9-mer

Fig. 6. The sequence of oriC and its organisation.

Page 17: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 15

unwounded oriC region and thereby expanding the open bubble to about 65 bp before DnaC leaves the complex. The result is one active DnaB hexamer on the 5'-end of each strand called the pre-priming complex. The DnaB hexamers move in a 5'-3'-direction to open up the bubble long enough for the primase (DnaG) to bind at the 5' end of the DnaB hexamers (Fang et al., 1999). The primase synthesises the first RNA-primers. Two sliding-clamps, the β-subunits of DNA Pol III, and two Pol III dimers are assembled onto the primed site. Chromosome Replication and its co-ordination to the other cell cycle phases. At moderate and high growth rates there are overlapping replication cycles, and replication is initiated at two or more origins per cell and cell cycle (see figure 2). During rapid growth many oriC are initiated for replication, and they are all initiated at the same time (Skarstad et al., 1986). These multiple initiations have to be co-ordinated to avoid asynchronous initiation of replication. Many parallel control mechanisms guarantee the synchrony. (i) A titrating DnaA-binding site datA duplicates soon after initiation of replication and thereby titrates away the DnaA-protein. (ii) The SeqA protein prevents Dam-methylase to remethylate the 11 hemimethylated GATC-sites in the newly replicated oriC region by binding to them. (iii) SeqA inhibits the transcription of the dnaA gene. (iv) SeqA prevents DnaA to bind to oriC. (v) The active ATP-DnaA is hydrolysed to inactive ADP-DnaA by the β-subunit of the DNA-polymerase Pol III). DnaA a protein with many functions The DnaA is the initiator protein for the chromosomal replication (Skarstad and Boye, 1994; Messer and Weigel, 1996). The DnaA is a 52 kDa protein belonging to the large protein family AAA+ (ATPases Associated with various cellular Activities) found in all three domains of life. The AAA+ class of ATPases have common motifs. Many of them are initiator proteins of DNA replication but they also have other roles in DNA metabolism like remodelling and loading of proteins. DnaC is also a member of this AAA+ family (Neuwald et al., 1999). An average E. coli cell contains 1000-2000 DnaA molecules (Sekimizu et al., 1987; Bramhill and Kornberg, 1988). DnaA binds as a monomer with its C-terminal domain 4 to the DnaA-boxes with the consensus 9-mer sequence 5'-TTA/TTNCACA-3' (Roth and Messer, 1995; Schaper and Messer). DnaA binds to the 9-mer DnaA-boxes complexed with either ATP or ADP with the same affinity (Sekimizu et al., 1987). DnaA binds with the same affinity to linear DNA as to supercoiled DNA (Weigel et al., 1997). However, only the ATP-DnaA can open up the AT-rich region upstream to the five DnaA-boxes by co-operative binding of ATP-DnaA to the 6-mer ATP-DnaA-boxes. DnaA autoregulates its own expression if ATP is bound to it (Speck et al., 1999). The transcription of the dnaA-gene is known to be cell-cycle-dependent; The transcription decreases almost 20-fold after initiation of chromosome replication during ca 15 minutes. (Campbell and Kleckner, 1990; Bogan and Helmstetter, 1997). The dnaA-gene has two promoters with two DnaA-boxes and three ATP-DnaA-boxes in between them (Kücherer et al., 1986; Polaczek and Wright, 1990). DnaA is known to act as an activator, repressor or terminator of transcription. DnaA regulates many genes through different locations and arrangements of DnaA-boxes (for review, see Messer and Weigel, 1997)

DnaA protein is found in cell membranes and there are strong indications that DnaA interacts with the cell membrane and vice versa. (Sekimizu et al., 1988; Newman and Crooke, 2000). DnaA mutants affect the composition of fatty acids in

Page 18: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

16 Jan Olsson

the phospholipids of cell membranes (Suzuki et al., 1998). Acidic phospholipids decrease the binding-efficiency of adenine nucleotides to DnaA and, hence, speed up the reactivation of ADP-DnaA with ATP (Sekimizu and Kornberg, 1988). Phospholipids also decrease DnaA's affinity to oriC (Crooke et al., 1992). After initiation of chromosome replication, the hemimethylated oriC is translocated to the membrane, suggesting that phospholipids in the membrane interact with the oriC and DnaA complex (Ogden et al., 1988; Landoulsi et al., 1990). DnaA has a distinct membrane-binding domain centred around residue 372 (Garner and Crooke, 1996; Garner et al., 1998).

Chromosome replication is initiated at a specific cell mass per oriC (Donachie W. D., 1968). The concentration of DnaA sets the initiation mass. Overproduction of DnaA decreases and underproduction of DnaA increases the initiation mass (Langer et al., 1996; Skarstad et al., 1989; Løbner-Olesen et al., 1989; Atlung and Hansen, 1993). DnaAts mutants have an elevated initiation mass at permissive temperature 30°C (Boye et al., 1996).

When a chimerical pBR322 derived plasmid with an oriC region was transformed into an E. coli cell, the concentration of DnaA was increased. The proposed explanation was that the DnaA-boxes titrated DnaA molecules. To respond to this titration, the cell increased DnaA expression to maintain the concentration of free DnaA constant (Hansen et al., 1987). The initiator-titration model was first suggested as a possible replication-control mechanism by Hansen et al., (1991). At the end of the last century, a site was found which binds DnaA more efficiently and in eight times higher amounts than oriC. This so-called datA-locus is a ca 1 kbp sequence located at 94,7 minutes on the physical map and clockwise to oriC. The datA-locus binds eight times more DnaA proteins than the region containing oriC and mioC. Therefore, when the datA locus is replicated, the DnaA-binding capacity will double and thereby withdraw free DnaA, and the intracellular concentration of free DnaA is reduced. The importance of the datA locus was confirmed when a strain with a deleted datA locus showed a phenotype in which chromosome replication was overinitiated. Mutations in either DnaA-box 2 or 3 in the datA locus gave the same results as deletion of datA. The asynchrony and overinitiation with the deleted datA locus could be recovered with a DnaA-titrating plasmid containing the DnaA-boxes from mioC and an oriC in which the left AT-rich 13-mer region was deleted (Kitagawa et al., 1996; 1998). The GATC sites in oriC remain hemimethylated for 1/3 of a cell generation Except for the AT-rich region, the three 13-mers and the five 9-mer DnaA-boxes oriC contains 11 GATC-repeats (see fig. 6.). The GATC sites are scattered all over the chromosome but are 10 times more frequent at oriC (Meijer et al., 1979; Oka et al., 1980). The adenine in a GATC site is normally methylated. The nucleotides incorporated during replication lack the methyl group and the newly replicated DNA remains hemimethylated until dam methyltransferase remethylates the nascent DNA-strand. It appears that the GATC sites need to be re-methylated before initiation of replication can start in vivo. The transformation frequency of methylated oriC-plasmids was much less than that of unmethylated plasmids in dam-deficient cells. Plasmids with ColE1 origin instead of oriC was transformed more efficiently in the dam-deficient strain (Messer et al., 1985; Smith et al., 1985). The fully methylated oriC-plasmids that did transform the dam-deficient strain replicated one time and were then accumulated as hemimethylated DNA unable to replicate any further (Russel and Zinder, 1987). Hemimethylated DNA, however, is as good a substrate for

Page 19: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 17

in vitro replication as fully methylated DNA (Boye, 1991). It was shown that hemimethylated oriC DNA was associated with the membrane fraction; this was proposed to be the sequestration of newly replicated oriC-region (Ogden et al., 1988). Further description of the sequestration mechanism is found in the next chapter. The timing of chromosome replication is dependent on the level of dam expression. The dam expression was varied with temperature by the use of a dam gene downstream of the lambda promoter λPR and the temperature sensitive repressor cI587 in a host with the dam-3 mutation lacking methylation activity. The strain was synchronous within a narrow temperature interval (the asynchrony index (A) dropped at 34-39°C). The strain was asynchronous at lower and higher temperatures when the dam gene was obviously under- or over-expressed (Boye and Løbner-Olesen, 1990).

Specific restriction enzymes can distinguish between methylated and hemimethylated 6-mer sequences containing GATC-sites. DnaCts mutants can be synchronised by a period at non-permissive temperature. The kinetics of re-methylation after initiation of chromosome replication in synchronised DnaCts mutant was studied by restriction analysis. Both the oriC region and the dnaA promoter P2 region were hemimethylated for a significantly longer period than other regions on the chromosome. Thus initiation of chromosome replication and dnaA transcription were both inhibited (Campbell and Kleckner, 1990). When SeqA is overexpressed, the period of hemimethylation is prolonged, and segregation of chromosomes and cell division are delayed for more than one cell generation (Bach et al., 2003). Sequestration control The repressor of re-initiation was identified as the SeqA protein responsible for sequestration of the hemimethylated GATC sites in oriC. There are about 1000 SeqA molecules per cell (Slater et al., 1995). The SeqA protein can bind fully methylated GATC-sites but prefers hemimethylated and does not bind to unmethylated DNA at all (Slater et al., 1995; Skarstad et al., 2000). In a seqA strain, the synchronisation of initiation is lost and the sequestration from the dam methylase is lost as shown by flow cytometry and restriction analysis, respectively (Lu et al., 1994). The absence of SeqA decreases the initiation mass about 10-20% (Boye et al., 1996). SeqA protein binds more efficiently if two or more hemimethylated GATC sites are clustered with a distance of up to 31 bp (Brendler et al., 1999; 2000).

The temporary suppression of re-initiation of replication due to inhibition of remethylation of hemimethylated GATC sites in oriC by the binding of SeqA is called sequestration control. Additional roles of SeqA in the cell cycle. The SeqA protein most probably has additional roles than sequestration by the binding to hemimethylated GATC-sites.

SeqA can inhibit replication of both fully methylated and unmethylated oriC-plasmids in vitro, fully methylated oriC-plasmids are however more sensitive to SeqA than the unmethylated. The pre-priming complex is the complex formed by oriC-DNA, after the formation of the open complex, when DnaB and DnaC-proteins binds to it. SeqA inhibits a proper assembly of the pre-priming complex. SeqA inhibits initiation even when added to already properly formed pre-priming complexes (Wold et al., 1998)

The nucleoid of SeqA-deficient strains has an increased degree of negative supercoiling, deduced after measure of sedimentation coefficients in sucrose density gradients (Weitao et al., 1999 and 2000). Topoisomerase I encoded from topA is

Page 20: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

18 Jan Olsson

known to decrease negative superhelicity (Wang, 1971). SeqA can bind non-specifically to supercoiled DNA, protecting the supercoiled oriC plasmids from Topoisomerase I activity. SeqA also restrains unmethylated supercoiled oriC plasmids from activity by Topoisomerase I. This might be the mechanism by which SeqA inhibits the formation of the open complex (Torheim and Skarstad, 1999).

The MukB-protein is structurally closely related to the SMC-proteins (Structural Maintenance of Chromosomes)(Niki et al., 1991). The MukB protein increases nucleoid condensation and the degree of negative superhelicity, and the null mutant has a very reduced nucleoid condensation and superhelicity. The null mutant is also temperature sensitive. The null mutant of seqA shows increased nucleoid condensation and increased negative superhelicity. The double mutant seqAmukB has regained the wild-type condensation. This suggests that SeqA has an additional role as the opposite force to the condensation and negative superhelicity ability of MukB (Hu et al., 1996; Weitao et al., 1999 and 2000).

SeqA is localised intracellularily at a distinct foci, in a mukB null mutant the SeqA protein is scattered all over the nucleoid in foci with various size (Onogi et al., 1999). The SeqA foci are not dependent on the oriC-region shown by an oriC-deleted mutant. The number of SeqA foci is proportional to the number of oriC at different growth rates. Its also interesting that the oriC and SeqA foci are not co-localised (Hiraga et al. 1998). Overexpression of SeqA does not increase the number of foci; rather the intensity of each focus is increased (Bach et al., 2003).

SeqA is proposed to be involved in chromosome organisation and segregation during replication and throughout the cell cycle (Dasgupta et al., 2000). Overexpression of SeqA inhibits cell division (Hiraga et al., 1998; Torheim and Skarstad, 1999; Brendler et al., 2000; Bach et al., 2003).

SeqA is reported to prevent the formation of the open complex at high concentrations (Torheim and Skarstad, 1999). SeqA preferentially binds to the region (the left part of minimal oriC) for the initial strand separation (Slater et al., 1995).

The oriC-DNA is associated to the membrane through a multi-protein complex including SeqA and a factor called SeqB, (Shakibai et al., 1998).

Page 21: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 19

THE R1-PLASMID

Plasmid ecology in bacteria The bacterial chromosome is a large DNA molecule containing all genes essential for living, metabolism, proliferation, etc. Since bacteria do not reproduce sexually, this means that mixing genes in the next generation by mating does not drive the evolution. It is now obvious that many bacteria, maybe all bacteria, share a common gene pool. The bacteria can obtain new genetic material by uptake of free DNA molecules, by transformation (Griffith, 1928). There are numerous shuttle mechanisms for exchanging genetic materials and they can transfer new genetic material between members in the "common gene pool" (Dionisio et al., 2002). Bacterial viruses (bacteriophages) transfer DNA in lipid and protein capsules. Some shuttles are naked DNA mostly circular molecules that can be transferred by a direct cell-to-cell contact a mechanism called conjugation. Every shuttle has its own specific replication mechanism but they use more or less all proteins from the replication machinery of the host, e g. DNA polymerase, primase, SSB etc. These naked shuttles had many different names until they got the name plasmids (Lederberg, 1952). Conjugation requires products from the "transfer genes" (tra) encoded within the plasmid itself.

The plasmids have a parasitic behaviour and are sometimes described as selfish DNA. They use the metabolites and the replication, transcription and translation machineries of the host. The growth rate is sometimes lower in plasmid-containing than in plasmid-free cells. Plasmids have many exciting and diverse strategies to survive and colonise new hosts. Some plasmids have a killer system, which means that they produce a stable toxin and an antidote with short lifetime. This ensures that the daughter cells, which have lost their plasmid, would die since the antidote would decay while the toxin is stable. Some plasmids have an almost altruistic behaviour where a low frequency of the plasmid containing cells commit suicide by lysis to spread the exotoxin which kills plasmid free cells; e.g. the ColE1 plasmid produces the toxin colicin E1 that kills cells without the plasmid. Some plasmids produce conjugation inhibitors that prevent other plasmids to enter the same host. The plasmids became more interesting when scientist discovered that many plasmids contain genes responsible for antibiotic resistance and are responsible for the spread of antibiotic resistance among pathogen bacteria. Plasmids can give its host other kinds of resistance e.g. against heavy metals, bacteriocins, radiation and toxic anions. In this situation, the parasite host relation is more symbiotic, since the host gain some benefits. Plasmids seem to be selfish DNA molecules spreading resistance against various antibiotics. However, many plasmids today are very important tools in gene technology and science e.g., as cloning vectors and expression vectors. The R1-plasmid Plasmid R1 (resistance plasmid) was discovered as an extrachromosomal factor that could transfer resistance to many antibiotics between cells (Meynell and Datta, 1967). Plasmid R1 belongs to the incompatibility group FII (IncFII).

Page 22: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

20 Jan Olsson

R1 is a low-copy plasmid with about 3 copies per newborn cell (Nordström et al., 1980). Low-copy plasmids need many extra systems (e.g. active partition, toxin-antitoxin system) to secure their stability and are therefore larger than high-copy plasmids; the size of R1 is about 90 kb. Plasmid R1 also has a conjugation system encoded by the tra-operon. Several antibiotic-resistance genes are carried by R1 to confer upon the host cell resistance to kanamycin (Km), ampicillin (Ap), sulfonamides (Su), streptomycin/spectinomycin (Sm/Sp) and chloramphenicol (Cm). The parD locus encodes one of the two killer systems, the kis gene encodes the poisonous product and the kid gene encodes the antidote (Bravo et al., 1987; Ruiz-Echevarria et al., 1991). The second killer system is the hok/sok system encoded from the parB locus. The hok-product is a poisonous protein and the sok product is an antisense RNA that inactivates the hok-mRNA (Gerdes et al., 1986a and b; Rasmussen et al., 1987).

The parA region with the ParM and ParR proteins and the adjacent parC region, a centromere-like DNA region to which the ParR protein binds encode the true partition system. The parC region complexed with ParR functions as a nucleation point for the actin-like polymerisation of the ParM protein (Møller-Jensen et al., 2002; Jensen and Gerdes, 1999; Dam and Gerdes, 1994). The basic replicon (see figure 8A.) The basic replicon of R1 plasmid is 2,5 kb (Kollek, et al., 1978; Molin et al., 1979). The activator of replication is the RepA-protein. Replication is initiated by the binding of many cis-acting RepA-proteins and trans-acting DnaA-proteins. R1 replicates mainly uni-directionally with a low frequency of the rolling-circle mode of replication (Krabbe et al., 1997). The RepA protein has to be synthesised de novo between two subsequent replications of the same plasmid (Austin and Nordström, 1990; Masai et al., 1983; Masai and Arai 1987). However the DnaA protein is not essential for in vivo initiation but the initiation frequency is decreased if DnaA activity is shut off (Bernander et al., 1991). Synthesis of the RepA protein requires a

Fig. 7. Schematic map of plasmid R1.The tra-region is a group of genesinvolved in conjugation. The killersystem are parB and parD. The partition system is encoded by the parAoperon. Five genes responsible forresistance against antibiotics aredenoted KmR. ApR, SuR, SmR and CmR. The basic replicon is enlarged to showthe positions of genes involved inregulation of replication of the plasmid(see figure 8).

R1 ca 90 kb

tra

parB parA

CmR

parD

Basic repliconca 2,5 kb

PrepA

KmR

SmR

SuR

PcopB

PcopA

copA copB tap

copT

oriR1 RepA

ApR

Page 23: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 21

translational coupling of the small leader peptide, Tap (translational activator peptide), located just in front of the ORF for RepA. The ribosome-binding site (RBS) is sequestered in a secondary structure in the mRNA until the upstream translation of tap-region opens it. The distance between the RBS of repA and the start codon is not optimal (Öhman and Wagner 1991; Blomberg et al., 1992 and 1994). The RepA-mRNA is transcribed from two promoters PrepA and PcopB, the latter is constitutively expressed. The 2,5 times stronger PrepA is downstream of PcopB. The transcript from PcopB also contains the mRNA for CopB. The CopB protein is 86 aa and 33kDa protein. There are about 2000 CopB-monomers per cell, they bind as tetramers to PrepA and reduce the transcription from PrepA by 90% (Light and Molin, 1981and 1982a; Riise and Molin, 1986;Womble et al., 1985; Light et al., 1985). Reduced concentration of CopB reduces drastically the concentration of tetramers this turns on PrepA. The CopB-loop is not believed to play much of a role under steady-state conditions. It is more probable that the CopB-loop has an important role at very low copy number, e.g. after conjugation to establish wild-type copy number or when a daughter cell only gets one single copy after cell division (See figure 8.;Light and Molin, 1982b; Light et al., 1985; Womble and Rownd, 1986; Riise and Molin, 1986, Nordström and Wagner, 1994).

PrepA PcopB

PcopA copA copB tap

copT

oriR1 RepA

CopA

Low concentration of CopA due to low gene dosage and

convergent transcription

cis-actingRepA

No CopB-tetramers

RepA-mRNA

CopB & RepA mRNA

Very low copy number B.

Fig. 8. The R1 basic replicon during normal copy number (A) and at very low copy number (B).The CopB-system is believed to be a rescue system at very low copy number, e g. after conjugationor newborn cells with only one plasmid. At normal copy number PrepA is repressed 90% by theCopB-tetramer. At very low copy number, CopB tetramers are rarely formed due to the low genedosage; therefore the stronger promotor (PrepA) will be depressed. The result is a higher mRNAconcentration and hence a higher RepA production; this might assure enough replication before thefollowing cell division.

promotor transcription/translation inhibition RNA

PrepA PcopB

PcopA copA copB tap

copT

oriR1 RepA

CopA

trans-acting CopA

cis-actingRepA

CopB- tetramer

RepA-mRNA

CopB & RepA mRNA

kT

kA

Normal copy number A.

kapp kDA kR

Decay of CopA

Page 24: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

22 Jan Olsson

The inactivator CopA, an antisense RNA, controls the copy number. The half-life time of CopA is 1-1.5 min (Stougaard et al., 1981; Söderbom et al., 1997; Söderbom and Wagner, 1998). Therefore, the concentration of CopA is proportional to the copy number which leads to a rapid adjustment of plasmid replication upon changes in copy number. The target for CopA is the complementary region CopT in the tap-gene (translational activator peptide). CopA is transcribed 2 bp upstream of the RBS of tap. CopA inhibits translation of this leader peptide when it binds to CopT. The binding starts with a reversible loop-loop interaction step called the Kissing complex; this step proceeds irreversibly to the Extended kissing complex and the Stable CopA-CopT complex (Light and Molin, 1981; Nordström and Aagaard-Hansen, 1984; Nordström and Wagner, 1994; Malmgren et al., 1996; Kolb et al., 2000). The copy number Cop is primarily determined by: Cop=k kR kDA kT /kA kapp kH (2) where kR is the rate of replication, kDA is the rate of decay of CopA, kT the rate of transcription of RepA-mRNA from PcopB and PrepA, and kA is the rate of transcription of the antisense RNA CopA, kapp is the association rate of CopA and the CopT-region of the RepA-mRNA, kH is the growth rate of the host (Nordström and Wagner, 1994). The rate of transcription of CopA, kA decreases if the transcription from PcopB and PrepA increases by convergent transcription leading to an increased copy number Cop (Stougaard et al., 1982; Nordström and Uhlin, 1992). The nature of copy number control (CNC) The plasmid has to synchronise its replication with the growth of the host cell. A too low replication frequency would lead to plasmid loss. A too high replication frequency would lead to an overcrowded plasmid population resulting in a too heavy metabolic load (Proctor, 1994; Paulsson and Ehrenberg, 1998, Wouters et al., 1980). The main properties of a copy-number control is that the plasmids replicate more than once per cell generation if the copy number is lower than the average and replicates less than once per cell generation if the copy number is too high. Hyperbolic copy number control (CNC) versus On-Off CNC was discussed in the review Nordström et al. from 1984. The relative copy number was modified by metabolic shifts,

1 2 3

4

3

2

1

4

Rep

licat

ions

/pla

smid

in

eac

h ce

ll cy

cle

Relative copy number

A

B CFig. 9. The behaviour of different Copy-number-control strategies: A No control, each plasmid replicates once in each cell cycle. B Hyperbolic copy-number control; the replication frequency is inversely proportional to the copy number. C On-off control; small changes around the normal copy number switch replication on or off. Within the light grey box, the plasmid will be lost. In the dark grey area, replication is uncontrolled, run away replication (Nordström et al., 1984)

Steady state

Page 25: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 23

temperature shifts with temperature-dependent copy number mutants and temperature shifts with R1 cloned into a temperature-sensitive vector. All these changes coincide with line B in figure 9, supporting the hyperbolic model (Nordströmet al., 1984; Nordström, 1990). If the relative copy number is under 1 and the replication frequency under 1/cell cycle, the plasmid would be lost. If the relative copy number is above 1 and the replication frequency above 1/cell cycle the plasmid will replicate uncontrolled i.e. run away phenotype (see figure 9.). +n mode of plasmid replication The model deduces that if the average copy number in new-born cells is n, on average n plasmid, with a Poissonian distribution would be produced per cell before cell division (Nordström and Aagaard-Hansen, 1984). Flow-cytometry experiments with a cloned GFP-gene (Green Fluorescent Protein) on R1-plasmid did show a fairly wide copy-number distribution, maybe Poissonian (Løbner-Olesen, 1999). It was however impossible to correlate it to exact numbers as it is not clear how proportional the GFP signal was to the actual copy number.

The +n model was supported experimentally from copy number upshifts and downshifts with temperature-sensitive copy-number mutant pKN301 as well as with metabolic shifts. Plasmid synthesis was followed by pulse labelling with [3H]-thymidine and the rate of plasmid synthesis was rather changed in one single step then gradually (Gustafsson and Nordström, 1980). Eclipse and the Multistep Control model. Replication of plasmid R1 is random, i.e. spread in time over the entire cell cycle and all copies have the same probability of being selected for replication (Gustafsson et al., 1978). The newly replicated plasmid is, however, sequestered from new initiation during ca 0.2 generations. After this eclipse period the plasmid has the same probability as the other plasmids in the cell to be selected for the next replication. This eclipse might have two additive causes, the structural and the kinetic factor. A newly replicated plasmid is open-circular, it has to regain the covalently-closed-circular (CCC) structure to be a good substrate for a new round of replication. Plasmid replication is dependent on right superhelicity (Uhlin and Nordström, 1985). This cycling from a CCC parent plasmid to CCC daughter plasmid takes about 4 minutes (Gustafsson et al., 1978; Koppes and Nordström, 1986).

The second cause is derived from a hypothetical model for CNC and has only been investigated theoretically. The model for this second factor is based on many intermediate states (Xi and Zi; see figure 10) for the newly replicated plasmid to regain the potential to replicate again (Ehrenberg and Sverredal, 1995). The model partly explains the eclipse measured by Koppes and Nordström (1986). Simulated data could imitate the experimental eclipse but the kinetics of the probabilities to replicate after the eclipse period was different. The simulations predicted 2-3 steps indicating that each mRNA translates a burst of RepA proteins and the cycle needs about 2-3 bursts this means that the plasmid visit the Xi states 1-2 times on average (Ehrenberg and Sverredal, 1995; McAdams & Arkin, 1997). Paulsson and Ehrenberg (2000) further discussed clock-like behaviour of the multistep model. The multistep control was shown to be probable for the replication of the ColE1-plasmid supported by molecular background and theoretical analyses of the kinetics (Ehrenberg, 1996; Paulsson et al., 1998).

Page 26: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

24 Jan Olsson

The par-locus a true partition system The par-locus (formerly parA) of plasmid R1 is one single operon encoding two trans-acting proteins ParM and ParR. Besides these two proteins, the par-locus also has one cis-acting centromere-like region called parC, upstream of parM and parR. The ParR autoregulates the promotor (Min et al., 1991;Dam and Gerdes, 1994; Jensen et al., 1994; Gerdes and Molin, 1996). Electron microscopy and ligation kinetic assay indicates pairing of two parC, containing plasmids in presence of ParR, and pairing was even more efficient in presence of ParM and ATP. The assay showed that supercoiled plasmids are better substrate for pairing than open-circular plasmids. (Jensen et al., 1998). R1 plasmids are clustered at a few foci and not scattered all over the cell. In high copy mutants is the fluorescent intensity of each foci instead of the number of foci is increased (Weitao et al., 2000; Pogliano et al., 2002). In vivo ParM is co-localised with parC-containing plasmids (Jensen and Gerdes, 1999). The parC-ParR-complex forms the nucleation point of an actin-like polymerisation of a spindle-like filament responsible for the plasmid segregation. Polymerisation requires ATP and breakdown of the filaments is through hydrolysis (Møller-Jensen et al., 2002).

Fig. 10. Scheme for the R1 replication cycle with n cis-acting RepA molecules. The Zi state is a plasmid with i RepA molecules bound to oriR1 and an active mRNA. Inactivation of mRNA by CopA (rate q) leads to state Xi. The rate constant q is dependent on the rate of CopA transcription and the rate of CopA/CopT duplex formation. Transcription of a new mRNA (rate kT) returns the plasmid to state Zi. Translation of RepA molecule (rate kIT) of a plasmid in state Zi moves the plasmid to state Zi+1. (Ehrenberg and Sverredal, 1995)

X1 X2 Xn-1X0

Z0

Z0 Z2Z1 Z0

kIT

Zn-1 Z0

Zn-1 kIT kIT kIT

kIT kT

kIT

kT kT kT q q q q

kIT

Page 27: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 25

PRESENT INVESTIGATIONS & DISCUSSION

My work has been focused on initiation of chromosome and plasmid replication. A key feature of chromosome replication is that it is tightly controlled to ensure that replication occurs once and once only during one cell cycle. One aspect of this control is the eclipse period, the period during which re-initiation of replication is prohibited. Reduction of the length of the eclipse period would then effect the fidelity of chromosome replication. I have therefore studied what genes determine the length of the eclipse period, and the correlation between the length of the eclipse period and synchrony of initiation of replication (papers I and II)

There is also an eclipse during plasmid replication. I have studied the causes of the eclipse period by comparing the eclipse period in a normal plasmid population with a plasmid population replicating without copy number control; both structural aspects and the copy-number-control system contribute to the eclipse period (paper III and IV)

Finally, plasmid R1 has an auxiliary copy-number-control system, CopB. I have determined the efficiency of this system in assisting the maintenance of plasmid R1 in a growing population of host bacteria (paper V).

ECLIPSE PERIOD MEASURED BY MESELSON STAHL DENSITY SHIFT

The eclipse period The E. coli chromosome has been shown to initiate once and only once per generation. This means that further initiation of replication from oriC is repressed during most of the cell cycle. The inter replication times are distributed around an average of one generation time τ. Depending on how accurate the control of replication is, the distribution of the inter replication time will vary. The minimal inter replication time is defined as the eclipse period (see figure 11.). It has been of interest to microbiologists to measure the eclipse period in order to investigate how cell-cycle mutations affect the timing of reinitiation of chromosome as well as plasmid replication.

10,15 0,4 0,6

Inter replication time (generation times τ)

P(ne

xt r

epli

cati

on) Fig. 11. The distribution of inter-

replication times have an average of 1generation time (τ) in steady state. Theeclipse period is the minimal inter-replication time. Strains with differenteclipse periods (indicated with arrows)have different distributions.

0

Page 28: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

26 Jan Olsson

Previous measurements of the eclipse period Many experiments have been done for estimation of the minimum inter-replication time, which we call the eclipse period. One method is Meselson-Stahl density-shifts experiments in which a culture is shifted from a medium with heavy isotopes to a medium with light isotopes. The DNA before a density shift contain two heavy strands (HH), after one round of replication in light medium two hybrid DNA molecules (HL) are formed. Further replication of HL-DNA results in one hybrid-DNA (HL) and one light DNA molecule (LL). Koppes and Nordström (1986) estimated the eclipse period by Meselson-Stahl density shift experiment in which they grew the cells in heavy medium and pulse-labelled the DNA with [3H]-thymidine just before they shifted the cultures to light medium. They followed DNA replication by harvesting DNA at different times. Each DNA sample was separated on a CsCl-density gradient by ultracentrifugation. The recently pulse labelled DNA will remain in the HH fraction until it replicates a second time, this time in the light medium, and then appear in the HL fraction. The time between the density shift and the appearance of radio-labelled HL-DNA is a measure of the eclipse period. It is important to point out that the HL-fraction will start to contain DNA directly after the density shift but the pulse labelled DNA is prohibited for a subsequent replication during the eclipse period. Koppes and Nordström measured the eclipse period for the integrated R1-origin in the chromosome, other groups have measured the inter-replication time on chromosome origins (Bakker and Smith, 1989; Smith et al., 1992). The same method has been applied for other integrated plasmid origins and on mini-chromosomes (Eliasson et al., 1996; Eliasson and Nordström 1997).

Other attempts to estimate the eclipse period utilised synchronised cells by temperature-sensitive cell-cycle mutants, dnaC2 and dnaA46. In these experiments initiation of replication was stalled by shifting an exponentially-growing population to non-permissive temperature; when the cells were shifted back to permissive temperature, initiation of replication in all cells was synchronised at initiation of chromosome replication. The rate of DNA synthesis after the shift back to permissive temperature was followed by [3H]-thymidine pulse incorporation for 3 minutes at different time samples. The relative DNA synthesis is a measure of active replication forks, this activity is supposed to increase stepwise with a two-fold increase each cell cycle. The length of each step is a measure of the eclipse (Bogan and Helmstetter, 1997). The other strategy was to follow the increase of oriC copies per cell by flow cytometry. In this experiment, cells were grown at non-permissive temperature for 90 minutes before shift back to permissive temperature. The number of oriC copies per cell was measured at different times after the shift back to permissive temperature. The number of oriC was measured by flowcytometry analysis of rif-runout population. The numbers of origins were expected to increase stepwise with an increase of the power of 2 at each step. The length of the step is an estimate of the eclipse period (von Freiesleben et al., 2000). Meselson-Stahl Density shift combined with Southern blot. (general description) We have combined two frequently used techniques and analysed and compared the results by theoretical simulations (Olsson et al., 2002; papers I, II, III and IV). The method is as in previous Meselson-Stahl density-shifts based on the semiconservative replication of DNA (Meselson and Stahl, 1958). The cells were first grown at least 10 generations in heavy medium with heavy isotopes (13C6-glucose; 15NH4

+). At an OD550≈0.2 the culture was diluted five-fold by light medium (12C6-glucose; 14NH4

+).

Page 29: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 27

DNA was extracted at different times after the shift. DNA fragments of different density were separated in a CsCl-gradient by equilibrium ultra-centrifugation. The gradients were fractionated on microtiter plates. The DNA contents in each fraction were quantified by dot-blot hybridisation; gene specific probes were used. The range of DNA contents in our samples was shown to be proportional to the signal in the phosporimager (Paper I: Olsson et al., 2002). The kinetics of the DNA replication and distribution of Heavy-Heavy-dsDNA (HH), Heavy-Light-dsDNA (HL) and Light-Light-dsDNA (LL) were plotted and compared with theoretical computer simulations (see the next paragraph). Each gradient fractionated on a microtiter plate can be blotted on 3-4 membranes. The membranes can be stripped and repeatedly rehybridised with different probes. This allowed us to study the replication kinetics of different segments of the same DNA molecule or separate DNA molecules in the same cells from the same density-shift experiment. Theoretical analysis of DNA replication with different eclipse periods The DNA before a density shift contain two heavy strands (HH), after one round of replication in light medium gives two hybrid DNA molecules (HL). Further replication of HL-DNA results in one hybrid-DNA (HL) and one light DNA molecule (LL). Thus after a density shift will three reactions take place: HH-DNA+ L-deoxyribonucleotides= 2 HL-DNA HL-DNA+ L-deoxyribonucleotides= HL-DNA + LL-DNA LL-DNA+ L-deoxyribonucleotides= LL-DNA + LL-DNA

We assumed that all three reactions take place with the same rate constant, k. Another assumption was that the whole population grows with the same generation time. The kinetic behaviour of the replication and the subpopulations of HH-DNA, HL-DNA and LL-DNA was theoretically analysed (Olsson et al., 2002). The simulations are robust and fits any distribution of inter replication times.

In the ideal case when all DNA molecules replicate with intervals of exactly 1 generation time, all HH-DNA will replicate to hybrid DNA (HL) before any LL-DNA will appear; hence, the HL-fraction will be 100% of the DNA before HL-DNA will replicate and result in an increasing fraction of LL-DNA. However if the eclipse period (∆t) is shorter then one generation time, some DNA-molecules will replicate twice, while others will not replicate at all. This scenario will result in HH, HL and LL-fractions simultaneous after one generation time (See figure 12.). However since one assumption was that the DNA population was homogenous the HL-fraction cannot be less than 50%. The maximum value of HL-DNA (HLmax) is always reached after one generation time. Depending on the length of the eclipse period ∆t, the HLmax value will vary in the interval 50-100% according to the following equation (Olsson et al., 2002).

∆t = 2 HLmax -1 (3) The eclipse period is defined as the first appearance of LL-DNA but since the LL-DNA appears asymptotically is it far more accurate to use the total kinetics of the DNA population. We used the HLmax value and compared it to the computer simulations in figure 12. to estimate the eclipse period.

Page 30: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

28 Jan Olsson

Properties and advantage of the method The measurement of the eclipse period can be applied on exponentially-growing cells without drug-induced or mutation-aided disruption of the cell cycle. The method does not require synchronisation or temperature shifts. The method is therefore useful if one wants to study how a single mutation affects the eclipse period without side effects caused by the method, e.g. synchronisation by drugs. The replication kinetics of specific regions on the chromosome can be studied, with the use of specific probes at the hybridisation step. We could show that the value of the eclipse period was identical when we used probes close to the oriC-region (dnaA-probe at 84 min), the terC-region (tus-probe at 36 min) and at an intermediate position (dnaC-probe at 99 min). The eclipse periods measured in the wild-type strain MG1655 with the terC-probe reflects initiations of replications before the cells were shifted to light medium while the eclipse period obtained with the oriC-probe reflects initiations of replications after the density shift. The eclipse period was 0.60 with all three probes. This indicates that the eclipse period is not different before and after the density shift. The growth rate was measured by optical density and did not change after the density shift to light medium. We could choose probes specific for an R1 plasmid and the chromosome and hybridise on two separate membranes with DNA from the same CsCl-gradient. In conclusion: The method can measure the eclipse period in exponentially-growing cells, that therefore are unperturbed in their growth and replication. The method does not change the growth rate. The eclipse period can be studied for any region on the chromosome and for any other replicating DNA molecule.. Probes close to the oriC reflect initiations after the density shift while probes close to the terC reflects initiations before the shift. The eclipse period can be measured for the chromosome and coexisting plasmids in cells from the same density shift experiment.

Fig. 12. Computer simulation of the replication kinetics during a density shift experiment. The HH-DNA drops and HL-DNA increase directly after the shift to light medium. The LL-DNA increases gradually depending on the length of the eclipse period. The very thick lines correspond to the situation with an eclipse of exactly one generation time. The lines of medium thickness correspond to replication with no eclipse. The thin lines correspond to intermediate length of the eclipse period in steps of 0.1 generations. The eclipse period is easily estimated from value of the HLmax. Data from a typical experiment are plotted on top of the simulated curves.

Page 31: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 29

CONTROL OF CHROMOSOME REPLICATION

(papers I and II)

Our focus is on the initiation of the chromosome Replication of the E. coli chromosome is under strict cell-cycle control. This statement is supported by many results: (i) the initiation mass is relatively constant over a wide range of growth rates (Donachie, 1968; Boye et al., 1996), (ii) Initiation of replication from a unique locus oriC, takes place once and only once every cell cycle (Boye et al., 2000),and (iii) synchrony of initiation from all oriCs in a cell (even for multicopy oriC-plasmids) during exponential growth (Skarstad et al., 1986; Leonard and Helmstetter, 1986). There are many requirements to be fulfilled to allow an initiation (a) the initiator protein DnaA has to be available in required concentration and in the active ATP-DnaA form, (Messer and Weigel, 1996; Bogan and Helmstetter 1997; Katayama, 2001a) (b) the Dam-methylase has remethylated the hemimethylated GATC-sites in oriC (Russel and Zinder, 1987; Cambell and Kleckner, 1990; Lu et al., 1994; von Freiesleben et al., 2000).

We have in paper I and II tried to understand some already well studied and some less investigated control systems of the initiation of chromosome replication and the interplay between them. We have used numerous mutants that we knew or believed could disturb the control of chromosome replication resulting in asynchronous replication. Meselson-Stahl density-shift experiments combined with southern hybridisation was developed to measure the minimal inter-replication time, the eclipse period. Cell size and DNA content were measured by flow cytometry. We analysed rif-runout cells with the same method to investigate the loss of synchrony.

We wanted to investigate the nature of asynchrony, eclipse period, and try to understand the correlation between asynchrony and reduced eclipse period. The eclipse period in cells growing exponentially phase has never been studied for specific segments on the chromosome before these studies.

Eclipse period at different growth rates Replication in wild-type E. coli is under strict control. The inter-replication time is known to be narrowly centred around one generation time (Koppes and Nordström, 1986). The C- and the D-periods are quite constant (40 min and 20 min, respectively, at 37°C) at different growth rates (Cooper and Helmstetter, 1968). We wanted to investigate whether the length of the eclipse period is fixed at different growth rates or if it is a constant fraction of the generation time.

We determined the length of the eclipse period for the wild-type strain MG1655 in two different media and at two different temperatures. As shown in Table I, that the eclipse period was a constant fraction (0.6) of the generation time and not a fixed physical time like the C- and D-periods. This may be important when different mutants have different growth rates in the same medium. Wild-type strains of E. coli are known to have synchronous initiation of replication. Hence, synchrony is possible with an eclipse of 0.6 generation times.

Page 32: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

30 Jan Olsson

Table I. The length of the eclipse period in the wild type E. coli K-12 strain MG1655

Minimal medium supplemented witha

Temp. (°C)

Generation time τ (min)b

Eclipse (Generation time)

Eclipse (min)

Glucose 30 137 0.60 82 Glucose 37 76 0.60 46

Glucose and CAA 37 40 0.60 24 a glucose (6 mM) and casamino acids (CAA) 0,1% (w/vol) b estimated from density shift data when the size of the HH-fraction = LL-fraction Eclipse period in the absence of sequestration The sequestration machinery is crucial in controlling the timing of synchronous initiation of replication. The SeqA protein and the Dam methyl-transferase in concert with the GATC-sites in the oriC-region are components in the sequestration process (see The sequestration control; page 17). It is known that the GATC-sites are remethylated much faster in seqA-mutants than in the wild-types. Flow-cytometry analyses by rif-runout experiments with seqA- and dam-mutants have shown asynchronous initiation of replication (Boye and Løbner-Olesen, 1990; Lu et al., 1994). Asynchronous initiation of replication indicates that initiation of the individual origins within one cell is scattered during the cell cycle. Asynchronous initiation does not mean per se that the eclipse period is short. An eclipse period of 0.6 generations for each individual origin is not in conflict with asynchrony.

Eclipse periods have been studied previously in seqA and dam mutants. But those experiments were done on synchronised cells in a parent strain with the dna46 mutation. The authors found that the eclipse period disappeared in the seqA mutant. They could extend or reduce the eclipse period if the dam gene was overexpressed or underexpressed, respectively. They measured the eclipse period indirectly by flow cytometry and in cells with the dnaA46 mutation during non-exponential growth conditions (von Freiesleben et al., 2000)

We wanted to measure the eclipse period of the chromosome in cells during unperturbed exponential growth with a direct method. We measured the eclipse period in seqA, dam mutants, and in the double mutant seqAdam. We used two different seqA mutants, inactivated by insertion transposon Tn10 (seqA::Tn10) and one in frame deletion (seqA∆10). The seqA gene belongs to the same operon as the gene pgm encoding the enzyme phosphoglucomutase (Lu and Kleckner, 1994). The insertion of transposon Tn10 in the seqA gene (seqA::Tn10) inactivates the down stream pgm-gene. The in frame deleted seqA mutant has a normal pgm-expression. All mutants were variants of MG1655. We wanted to answer these following questions: • How is the eclipse period changed in the mutants relative to that of the wild type

strain? • Since SeqA is known to have additional roles than in sequestration, will the

eclipse period in the seqA mutant and the dam mutant be different? • Is sequestration of hemimethylated oriC the only contributor to the eclipse or

does SeqA have additional roles that influence timing of initiation? • Does the absence of pgm activity affect the eclipse period in the seqA mutant

(seqA::Tn10)?

Page 33: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 31

• Is it possible to see additive effect on the eclipse period in the double mutant seqAdam?

The eclipse period was significantly reduced in all strains relative to the wild-type strain (Table II). This is the first time the inter-replication time has been measured in these mutants during unperturbed exponential growth.

The eclipse period in the seqA mutant (seqA::Tn10) was reduced to 0.14 generations while the eclipse period of the dam mutant (dam::KmR) was reduced to 0.40 generations. The difference between the seqA mutants and the dam mutant was significant. The difference between the length of the seqA mutant (seqA::Tn10) and the dam mutant (dam::KmR) was surprising to us. We know that SeqA might have additional roles in the control of the cell cycle. However our seqA mutant (seqA::Tn10) has a disrupted phosphoglucomutase (pgm) activity because the pgm gene is located down stream of the seqA gene at the same operon (Lu and Kleckner, 1994).

Strains lacking pgm-activity exhibit poor growth on minimal medium supplemented with galactose as the carbon source. It is proposed that pgm mutants are defect in cell-wall structure after reports showing that pgm mutants are more sensitive to phage C21and sensitive to detergent (Lu and Kleckner, 1994).

To investigate whether the disrupted Pgm-activity can explain the differences between the eclipse periods in our seqA-mutant (seqA::Tn10) and the dam-mutant., we measured the eclipse period in a seqA-mutant (seqA∆10) with an in-frame deletion, so the Pgm activity is not affected.

Both seqA mutants (pgm+ and pgm-) reduced the eclipse period to ca 0.14 ±0.02 generations. Hence, we could not see any effect from the disrupted pgm-expression in the seqA::Tn10 strain.

The picture of the control of the initiation of replication is complex. If the effect from the seqA and the dam mutations were additive, we would see an even shorter eclipse period in the double mutant damseqA than in the seqA mutant. The double mutant seqAdam got an intermediate reduction of the eclipse period to 0.32 generations, Hence the effect was not additive. This is surprising to me. What can explain this intermediate length of the eclipse period? I would like to attack this strange result from two points of view.

The eclipse period of the seqA mutant (0.14) was shorter than the eclipse period in the double mutant seqAdam (0.32). The sequestration control (see chapter The sequestration control page 17) is bypassed both in a seqA mutant and in a dam mutant. However, the seqA mutant has to be remethylated before the initiator protein DnaA binds to oriC, while the chromosome is completely unmethylated in the dam mutant. A difference in binding affinity for DnaA to methylated and unmethylated oriC might explain why the eclipse period is longer in the double mutant seqAdam than in the seqA mutant.

The eclipse period of the dam mutant (0.40) was longer than the eclipse period in the double mutant seqAdam (0.32). Results presented later in the thesis show that the increased degree of negative superhelicity reduces the eclipse period (paper II). Our lab has reported an increased nucleoid condensation and increased negative superhelicity in seqA-mutant compared to the wild type strain (Weitao et al., 1999; 2000). Both the double mutant seqAdam and the dam mutant have bypassed sequestration control and both have unmethylated oriC. We don not know if the superhelicity is similar in the seqA mutant and in the double mutant seqAdam. The

Page 34: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

32 Jan Olsson

shorter eclipse in the double mutant relative the eclipse period in the dam mutant could reflect different degrees of negative superhelicity.

We used different probes close to oriC, terC and at 99 minutes (DnaC). We could not see any significant differences depending on which probe we used, hence, the treatment during the density shift did neither effect the eclipse period in the mutants nor in the wild-type strain, (see table II). The generation time was not affected by the mutations. All strains were grown at 37°C in minimal medium supplemented with 6 mM glucose. The eclipse period was not completely abolished in any of the mutants. Table II. The eclipse period in mutants without intact sequestration. Strain Generation

time τ (min)a DNA probe Eclipse

(Generation time) dam::KmR 75 dnaA 0.40 tus 0.40 seqA::Tn10 pgm- 77 dnaA 0.15 dnaC 0.14 tus 0.14 seqA∆10 pgm+ 80 dnaA 0.12 dnaC 0.12 tus 0.16 dam::KmR seqA::Tn10 pgm- 80 dnaA 0.32 dnaC 0.32

a estimated from density shift data when the size of the HH-fraction=LL-fraction In conclusion Mutations in the sequestration machinery reduced the eclipse period significantly. The absence of pgm activity in the seqA mutant (seqA::Tn10) did not affect the eclipse between the two seqA mutants. The eclipse was most reduced in the seqA (0.14), was 0.40 in the dam mutant, and 0.32 in the double mutant seqAdam. The density shift did not have any effect on the eclipse periods, since the length of the eclipse period was similar were probes close to oriC and close to terC were used. The double mutant had an intermediate eclipse period between the short one in seqA-mutant and the less reduced dam-mutant. The eclipse was not reduced completely in any of the mutants. DnaA(ts) and SeqA have additive effect on the eclipse period The initiator protein DnaA binds to DnaA-boxes in oriC with remethylated GATC-sites. ATP- and ADP-DnaA binds to DnaA-boxes. DnaA without ATP or ADP cannot bind to DnaA-boxes. ATP-DnaA is required to break up the strands in the AT-rich region of oriC (see chapter DnaA a protein with many functions, page 15 and figure 6, page 14)

It is known that many temperature-sensitive dnaA mutants replicate asynchronously at permissive temperature (Skarstad et al., 1986; Skarstad et al., 1988). One of these temperature sensitive dnaA mutants is dnaA46. The DnaA46 protein lacks the ability to bind ATP and ADP also at permissive temperature (Hwang and Kanugi, 1988). The aim of our experiments was to study the eclipse period in the dnaA46 strain. The mutation was introduced into the wild type strain MG1655. We wanted to see if the sequestration machinery was the sole control system required to guarantee a narrow distribution of the inter-replication time centred around one

Page 35: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 33

generation time. The reducing effects in the seqA and dam mutants were not additive. We wanted to investigate the effects on the eclipse period in the double mutant dnaA46seqA strain.

The eclipse period was reduced to 0.30 generation times in the dnaA46 strain at the permissive temperature 30°C. The GATC-sites in oriC remains hemimethylated during 1/3 of a generation time (Campbell and Kleckner, 1990). The length of the eclipse period in the dnaA46 mutant is almost identical with the time period in which the oriC remains hemimethylated due to the sequestration control. The reduced eclipse period in the dnaA46 mutant demonstrate that other control systems besides the sequestration machinery are required to maintain the long eclipse period in the wild-type strain. A new DnaA-binding region 10.7 minutes clockwise of oriC was found recently, called the datA-locus. The datA-locus contains DnaA-boxes (Kitagawa et al., 1996; 1998). It is not known if the datA-locus contains ATP-DnaA-boxes. The initiator-titration model suggests that upon replication of the datA-locus, all available DnaA molecules titrates to the duplicated datA-locus instead of being used to reinitiate replication from oriC (Hansen et al., 1991). It was reported that the DnaA46 protein cannot bind neither ADP or ATP at permissive temperature (Hwang and Kanugi., 1988). Only the ADP- or ATP-bound forms of DnaA can bind to the DnaA-boxes. Hence DnaA-titration is lost in the dnaA46 mutant (Hansen et al., 1991;Kitagawa et al., 1996,1998). The reduced eclipse period in the dnaA46 mutant may reflect a loss of DnaA-titration.

The RIDA (Regulatory Inhibition of DnaA) is based on the regulation of the hydrolysis of the ATP bound to DnaA by linking it to the β-subunit "sliding clamp" of the pol III holoenzyme (DnaN) (Katayama et al., 1998). This inactivation of ATP-DnaA is believed to contribute to the timely initiation of chromosomal replication (Katayama, 2001a). The RIDA system is probably bypassed in the dnaA46 mutant since the DnaA46 protein lacks affinity for ATP (Hwang and Kanugi., 1988).

ATP-DnaA is required to open up the AT-rich region at oriC to initiate replication (Speck et al., 1999). It seems as dnaA46 mutants can initiate chromosome replication. I do not know whether the mutated DnaA46 protein is involved in the initiation process and whether the replication is initiated at oriC. Marker-frequency analysis could give an answer to the last question. Maybe the DnaA46 protein can bind ATP but less efficiently and therefore initiate replication (after correspondence with Prof. Katayama). It is not obvious that the dna46 mutant is under sequestrational control. It is therefore important to measure the eclipse period also in the double mutant dnaA46seqA.

The eclipse period was further reduced to 0.10 generation time in the double mutant dnaA46seqA strain. It seems as if the dnaA46 and the seqA mutation have mild additive effects on the length of the eclipse period in the double mutant dnaA46seqA. The shorter eclipse period in the double mutant dnaA46seqA relative to the dnaA46 mutant shows that replication in the dnaA46 mutant is under sequestrational control. The length of the double mutant dnaA46seqA (0.10) is shorter than the eclipse period in the seqA mutant (0.14). The eclipse in the seqA mutant is probably longer, since the RIDA system is intact. In conclusion: We could show a reduced eclipse period in a strains with intact sequestration machinery. The length of the reduced eclipse period in the dnaA46 mutant (0.3) is the same as the length of the time oriC is sequestered from remethylation by the sequestration control.

Page 36: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

34 Jan Olsson

Since the DnaA46 proteins cannot bind ATP probably both the RIDA system and the DnaA-titration controlled system are bypassed.

The eclipse in the seqA mutant (0.14) is probably longer than the eclipse in the double mutant dnaA46seqA (0.10) because the seqA mutant has an intact RIDA system. Superhelicity and the eclipse period The seqA mutant had a shorter eclipse period (0.14) than the dam mutant (0.40). Our laboratory has shown that the seqA mutant has a more condensed nucleoid due to increased negative superhelicity than the wild type strain (Weitao et al., 1999 and 2000). Could the difference in superhelicity cause further reduction of the eclipse period in the seqA mutant relative to the dam mutant? To isolate the effect due to increased negative superhelicity we used a double mutant gyrB225topA which is known to have an increased negative superhelicity and its parent strain JJC82 (Sternglanz et al., 1982; DiNardo et al., 1982).

The cells were grown in minimal medium supplemented with 6 mM Glucose at 37°C. The eclipse period was 0.75 in the parent strain JJC82 and the double mutant gyrB225topA had an eclipse period of 0.49.

The eclipse period in the parent strain was longer in our wild type strain JJC82 than in the previous wild type strain MG1655, 0.75 and 0.60 generations respectively. The eclipse period in the double mutant gyrB225topA was reduced 35 % to 0.49 generations. The eclipse period was reduced in the presence of intact sequestration machinery. Asynchrony index a rough parameter Cells growing in very poor medium, e.g. minimal medium supplemented with acetate as carbon source, cannot have a high asynchrony index, since the average number of oriC is very low. We wanted to investigate the correlation between the asynchrony index and number of oriC. We plotted our calculated asynchrony indices against the average number of oriC in the cell. We saw two distinct clusters, corresponding asynchronous cells and synchronous cells. Both clusters were extended and the asynchrony index was proportional to the number of origins. The slopes of the two clusters were clearly different, the cluster corresponding to the wild-type strains had lower asynchrony indices and the increase was much less than that of the asynchronous clusters (see table II in paper II; see figure 13.A.). The plot has trend-lines for the two clusters. We show that the asynchrony index (A) is a quite difficult parameter. It is very important to calculate the average number of oriC to see whether different asynchrony index is due to different degree of asynchrony or due to different number of oriC.

We wanted to see if this relation between asynchrony index and number of oriC was consistent with previous experiments. We therefore calculated the average number of oriC and the asynchrony index from data in two previous papers (Eliasson et al., 1996; Skarstad et al., 1986). The asynchronous strains were a recA mutant, a dnaA46 mutant, intP1, and intF-strains, and the synchronous strains were four different wild-type strains, both K12 and B-strains. The asynchrony indices did coincide with our two clusters (see table II in paper II; see figure 13.A.).

Previous report showed that dnaA-mutants have different asynchrony index (A=(f3)/(f2+f4)) depending on the position of the mutation (Skarstad et al., 1988). We wanted to see if those data depended on different number of oriC or if was due to different levels of asynchrony so we recalculated the asynchrony index from the

Page 37: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 35

published data: A=(f3+f5+f6+f7)/(f2+f4+f8) (Table III). We could see that some of the mutants coincided very well with our cluster while other dnaA-mutants are intermediate (figure 13.B). I will not discuss the locations of the different dnaA mutants in relation to their degree of asynchrony. I just wanted to show that strains with intermediate levels of asynchrony do exist.

It is possible to increase the average number of origins by changing the growth rate. Is it possible to use the slopes of the calculated trend lines of the two clusters to predict a new asynchrony index? If one already know the asynchrony index and average number of origins in one condition, e.g. minimal medium with glucose as carbon source, is it possible to estimate the new asynchrony index in another growth condition just from the conditions in the previous growth conditions, the slope of the trend line and the average number of oriC in the new growth condition? We studied the increase in asynchrony indices and average number of origins for the strains grown in minimal glucose medium with and without supply of casamino acids. We saw that the increase of the asynchrony indices was not parallel with the trend lines (see figure 13C.). We cannot find any obvious explanation of this puzzling result.

We studied also whether an increased degree of negative superhelicity caused asynchrony. We studied one topA mutant and the double mutant gyrB225topA. The Topoisomerase I (encoded from the topA gene) decrease the negative superhelicity. The topA mutant is believed to have an increased negative superhelicity similar the superhelicity of the seqA mutant. The topA mutant did hardly show any loss in synchrony. The change in superhelicity has not been investigated in detail. It is possible that negative supercoiling is less in the topA mutant than in the seqA mutant due to homeostatic compensation by the gyr gene (Mentzel and Gellert, 1983). The double mutant gyrB225topA is known to have an increased level of negative superhelicity (Sternglanz et al., 1981; Di Nardo et al., 1982). The double mutant did have a distinct level of asynchrony (see table II in paper II). This is not surprising since the double mutant did also have an reduced eclipse period. Table III. Asynchrony indices and average number of origins/cell recalculated from published flow-cytometer data on dnaA-mutants grown in LB medium at 29°C (Skarstad et al., 1988). Strain Origins/cell A=(f3)/(f2+f4) A=(f3+f5+f6+f7)/(f2+f4+f8) CM735 3,66 0,10 0,08 CM735 4,33 0,11 0,13 CM742 dnaA46 4,38 1,25 1,21 CM742 dnaA46 3,89 1,32 0,93 CM740 dnaA5 4,04 1,24 1,09 CM746 dnaA204 3,70 0,25 0,22 CM748 dnaA203 4,03 0,26 0,29 CM744 dnaA205 3,15 0,71 0,39 CM2735 dnaA601 4,01 1,29 0,98 CM2733 dnaA602 4,04 1,25 1 CM2738 dnaA604 4,37 1,33 1,03 CM2555 dnaA508 3,53 0,27 0,24 CM2556 dnaA167 4,17 0,38 0,38

Page 38: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

36 Jan Olsson

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,5 2 2,5 3 3,5 4 4,5 5

M9 glucose M9 glucose CAA

average number of oriC

Asy

nchr

ony

inde

x

medium and asynchrony

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,5 2 2,5 3 3,5 4 4,5 5

205

203

167

46

46

46

46

5

601602

604

508204

average number of oriC

Asy

nchr

ony

inde

x

dnaA mutants

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,5 2 2,5 3 3,5 4 4,5 5

Asy

nchr

ony

inde

x Aour data Eliasson et al . 1996Skarstad et al . 1986

average number of oriC

two clusters

Fig 13. (A) Correlation between asynchrony index and average number of oriC. The data were collected from flow cytometry from our own experiments (♦) and from two separate published papers; the latter coincide with our clusters (Eliasson et al., 1996 (o); Skarstad et al., 1986 (*)). Trend lines of the asynchronous (thick line) and the synchronous cluster (thin line) are drawn in the plots. Fig. 13. (B) Asynchrony index as a function of the number of oriC. The asynchrony index and average number of oriC were re-calculated from published data of dnaA mutants (Table III in this paper; Skarstad et al., 1988). The data from some of the mutants (◊) do not coincide with the defined clusters. Thenumbers of the dnaA mutants are labelled in the plot. The trend lines are the same as in figure 13.A. Fig. 13. (C) Asynchrony index of a number of strains, grown at two media, as a function of the number of oriC. The relative increase in asynchrony index is not parallel with the trend lines. The cells were grown in glucose-minimal medium with (♦) and without casamino acids (◊). Each arrow indicates the data for one strain in the two media.

Page 39: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 37

Correlation between eclipse and asynchrony The eclipse period was reduced by different degree in different mutants. It seems that many causes to asynchrony also cause a reduced length of the eclipse period. We wanted to investigate whether there is any correlation between the length of the eclipse period and the degree of asynchrony. We calculated the asynchrony index (A=(f3+f5+f6+f7)/(f2+f4+f8)) from flow cytometry of rif-runout treated cells (Boye and Løbner-Olesen, 1990). We studied cells in exponential phase and cells after rif-runout. The cells were harvested in exponential phase at an OD550=0.1. The rifampicin and the cephalexin were also added at the same optical density. The rif-runout was incubated another 2 hours. We calculated the average number of oriC per cell from the rif-runout experiment. (4) n is the highest numbers of oriC among the cells in the sample, fi is the fraction with i chromosome equivalents.

The flow-cytometry analysis was performed during one specific calibration of the flow cytometer instrument for each growth condition. It is important to avoid fluctuations between separate calibrations of the instrument especially when we want to calculate the average cell mass and DNA-content of the cells grown in exponential phase. The strains were grown in glucose-minimal medium with or without a supplement of casamino acids.

The results are summarised in tables I and III in paper II. The average number of oriC varied among the strains and between the growth conditions. All strains were grown at 37°C except for the strains with the dnaA46 mutation; they were grown at 30°C.

We plotted the asynchrony indices against the corresponding eclipse period. We could suspect that the cells fail to replicate synchronously at a threshold value under 0.5 generations of the eclipse period. This threshold became even more obvious when the asynchrony indices were normalised to the average number of origins (see fig. 14.).

As we could see the parameter asynchrony index (A) is a quite rough measure of the level of asynchrony. The asynchrony index increased with increased number of oriC when asynchronous strains were grown in different rich media. It was therefore reasonable for us to compensate for this effect when we wanted to investigate the relation between asynchrony and the eclipse period. The asynchrony does not seem to be a smooth function of the length of the eclipse period. It is rather a step like function. The shortest eclipse period with asynchronous initiation of replication was 0.5 generations. And the step between synchrony and asynchrony seems to be very abrupt. However our eclipse periods with the length of 0.49 generations is actually a reduction from an eclipse period of 0.75 generations in the parent wild-type strain JJC82. The wild-type strain MG1655 had an eclipse of 0.6 generations. The dam mutant in the MG1655 strain had an eclipse of 0.4, hence, a reduction of 33%. So we can say that an eclipse reduction of ca 35% is enough to obtain asynchrony.

origins/cell=

Σi=1

n

i⋅ fi

Σi=1

fi n

Page 40: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

38 Jan Olsson

0

0,2

0,4

0,6

0,8

1

1,2

1,4

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,80

0,05

0,1

0,15

0,2

0,25

0,3

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8

Eclipse period Eclipse period

Asy

nchr

ony

inde

x (A

)

Asy

nchr

ony

inde

x / n

umbe

r oriC

Fig. 14. (A) Correlation between asynchrony index A: and the length of the eclipse period. A=(f3+f5+f6+f7)/(f2+f4+f8). (B) Correlation between the ratio asynchrony index/number oriC and the length of the eclipse period

A B

Page 41: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 39

THE NATURE OF PLASMID R1 REPLICATION (PAPERS III AND IV)

Plasmids are autonomous replicating DNA molecules. Plasmids can be either circular or linear molecules with a size range from 1.5 kbp-Mbp. E. coli can harbour many different kind of plasmids. The plasmids are classified in incompatibility groups. Plasmids with the same copy-number-control (CNC) system cannot coexist in the same cell, they are incompatible. Plasmid R1 belongs to incompatibility group IncFII. Plasmid R1 is a low-copy plasmid with about 3 copies per newly born cell (Nordström et al., 1980). The size of plasmid R1 is about 90 kb. We have focused on the kinetics of the R1 basic replicon. We have therefore used different mini-plasmids with a basic replicon and one selection marker. We wanted to do in vivo experiments with the aim to see if theoretical models of the CNC system are relevant.

We have used Meselson-Stahl density-shifts experiments with R1-specific probes (Olsson et al., 2002) to follow plasmid replication in vivo. The function of the CNC system is to increase or decrease the replication rate of plasmid population within the cell depends whether the copy number is to low or to high, respectively. We needed to be able to study the plasmid in steady-state replication as well as in up shift and down shift of the copy number. We used an R1 plasmid with a temperature-regulated copy number. Theoretical prediction on the Multistep control system The minimal inter-replication time was measured on an intR1-strain with Meselson-Stahl density shift experiments. The eclipse period turned out to be ca 0.2 generation times (Koppes and Nordström, 1986). Ehrenberg and Sverredal 1995 first described the Multistep control model. The Multistep model was shown to be true for the ColE1-plasmid (Paulsson et al., 1998; Paulsson and Ehrenberg, 2000). R1 replication requires de novo synthesis of many copies of the replication protein RepA. RepA has been shown to act in cis, meaning that the RepA proteins bind to the origin of replication oriR1 on the plasmid from which the RepA mRNA was transcribed. The RepA synthesis is a rate-limiting step in replication. The RepA synthesis is inhibited by CopA, an antisense RNA to the CopT region of the RepA mRNA. The reaction scheme describing the initiation of a single plasmid copy is shown in figure 10 (Ehrenberg and Sverredal, 1995). They could from this scheme simulate the eclipse period obtained by Koppes and Nordström 1986.

The simulation predicts 2-3 steps indicating that each RepA mRNA translates a burst of RepA proteins. If the Multistep model is true number of steps should increase if the inhibition frequency by CopA was increased. The inhibition frequency increases if the CopA concentration is higher than normal, e.g. during a downshift in copy number. An increased number of steps predicts an extended eclipse period. Conversely the number of step would decrease if the CopA concentration is low e.g. during an upshift. We used upshifts and downshifts to investigate whether the Multistep model is true for the CNC system of plasmid R1.

Page 42: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

40 Jan Olsson

R1 plasmid with a temperature-dependent copy number We wanted to investigate the relevance of the Multistep-control model for plasmid R1 studying the eclipse period of the plasmid in transient states at downshift and upshift. The eclipse period should be shorter during an upshift and consequently longer during a downshift. We used a mini R1-plasmid called pOU71, which is a temperature-induced runaway-replicating plasmid at temperatures >39°C (Larsen et al., 1984). The RepA mRNA synthesis can be increased by a cloned lambda promotor λPR upstream of PcopB. The lambda promotor is much stronger than the down stream promotors. The lambda promotor is repressed by a repressor encoded from cI857 at temperature ≤34°C. (see figure 15A) The copy number of pOU71 at 30°C is equal to the wild-type plasmid.

Figure 15B describes the situation during an upshift; The temperature sensitive repressor cannot repress the strong lambda promotor λPR. The increased transcription of RepA mRNA reduces the transcription of the antisense RNA CopA by convergent transcription (Stougaard et al., 1982). At a temperature interval (ca 34-39°C) will the CNC regulate the plasmid replication but the copy number at steady state will be higher than at 30°C. When the temperature is high (>39°C), the CopA expression will drop a lot due to convergent transcription and, hence, fail to catch up the increased RepA mRNA synthesis, the replication are therefore completely uncontrolled, denoted runaway replication (see the large dark grey square in figure 9).

When cells with increased copy number due to an upshift to high temperature, are shifted back to 30°C, the elevated copy number will drop to normal copy number. The kinetics of the drop will probably follow the +n-mode. This situation reflects a downshift situation. During a downshift experiment the relative CopA concentration is higher than at steady state in the same temperature. The RepA synthesis drops because most RepA mRNA will be inactivated very fast due to the high concentration of CopA (see figure 15C).

Eclipse period during Upshift (paper III) The Multistep model predicts an abolished eclipse during an upshift in copy number (Ehrenberg and Sverredal, 1995; Paulsson and Ehrenberg, 2000). We used plasmid pOU71 a mini-R1-plasmid with runaway phenotype at high temperatures, to test this prediction (Larsen et al., 1984).

The speed of the inactivation of the repressor cI857 due to temperature shift was not clear for us. We did not know if the folding of the repressor was temperature sensitive or if the 3D-structure itself was sensitive to high temperatures. If only the folding is temperature sensitive the copy number increases will start when growth has

Fig. 15. The basic replicon during normal copy number (A), increasing copy number during an Upshift (B), and decreasing copy number during downshift (C). Plasmid pOU71 is an R1-plasmid with the lambda promotor (λPR) in front of the basic replicon. The lambda promotor is repressed at temperatures <34°C by the repressor encoded from cI857 (A). After a shift to temperatures >39°C the much stronger promotor (λPR) is derepressed. The increased transcription of RepA-mRNA inhibits transcription of the antisense-RNA CopA, so called convergent transcription. This results in an increased copy number (B). A downshift experiment can be obtained if the cells are shifted back to 30°C. The strong lambda promotor (λPR) is repressed, The concentration of CopA is very high due to high gene dosage. The high concentration of CopA efficiently inhibits translation of the normally expressed RepA-mRNA and, hence further replication is reduced(C).

promotor

transcription/translation

inhibition

RNA

Page 43: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 41

PrepA PcopB

PcopA copA copB tap

copT

oriR1 RepA

CopA

Low CopA-expression due to convergent transcription

cis-actingRepA

CopB- tetramer

RepA-mRNA

CopB & RepA mRNA

λPR

cI857

PrepA PcopB

PcopA copA copB tap

copT

oriR1 RepA

CopA

trans-acting CopA

cis-actingRepA

CopB- tetramer

RepA-mRNA

CopB & RepA mRNA

λPR

cI857

PrepAPcopB

PcopA copA copB tap

copT

oriR1 RepA

CopA

High concentration of CopA due to high gene dosage

cis-actingRepA

CopB- tetramer

RepA-mRNA

CopB & RepA mRNA

λPR

cI857

Normal copy number at 30°°°°C

Increasing copy number at Up-shift 30°°°°C⇒⇒⇒⇒42°°°°C

Decreasing copy number at Down-shift 42°°°°C⇒⇒⇒⇒30°°°°C

15A.

15B.

15C.

Page 44: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

42 Jan Olsson

diluted the repressor. We did not know the behaviour of the kinetics of the release from the lambda-repressor cI857. To be assured that the upshift had started, the eclipse period was measured 30 minutes after the temperature shift (see figure 15).

The cells were grown at least 10 generation in the exponential phase in heavy-minimal medium. At OD550≈0.1 the cells were shifted to the new higher temperature (37, 38, or 42°C). The cells were grown another 30 minutes before the cells were shifted to pre-warmed light medium by dilution. Time samples were taken. The rest of the experiment followed the procedure presented in the chapter describing Meselson-Stahl density-shift combined with Southern blot.

We made two control experiments. The first the culture was kept at 30°C. In the second control experiment, we used the mini-R1 plasmid pGW71, which is identical to the runaway plasmid pOU71 except that pGW71 lacks the lambda-promotor and repressor-gene. We used this to measure the eclipse period in this control plasmid at exponential growth at 42°C.

The optical density was measured during the experiment. The growth rate seemed to change immediately after the temperature shift. The addition of light medium at the density shift did not change the growth rate. The generation time of the cells was shorter at 37, 38 and 42°C than at 30°C; however the generation time was longer at 42°C (90 min) than at 38°C (80 min). I think plasmid replication during runaway replication recruits so much energy, metabolites and replication-associated proteins in competition with the needs of the host at 42°C. The generation time for our control strain (MG1655 - pGW71) at 42°C was shorter (65 min) (see figure 17A).

We used a chromosome-specific probe close to the terC-site (tus) and an R1-specific probe targeted to the basic replicon. Cross hybridisations were not detectable (data not shown); hence, the probes we used were specific.

The replication rate of the chromosome increased significantly at the higher temperatures relative to that at 30°C (see figure 17A). The replication rate of the plasmid increased more than the rate of chromosome replication at higher temperatures (see figure 17 B, dark grey bars). Panel A shows that the eclipse period for the plasmid decreased 38%, from 85 minutes to 50-55 minutes at the higher temperatures (light grey bars). The eclipse period of the plasmid pOU71 dropped more drastically (85%) from 27 minutes at 30°C to 4 minutes at 42°C.

Panel B in figure 17 illustrates that the eclipse period (light grey bars) of the chromosome was quite constant in all five experiments, ca 65 % of the doubling time of the chromosome.

The eclipse period oft the plasmid was ca 20% and 15% for the runaway plasmid at 30°C steady state and the control plasmid pGW71 at steady state 42°C, respectively (see figure 17C). The eclipse period of the plasmid relative to doubling time of the plasmid are increased at 37°C and 38°C to ca 25% while it drops to ca 10%.

Density shift

Temperature shift at OD550=0.1

30°°°°C, Heavy medium

37,38 & 42°°°°C 37,38 & 42°°°°C, Light medium

Fig. 16. Experimental schedule for measurements of the eclipse period during the upshift experiment

30 min

Time samples

Page 45: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 43

The plasmid-eclipse at 30°C was 0.2 generation times (plasmid); this is almost equal to the eclipse period measured on the IntR1-strain (0.22 generation)(Gustafsson and Nordström, 1978; Koppes and Nordström, 1986). So the eclipse period for the same basic replicon have been reproduced. It is a little bit surprising that our control plasmid pGW71 had a shorter eclipse period (0.15).

The temperatures and the coexisting plasmids did not affect the eclipse periods of the chromosomes. The chromosomal eclipse was found to be within the range of 0.60-0.70 generations. We reported previously eclipse periods for the wild type chromosomes 0.60-0.75 generations (Olsson et al., 2002; papers I and II).

At the intermediate upshifts (37 and 38°C) we could see increased eclipse period (0.25) which is opposite to the predictions based on the Multistep control model. One explanation could be that the plasmids had already managed to reach a new steady state in copy number, or maybe the plasmid copy number increased too much so the system has to down-regulate; this could explain the longer eclipse. The Multistep-control model predicts a longer eclipse in a downshift situation. I think, however, that the differences are too small to allow any definite conclusions.

In the larger upshift experiment (30°C=>42°C) the eclipse period was reduced from 0.20 to 0.10, an almost 50%-reduction. I think this difference is more significant and this supports the Multistep control model. In this more drastic upshift the plasmid replication was more out of control. The length of the eclipse period, however, was not completely abolished (4 min). We think that these 4 min reflect physical constrains; it is reported that it takes about 4 minutes for a newly replicated plasmid to regain its CCC-structure, which is the substrate for new initiation of replication (Gustafsson et al., 1978; Uhlin and Nordström, 1985).

Our results indicate that plasmid-copy number responded faster than we believed when we designed the upshift experiments. Therefore, we wanted to measure the copy number increase in detail.

0

25

50

75

100

125

150

30C=>

30C

30C=>

37C

30C=>

38C

30C=>

42C

42C=>

42C

min

ute

s

chromosome(gen)eclipse length (chrom)plasmid (gen)eclipse length (plasmid)

0

25

50

75

100

125

150

30C=>

30C

30C=>

37C

30C=>

38C

30C=

>42C

42C=>

42C

Rel

ativ

e th

e g

ener

atio

n t

ime

of

the

chro

mo

som

e (%

)

chromosome(gen)eclipse length (chrom)plasmid (gen)eclipse length (plasmid)

0

5

10

15

20

25

30

30C=>

30C

30C=>

37C

30C=>

38C

30C=>

42C

42C=>

42C

Rel

ativ

e th

e g

ener

atio

n t

ime

of

the

pla

smid

(%

)

eclipse length (plasmid)

Fig. 17. Data from the upshift experiments. The graphs show the length of the generation time of the chromosome (black bars), the eclipse period of the chromosome (light grey bars), the generation time of the plasmid (dark grey bars), and the eclipse period of the plasmid (white bars). The graphs include results from the upshifts from 30 to 37, 38 and 42°C. Two control experiments are included; the first is at steady state growth with plasmid pOU71 (30°C=>30°C) and the second is the control plasmid pGW71 at steady state growth at 42°C (42°C=>42°C). The left panel shows the data in physical time (A). The middle panel presents the data relative to the doubling time of the chromosome (B). The right panel shows the eclipse period of the plasmids relative to the doubling-time of the plasmid (C).

A B C

Page 46: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

44 Jan Olsson

In conclusion: The eclipse periods of the chromosome in our experiments remained long at all temperatures tested. The growth rate for our strain, containing the runaway-plasmid pOU71, was decreased at 42°C relative to the same bacteria with our control plasmid pGW71. We could not see any shorter eclipse period for pOU71 at the milder upshifts (to 37 and 38°C). The plasmid-eclipse was significantly shorter at the major upshift, 0.10 relative to the control at 30°C (0.20) this supports Multistep control. The eclipse period in this major upshift was 4 min, probably due to structural requirements for the plasmid to regain the right supercoiling.

We have shown that both the Multistep control system and the structural aspects contribute to the eclipse period of plasmid R1. Copy-number increase at temperature shifts from 30°C to 37, 38, 39 or 42°C We have investigated the copy number increase when the runaway-plasmid pOU71 was shifted to different temperatures. The growth medium was minimal-medium supplemented with glucose. Samples were taken at intervals after the temperature shift and analysed for the relative concentration (relative copy number) of the plasmid, expressed as the ratio between plasmid DNA (measured with an R1-specific probe) and the chromosome (the tus-gene). The relative copy number increased immediately after the temperature shift. The increase decelerated asymptotically until the plasmid copy number reached a plateau indicating the new steady state level (see figure 18).

In our upshift experiments we measured the eclipse period after 30 minutes in the new temperatures (see figure 15). We wanted to investigate if experimental measurements of the eclipse period could confirm the Multistep model. The model predicts an abolished or shorter eclipse period during upshift. The eclipse period was 0.10 generations at the upshift to 42°C. We could not detect any reduced eclipse for the upshifts to 37 and 38°C. I think we might be able to see a reduced eclipse if we repeats the density shift much closer to the temperature shift. We have indicated the relative plasmid-copy number at 30 minutes with an arrow in figure 18. The increase

ratio

pla

smid

/chr

omos

ome

time (min)

1

10

100

1000

0 50 100 150 200

4 2 °C

3 9 °C

3 8 °C

3 7 °C

3 0 °C

Fig. 18. Replication of plasmid pOU71 after shift from 30°C to 30, 37, 38, 39 and 42°C. The ratio plasmid/chromosome was determined by hybridisation with radioactively-labelled probes specific for the R1-basic-replicon and the chromosomal tus-gene.

30 min

Page 47: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 45

in copy-number was less at 30 min after the temperature shift than immediately after the temperature shift. Down shift experiment gave surprising results (paper IV) We have shown that both the Multistep control system and the structural aspects contribute to the eclipse period of plasmid R1. We want to investigate the Multistep control model further. During a downshift the gene-dosage of CopA is high and the transcription rate of RepA-mRNA/plasmid is not elevated. Since RepA translation and binding of RepA to oriR1 is in cis and CopA acts in trans the RepA synthesis will drop drastically (see figure 15B). The model predicts an extended eclipse period during these conditions (Ehrenberg and Sverredal, 1995; Paulsson and Ehrenberg, 2000).

The theoretical simulation was based on some assumptions. One assumption was

that the cell population is homogenous. All cells grow with the same growth rate. Another assumption was that all plasmids that are not restricted by the eclipse period have the same probability to replicate. With these assumptions the maximum level of the HL-peak is restricted to the interval 50-100%, according to the formulae 3 on page 27.

Figure 19 shows both the schemes of the procedures of the minor and major down shifts and the control experiment. The down shifts were from 7 and 50 times

Fig. 19. Experimental schedule for measurements of the eclipse periods during two different downshift experiments. The copy number was increased to ca 7 and 50 times higher copy number than at the steady state level at 30°C by incubation at 39°C in 40 min and at 42°C in 60 min, respectively. The black silhouettes show the distributions of HH, HL and LL-plasmid DNA in the CsCl-gradient using R1-specific probe. The right panels show the kinetics of R1-specific HH, HL and LL-DNA after the density shifts. The black symbols indicate the HL-DNA. The peak value for the HL-DNA was 35%, 28% and 60% for the small and large downshift and the control at 30°C, respectively.

Density shift

Temperature shift at OD550=0.1

30°°°°C, Heavy medium

39°°°°C

30°°°°C, Light medium 40 min

Time samples

Density shift

Temperature shift at OD550=0.1

30°°°°C, Heavy medium

42°°°°C

30°°°°C, Light medium

60 min

Time samples

Temperature and density shift at OD550=0.1

30°°°°C, Heavy medium 30°°°°C, Light medium

Time samples0 60 120 180 240 300

0 5 10 15 20 25

0 5 10 15 20 25HH HL LL

HH HL LL

HH HL LL hours

hours

min

Page 48: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

46 Jan Olsson

higher copy number than at steady state at 30°C. The plasmid amplification before the downshifts was estimated from the graphs in figure 18.

The result of our measurement gave an HLmax<50% which is impossible with the assumptions on which the theoretical simulations are based. The minor down shift from 7 times higher copy number than the wild-type had an HLmax of 35%. The major downshift from 50 times the normal copy number gave an HLmax of only 28%. One situation that could give similar results is that the cell population is non-homogenous e.g. if a subpopulation is dead at the downshift. This is however not probable since all HH-DNA specific for the chromosomes replicated within 1.5 cell-doublings times.

Another possible explanation is that a newly replicated plasmid has a higher probability of being selected for the next replication. This is however against previous assumptions and results, in which the plasmids are supposed to replicate randomly distributed over the whole cell cycle (Gustafsson et al., 1978). Our group and other groups have shown that R1 and other plasmids are localised in clusters (Eliasson et al., 1992;Weitao et al., 2000; Pogliano et al., 2001). Weitao et al. studied one plasmid with wild-type copy number pKN501 (Molin et al., 1979) and the copy-number (copA) mutant pKN907 that has 3.7-fold higher copy number. Both these plasmids were parA-. The number of clusters in the cell with increased copy number did not increase (Weitao et al., 2000).

We do not know if the number of clusters increase when we increase the copy number with our plasmid (pOU71) to 7 or 50 times the normal copy number at 30°C. Maybe only some clusters have the potential to replicate. This means that the eclipse period determined at the upshifts experiments might be underestimated if the plasmid population was non-homogenous. The HH-peak of plasmid DNA did disappear sooner or later, so a subpopulation of plasmids was not permanently withdrawn from replication. If the number of plasmids in each cluster within a cell varied, the larger clusters might have lower replication frequency relative to small clusters. Maybe the local CopA concentration varies between small and big clusters within the same cell.

It is also possible to explain our results if we assume that we have copy-number variation among the clusters in a cell and that partition system acts rather on cluster-level than at individual plasmid level.

There is no doubt that the plasmid population was not homogenous. The limitation with all our experiments is that we tries to study parameters within the cells but can only do macroscopical experiments. Therefore we can not exclude that our results reflects variation in plasmid copy number among the cells.

The +n mode of replication fits with our data from the downshift experiment The copy-number control of the plasmid adjusts the rate of replication at upshifts and downshifts. Each daughter cell gets n plasmids on the average after cell division. So on the average the number of plasmid copies increases with n copies during each cell cycle. Plasmid pOU71 has different copy numbers at different temperatures; hence, the number of plasmids in a newborn cell varies with temperature due to the temperature-regulated lambda promotor λPR. In the +n mode of replication the plasmid copy number in a cell increases with n plasmids irrespective of the number in the newborn cell.

The +n mode was suggested from experiments with downshifts from 3-5 times higher copy-number then the copy number at steady state. Our downshifts were from ca 7 and 50 times higher copy number; hence, more drastic downshifts then previous experiments. So we wanted to investigate whether the +n mode fits more drastic

Page 49: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 47

downshifts. We could determine the DNA-synthesis of chromosomal DNA and plasmid DNA from the density-shift experiment. The relative DNA synthesis was estimated from following formula:

DNA synthesis after the shift = (2LL+HL)/HH (5)

The DNA synthesis of the plasmid was also simulated for the +n mode for 10 different downshifts when the relative copy number before the shift were: 1, 4, 7, 10, 15, 20, 30, 40, 60 and 80 times relative to the normal copy number. We compared our experimental data with the simulated data (see figure 20).

The rate of chromosome synthesis (triangles) after the downshift was almost

constant. The plots of the DNA synthesis of the plasmid (squares) after the major downshift coincided perfectly with the predictions of a 50-fold downshift in copy number. Hence, our experimental data fits perfectly with the +n model (see figure 20). Our data for the minor down shift also agreed with the simulated plasmid synthesis after a 7-fold downshift (see paper IV).

In conclusion: We used plasmid pOU71 to investigate the Multistep-control model. The model predicts a longer eclipse during down shift. We could not see any eclipse, since the plasmid population was not homogenous during the downshift. Hence, it is the first time that anyone has observed a plasmid population in which newly replicated plasmid copies have a higher probability of replication than the rest of the population.

The measurements of chromosome and plasmid synthesis after our down shifts support the +n mode of plasmid replication.

1

1 0

1 0 0

0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 2 2 2 4

plasmidschromosome

1

1 0

1 0 0

0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 2 2 2 4

tim e (h)

Dow nshift from 50 (42°C) to 1 (30°C)

rela

tive

DN

A c

on

ten

t

1 4 7 10 15 20 30 40 60 80

FIG. 20. Comparison of experimental data from a downshift experiment with prediction of the +n mode of replication. The plots show the relative chromosome (triangles) and plasmid (squares) content after downshifts from 50 times, the normal copy number. The different lines are simulations for downshifts from N times higher copy number than the normal steady state value (N=1, 4, 7, 10, 15, 20, 30, 40, 60 and 80).

Page 50: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

48 Jan Olsson

COPB, A RESCUE SYSTEM? (paper V)

The basic replicon of plasmid R1 encodes a main copy-number-control system in which the antisense RNA CopA is the main player. In addition, there is the CopB system which is assumed to be a rescue system, since it speeds up replication at low R1 copy numbers by derepressing the repA promoter (see figure 8B).

The CopB system is believed to work as a rescue system for cells with low copy number. The CopB protein represses the stronger RepA-promotor located downstream the copB gene (see figure 8A). CopB act as a tetramer to repress the RepA-promotor and is believed that the concentration of tetramers is drastically reduced at low copy numbers due to gene-dosage effects. Hence, at low number of plasmid copies the RepA-promoter might be released from the inhibition and thereby the transcription rate of RepA mRNA increases (see figure 8B). In cells with low number of plasmid copies the rate of plasmid replication will increase and thereby an increased number of plasmid copies will be formed before the next cell division.

The probability for plasmid-free daughters increases with decreasing copy number in the mother cell. The loss frequency is described by this formula:

P(loss)= 2-m (m= the number of plasmids in the mother cell) (6) If the plasmid lacks a partition system, all plasmid copies are randomly

distributed to the daughters. This leads to a bias because mother cells with less than 2n will contribute with more plasmid free daughters than mother cells with more than 2n copies.

If the plasmid has a perfect equipartition system only mother cells with one single copy will contribute to plasmid free daughters. Therefore, the copy-number distribution is of utmost importance for the stability of maintenance of a plasmid.

The copy-number distribution of R1 is not exactly known. The distribution has been estimated indirectly by Løbner-Olesen (1999). He used an inducible gfp-gene as a reporter gene. The gene was induced during ca 25% of a cell cycle and the GFP expression was analysed in a flow cytometer. This was the first attempt to determine plasmid copy number distribution in E. coli population. The plasmid distribution for R1-plasmid was wider (CV∼40-50%) than the cell size distribution (CV∼21-24%).

The +n mode predicts a Poissonian distribution around 2n in the mother cells.

Previous experiments on the CopB-control system The CopB-system can be bypassed if extra CopB is synthesised from a coexisting compatible plasmid. Plasmid pOU18 and pOU16 are compatible with plasmid R1 and pOU16 has a higher copy number then pOU18 (Riise et al., 1982).

Previous experiments have compared the plasmid-loss frequency for parA--plasmids without or supplemented with extra CopB from plasmid pOU18. The loss frequency increased from 1.5% to 2.5% with the extra CopB-producing plasmid (Nordström and Aagaard-Hansen, 1984). They repeated the same experiment on a parA+-plasmid. They did not see any effect on plasmid stability if extra CopB was expressed. However, the killer system was present on their parA+-plasmid and did eliminate most of the plasmid-free cells.

Page 51: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 49

CopB-control system has a larger effect on the stability of parA+-plasmids We wanted to investigate whether this mild CopB effect on the parA--plasmids could be stronger if we repeated the experiment on parA+-plasmids without any killer system.

We choosed R1-plasmid pOU47 in our experiments. Plasmid pOU47 has a parA+-locus and the lac-operon. We used the lac-operon as a reporter for presence of pOU47 in E. coli. We grew the cells exponentially for at least 10 generations with selection for the R1-plasmid. After the release from selection the cells were kept in exponential growth by sequential dilutions with fresh medium.

We used plasmid pOU18 to supply the cell with extra CopB and screened plasmid free cells as white colonies during ca 100 generations. The loss frequencies were 1⋅10-4 per cell generation for plasmid pOU47 alone and 2.5⋅10-4 per cell generation for plasmid pOU47 in the presence of pOU18 (see figure 21). The effect of extra CopB was similar in the presence of the parA region as in absence of parA. However we were a little bit suspicious, maybe our extra CopB failed to repress the RepA-promotor enough?

Plasmid pOU18 and pOU16 have two different basic replicons, pOU16 has higher copy number. We showed that pOU16 repress the RepA-promotor 7 times more efficient than with pOU18.

We co-transformed pOU47 and pOU16 in our strain without functional lacZ-

gene. We grew the cells for 20 generations with selection before the release. The frequency of plasmid loss was now 6⋅10-4 per cell generation (see figure 21).

Hence, the frequency of plasmid loss was significantly increased when we bypassed the CopB-control system.

The fairly moderate effect of loss of derepression of the prepA promoter raises questions about the importance of the CopB system for the plasmid. There are at least two situations at which the CopB system might make a difference: i) in cells with a very low copy number due to statistical variations in replication and partition, and ii) during establishment of the plasmid directly after transfer to a plasmidfree cell; the rate of replication is six-fold higher during the first quarter of a generation time after

0 1 0 20 30 40 50 60 70 80 90 1 00 1 1 0

Fra

cti

on

pla

sm

id R

1 co

nta

ing

ce

lls

(%

)

generat ions

p O U 4 7

p O U 4 7 +pO U 1 8p O U 4 7 +pO U 1 6

100

99

97

98

96

95

94

93

FIG. 21. The fraction of the cells that carry plasmid pOU47 during exponential growth of populations of TOP10-pOU47 (closed diamonds), TOP10-pOU47 + pOU18 (open squares), and TOP10-pOU47 + pOU16 (closed triangles) as a function of time (generations). Note that the scale of the vertical axis is logarithmic.

Page 52: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

50 Jan Olsson

conjugal transfer of plasmid R1 compared to the steady-state level (Gustafsson and Nordström, unpublished). Since the CopB system has been kept by the plasmid it appears as even its fairly moderate effect has survival value for the plasmid.

In conclusion Plasmid R1 is a low-copy-number plasmid that encodes a partition system. The auxiliary copy-number-control system works as a rescue system that speeds up replication at low plasmid copy numbers. We could show a six-fold reduction of the rate by which plasmid-free cells are formed for parA+ plasmids, this is 4 times stronger effect than previous experiments on parA- plasmids. Wild type R1 plasmids are parA+, hence the CopB systems seem to have relevance.

Page 53: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 51

DISCUSSION

(INITIATION OF CHROMOSOME REPLICATION)

We have investigated two important parameters that describe the properties of the cell cycle of E. coli and the correlation between them. The minimum time during which a newly replicated oriC is prohibited to reinitiate replication is called the eclipse period. When E. coli grows with overlapping C- and D-periods two or four origins are initiated in each cell cycle. It is well known that all oriCs fire synchronously in E. coli. This is also true for minichromosomes.

Perturbation of different control systems in the cell cycle can cause asynchrony: 1. In the sequestration process the balance between the expression of the seqA-gene

and the dam-gene is important to assure synchrony. Lack of the seqA-gene or the dam-gene cause asynchrony. Also over- or under-expression of the dam-gene cause asynchrony (Boye and Løbner-Olesen, 1990). SeqA requires the SeqB-factor to associate with SeqA and hemimethylated oriC to the cell membrane (Shakibai et al., 1988). This binding of hemimethylated oriC to the cell membrane prohibits reinitiation of oriC-replication (Landoulsi et al., 1990), and maybe by blocking the remethylation.

2. Mutations in the replication-initiator protein DnaA cause different degrees of

asynchrony. DnaA co-ordinates the synchronous initiation. The level of DnaA expression sets the initiation mass. Oversupply of DnaA decreases the initiation mass but does not affect the synchrony (Skarstad et al., 1989; Løbner-Olesen et al., 1989). The active ATP-DnaA initiates replication from oriC (Sekimizu et al., 1987). The inactivation is regulated by the RIDA system. Reactivation of ADP-DnaA requires normal level of acidic phospholipids in the cell membrane. DnaA proteins that cannot bind ATP cause asynchronous replication, e.g. the dnaA46 mutant (Hwang and Kanugi, 1988; Skarstad et al., 1986) since the RIDA system is bypassed. It is a paradox that ATP-DnaA is required to initiate replication from oriC, but the DnaA46, which cannot bind ATP even at permissive temperature. The dnaA46 mutant does replicate. There are two possible explanations. (1) There are pathways for chromosome replication that do not require ATP-DnaA. (2) The ATP-bindings was to low to be measurable but enough for replication. DnaA proteins that cannot bind acidic phospholipids also cause asynchronous replication (Zheng et al., 2001). I think the balance of inactivation of ATP-DnaA and the reactivation of ADP-DnaA is important for synchrony.

3. The datA-locus is located ca 10.7 min clockwise to oriC. The datA-locus binds

more DnaA molecules more efficiently than oriC. When the datA-region replicates all active free ATP-DnaA molecules will be titrated while no new DnaA is produced as dnaA promotor is sequestered, in concert prohibiting reinitiation of oriC-replication. Deletion of the datA-locus causes asynchronous over-initiation (Hansen, et al., 1987; Kitagawa et al., 1996 and 1998).

Page 54: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

52 Jan Olsson

The mutants that we have investigated have defects in these three listed control systems: the sequestration control, the transitions between ADP-DnaA and ATP- DnaA and the DnaA-titration by the datA-locus. It is very difficult to separate the cause for the asynchrony in our mutants, since the control systems overlap and interplay with each other.

It seems obvious that a reduced eclipse period causes asynchrony and vice versa. But that statement is not completely logical. Because it is possible, theoretically to imagine asynchronous replication with long eclipse period (see figure22A). It is also possible to imagine the opposite, a strain with synchronous initiation but short eclipse periods (see figure 22B) Requirements for an eclipse period We decided to measure the eclipse period in our mutants and in the wild type. The eclipse period was found to be a constant fraction of the generation time independent of the growth rate, and not a fixed physical time like the C and D-periods.

The eclipse period was shorter in the seqA mutant (0.14 generations) than in the dam mutant (0.40 generations). This implies that the SeqA protein has additional roles in initiation control other than sequestration. Our lab showed that the seqA mutant has an increased level of negative superhelicity (Weitao et al., 1999 and 2000). We wanted to know if this increased negative superhelicity alone could cause the shorter eclipse period in the seqA mutant. We also wanted to investigate whether the different eclipse periods are reflected by different degree of asynchrony.

We studied the eclipse period and measured the asynchrony index in the gyrB225topA mutant, which is known to have increased negative superhelicity. The

Fig. 22. It is theoretically possible to imagine asynchronous replication with long eclipse periods (A) or synchronous replication with short eclipse periods (B). Hatched lines indicate cell divisions.

A

B

Page 55: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 53

double mutant initiated chromosome asynchronously. Increased negative superhelicity reduced the eclipse period by ca 30 % (0.75=>0.5). Hence, an intact sequestration machinery cannot solely assure synchrony and a long eclipse period.

The eclipse period has previously been measured on synchronised cells (Bogan and Helmstetter, 1997; von Freiesleben et al., 2000). The cells were synchronised by either a DnaC2 mutant or a DnaA46 mutant. Hence the previous eclipse measurements were never done on the wild type strain. Our results on the seqA mutants showing drastically reduced eclipse period are similar to von Freieslebens results. However, the eclipse period was completely abolished in his seqA mutant, but we saw a significant eclipse period (0.14). The difference between our experiments is that he measured the eclipse period in a synchronised double mutant dnaA46seqA and we measured the eclipse in an exponentially unperturbed growing single mutant seqA. It is known that the dnaA46 mutant initiates chromosome replication asynchronously. We wanted to investigate if the dnaA46 mutant has a reduced eclipse period and if we can see an additional reduction of the eclipse period in the double mutant dnaA46seqA. The RIDA system and the DnaA-titration by the datA-locus are bypassed in the dnaA46 strain since DnaA46 cannot bind ATP. DnaA46 had a reduced eclipse (0.3); we have thus found another strain with intact sequestration and reduced eclipse period. The length of the eclipse period in the dnaA46 mutant might reflect the time SeqA prohibits remethylation of oriC. We saw a mild reduction in the double mutant dnaa46seqA (0.10) relative to the single mutant seqA (0.14).

We have seen that a long eclipse period requires, a functional DnaA protein, an intact sequestration machinery and right degree of superhelicity.

Correlation asynchrony index against eclipse period Many different mutations can cause asynchrony. The asynchrony index (A) is a quantitative parameter for the degree of asynchrony. We have investigated the nature of the asynchrony index and the correlation between asynchrony against the length of the eclipse period. The value of the asynchrony index is growth-rate dependent. The asynchrony index increased if the growth rate increased, reflected by the average number of oriC per cell. It is therefore important to consider the growth rate when asynchrony indices shall be compared.

We wanted to investigate the correlation between the degree of asynchrony and the length of the eclipse. We wanted to avoid the variation of the asynchrony index due to different growth rate, so we divided the asynchrony index with the average number of oriC. We found a step-like correlation between the length of the eclipse period and the degree of asynchrony. Strains with a reduced eclipse ≤65% of the normal eclipse period were asynchronous.

Future projects: We have so far only seen asynchronous strains which have reduced eclipse period. We would like to measure the eclipse period in other asynchronous strains • It is interesting to measure the eclipse period on other asynchronous strains. We

would like to measure the eclipse period on the ∆datA mutant, which is an asynchronous strain with the right degree of superhelicity, intact sequestration machinery and intact DnaA-protein.

• I would like to investigate further how SeqA and superhelicity affect asynchrony and eclipse period. It is known that the mukB mutant recovers the normal

Page 56: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

54 Jan Olsson

chromosome condensation and degree of superhelicity (Weitao et al., 1999; 2000). It would be interesting to see if over-expression of SeqA can reduce the negative superhelicity in the double mutant gyrB225topA and if the eclipse period is extended and synchrony regained.

• The eclipse period in the seqA mutant is shorter than in the dam mutant. One explanation could be the increased negative superhelicity in the seqA mutant. However the superhelicity should be measured also in the dam mutant. If a difference in superhelicity that explains the different eclipse periods, should the triple mutant damgyrB225topA have similar phenotype to the seqA mutant and hence, a similar eclipse period.

• We have only measured the eclipse period in strains that are "truly" asynchronous. Some dnaA mutants are asynchronous on an intermediate level, it would be interesting to measure the eclipse in some of these strains.

• The dnaA46 mutant cannot bind ATP, which is believed to be required for normal initiation from oriC. It would of great interest to see where replication is initiated from in the dnaA46 mutant. We suggest marker frequency analysis.

• The eclipse period were extended in a strain with a 66% reduced dam-expression (von Freiesleben et al., 2000). Boye and Løbner-Olesen reported asynchrony when the dam expression was lower than in the wild-type strain (1990). Does this indicate that a long eclipse period and asynchrony is possible?

Page 57: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 55

DISCUSSION

(PLASMID)

The upshift supports Multistep control We wanted to investigate if the Multistep control contributes to the eclipse

period measured by Gustafsson et al., (1978) and Koppes and Nordström, (1986). The Multistep control suggests that many steps with exponential probability distribution improve the copy number control. The experimental data were theoretically simulated with the Multistep control model (Ehrenberg and Sverredal, 1995). If the Multistep control exists in plasmid R1, it should be possible to manipulate the length of the eclipse period by altering the no. of controlling steps. It is postulated that RepA proteins are translated in bursts from each RepA mRNA. Initiation of replication at oriR1 requires many RepA proteins. However the CopA antisense-RNA interrupts the RepA synthesis and a new RepA-mRNA has to be transcribed to synthesise the missing RepA-proteins. If the relative CopA concentration is decreased, the probability to translate all RepA proteins within a single burst, reduce the no. of steps required to relax the stringency of copy number control will increase. The Multistep control model predicts an abolished or at least reduced eclipse period under these circumstances. If the relative CopA concentration is increased less RepA will be synthesised per repA-mRNA; hence, the number of bursts will increase and the eclipse period will be longer (Ehrenberg and Sverredal, 1995; Paulsson and Ehrenberg 2000). The relative CopA concentration increases during a downshift of the plasmid copy number and decreases during an upshift. So the predictions is that the eclipse period will be reduced during an upshift and extended during a downshift. We made upshifts and downshifts with pOU7, an R1-plasmid, which increases its replication rate at temperature upshifts and decrease the rate at downshifts.

We could see a significant reduction of the eclipse period during a big upshift. However minor, upshifts resulted in a slightly longer eclipse, contrary to the predictions. The big upshift reduced the eclipse period from 0.20 to 0.10 plasmid duplications and this supports the Multistep control model. The eclipse was 4 min, but not completely abolished. This physical eclipse with the length of 4 min is probably due to structural constrains. The results were not very clear since the minor upshifts gave a little bit longer eclipse, which is opposite to the predictions.

When we tried to extend the eclipse period by a downshift experiment, the eclipse period was reduced to a negative value. This is not possible. We could not support the Multistep control model from our downshift experiment.

The plasmid synthesis after the downshift was delayed, following the kinetics predicted by the +n mode. This is the largest experimental downshifts / 50 and 7-fold that have confirmed the +n mode. Previous downshifts were 3 and 5 times higher copy number before the shift than the steady state copy number after the shift. Heterogeneous plasmid population The theoretical simulation of the eclipse period assumes that the plasmid population is homogeneous. The eclipse period is estimated from the maximum size of the fraction of Heavy-Light DNA, HLmax (see formulae 3) The HLmax was 35 and 28%, respectively which gives negative eclipse periods -0.3 and -0.44 generations during a downshift from 7 and 50 times the copy number after the density shift at steady state.

Page 58: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

56 Jan Olsson

A negative eclipse implies that the plasmid population is heterogeneous. Earlier experiments have proposed that the plasmid populations are homogenous and all copies have the same probability of being selected for replication (Gustafsson et al., 1978). Therefore, the computer simulation assumed a homogenous plasmid population.

The heterogeneity means that plasmids that replicates have a higher probability to replicate a second time relative the bulk of non-replicating plasmids. There are many possible explanations to this result. • A subpopulation of cells is dead during the downshift. This was not true since the

chromosomal HH-fraction disappeared within 1.5 cell doubling times. • Newly replicated plasmids have a higher probability to replicate than older

plasmids within the same cells. Our group and many other groups have shown that plasmids are localised in specific foci in the cell, (Eliasson et al., 1992, Weitao et al., 2000b; Pogliano et al., 2002). Weitao et al. (2000b) have shown that the average number of foci (2.3-2.4) were the same for a wild type plasmid and a copA mutant with 3.7 times higher copy number. If the foci have different size the local CopA concentration might vary between the clusters and therefore plasmids would replicate with different rates.

• It is also possible to explain these results even if the CopA concentration is homogenous in the cell if the copy number varies between the foci and the partition systems acts on cluster-level instead of on individual plasmids.

• If there is big variation in copy number between individual cells, we could get similar results.

There is no doubt that the plasmid population was heterogeneous during down shift. We do not know on which level we have the heterogeneity. We do not know whether the plasmid population is heterogeneous also at steady state and during upshift. If the plasmid population is heterogeneous at upshift and steady state our eclipse-periods are shorter than the true eclipse period. Hence our way to measure the eclipse period requires a homogeneous plasmid population, while measurement of the eclipse period with pulse labelling in combination with Meselson-Stahl density-shift just measures the eclipse period. Excess CopB reduce the plasmid stability The CopB system is believed to work as a rescue system for cells with very low copy number (see figure 8). The CopB system can be bypassed if extra CopB is supplemented from a compatible plasmid. Daughter cells have on average n plasmids. The plasmid copy number is believed to have a Poissonian distribution centred around 2n in mother cells just before cell division. There is a bias were mother cells with copy numbers ≤ 2n have higher probability to lose the plasmid P(loss) in one of the daughters than mother cells ≥ 2n. In mother cells with only one plasmid, one daughter cell will not receive a lose the plasmid copy even if the plasmid has an efficient partition system, e.g. parA in plasmid R1.

Previous experiments have shown a mild effect on the loss frequency if extra CopB is supplemented from a compatible plasmid pOU18 (2.5%/cell cycle) in comparison to intact CopB system (1.5%/cell cycle) if the plasmid is parA deficient. In a strain with a parA deficient plasmid, the plasmid can be lost from all mother cell following formulae (6). If the plasmid has a partition system, only mother cells with one plasmid will contribute to loss of plasmids. The release from CopB repression on

Page 59: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 57

the PRepA promotor is bigger in cells with very low copy number. So we thought that the CopB system is more important in plasmids with partition system.

We have studied plasmid stability on an R1-plasmid with partition system without extra CopB and an R1-plasmid supplemented with two levels of extra CopB. We saw that the loss rate was 1⋅10-4 per cell cycle for the sole R1-plasmid. The loss rate increased to 2.5⋅10-4 per cell cycle and 6⋅10-4 per cell cycle when the CopB system was bypassed with different gene-dosages of extra CopB.

We have shown that the CopB system probably has a more important role on the stability of plasmids with partition system. And the CopB system is probably even more efficient when a plasmid conjugate into a cell without any CopB protein. Our lab has shown that the RepA synthesis increased 6-fold after conjugation. It is however difficult to understand the importance of the CopB system for full size R1 plasmids (see figure 7) where probably many systems supplement each other: (i) two killer systems, parB and parD might kill cells that have lost the plasmid. (ii) the plasmids can conjugate to plasmid free cells. (iii) the CopB system speeds up the RepA synthesis in cells with low copy number.

Page 60: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

58 Jan Olsson

ACKNOWLEDGEMENTS

I first will mention the two most important persons during my time as a PhD-student. One helped me to focus my work; the other has been the visionary helping me to see beyond the horizon of present methods and knowledge.

• I want to thank Kurt Nordström my supervisor, you have always supported me even when I thought that everything was dark and completely hopeless. I am grateful for all exciting scientific discussions. It has always been a pleasure to discuss results with you. You have helped me to focus on the important experiments. I will always remember your nice cheerleader dress when you asked me to become your Ph.D.-student. The dress unmasked your sense of humor and complex personality.

• I have got a great friend during my time at BMC. Santanu you have offered me a very special gift,

your friendship. I cannot list all things you have done for me because my thesis has a limited number of pages. I will just mention some of them. You have shown me science and how fantastic the cell cycle is. Your sharpness supervised the optimization of the density shift experiments. I want to thank you and your generous family for hospitality, friendship and inviting me to India two times. Thanks for all Indian gastronomical pleasures, your lamb is the best in the world! You are very sharp and show a great amount of empathy and generosity.

Scientists need a lot of support. I want to thank all staff who gives our department continuity and a great atmosphere: Linda, Ingrid, Susanne, Sigrid and Christina F Thanks for all hidden job like salaries, organizing the economy etc. Christer and Joakim for repairing all machines. The cleaning staff and the staff at glassware-department. Solan and Ulrika for nice company in the lab. Eva F thanks for all media, plates, I have really appreciated your job and I have always felt that all media and plates are properly prepared, thanks for nice company when I used your scale. I would of course have failed to get a PhD without all other friends. I apologize if you cannot find your name here

• Sophie for introducing me to climbing, all the delicious sausiz from France, and great company in

Nepal. • Johan Paulsson, for friendship, good diving in Mexico and all plasmid-chats during my first three

years in Uppsala, your chanterelle place, Do you remember when we bought enormous amount of Coke because Pripps stopped the production in Sweden.

• Otto Berg for beeing a great teacher, you made thermodynamics very exciting and thank you for all help with the theoretical analysis on DNA replication after density shifts.

• Lena M and Tordan for friendship and delicious dinners, It is always nice to talk with you. • I thank Åsa-påsa for introducing me to Meselson-Stahl density shift experiment during your stressed

last months before you own defense. I thank Dorothee for all cookies and funny jokes. Björn Gullbrand the tivoli champ for all interesting chats about more or less scientific results, snowboarding in Santa Fe etc… . Tao for believing your dreams. Emilia for being such a warm and great friend., LinaJ, Helen W, Hywel

• Fredrik Söderbom helping me with my pdf-problems and all very nice chats over lunches. • Flärdh-group. Klas for discussions and Nina its always nice to talk food and sharing recipes. • Bernanders group, Rolf for all great feedback during all cell cycle meetings, Laurance (Philip) for

different French delicatessen, Andrzej being so crazy (like participating in Greece Jujutsu Open with yellow belt) and your warm personality, MagnusL Cecilia and Karin H, I will never forget the Budapestcake the best start for a long friendship.

• SolveigR my sweetest labteacher , offering icecreams, your hospitality in Boston when I came with fever and a cold.

• Helena S and Lotta I for all support at Amersham. • Other friends from BMC Kai, Hilde, Aneta, Gopal thanks for the socks, Sonchita I had a great

company at Nicopark.

Page 61: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 59

• The IBG-staff, Ylva, Kerstin, Cecilia, Gunnar A, 2x Janne , Matti, Elsbeth etc.and Eva Björkman it has been such a pleasure to collaborate with you and to come and chat now and then.

• Kirseboms group, JoannaK, Ema, Dave, Nisse I will never forget the fjords, shrimps and the great weddingparty, Fredrik and Mattias for great roomcompany. Björns S, thanks for borrowing your laptop.

• Diarmaid and his group, Patti all luck for the new family member, Tobias, may the force be with you and your enzymes , Peter, Mirjana one of my two favourite neighbours for lending me clothes, high heel boots and all other assesoars for a great party. Linda for being nice and good luck with your Capoeira

• Virtanens group Lotta, Helena N, Niklas, Nicos, Javier, Per my second favorite neighbor. Yangou and Christina K for always bothering how I am.

• Cristobals group, alltid trevligt att prata med dig, Jia lets eat Korean food someday • Björn Grynet, Henrik Kaessman. • Ingrid G and Jenny G for great summer collaborations. • Carlssons group, Karin for helping me to teach my most hard course, Anna-Chey for spontaneity,

humor and all agarose, I will do some plasmidpreps for you and Pernilla my Busbästis in BMC. • Wagners group, Gerhard himself always taking time to answer my questions, Copy and Monica being

great roommates, Magnus L for always being funny, Jörg, Klas U. Salme its soon your time good luck, Helena Borg for training me in the Slottbacken before the European championship.

• Minh and Anh-Nhi good luck with your defense and may we have a great party the 23:rd. • Kraftverket the superior ju-jutsu club. I thank Anna D for being the most spjuvrig I know and the

judotraining at café Linné together with Kraftverkets next worldchampion, JennieB, thanks for all training in early mornings and late evenings. All my other jujutsufriends in Uppsala, Erik and Björn, Micke M, Jörgen, Frida, AndersB, CarinaB, Anna B, MarkusT, LenaL, JohanO, Goce and Cecilia helping my voice, Maja and AndersM, Bagge and Magnus always offering sushi, lussebullar, dinner etc. UlrikaB for friendship,Thomas F my naprapat. Malin and Krister I will always dribble when I remembering our fläskmiddagar. My friendly worldchampions in Lyon. And all other friends I have got through the ju-jutsu. All my jujutsufriend in the whole Sweden you will find them on: www.sportju-jutsu.com.

• LinaD and Andy for delicious clams and great pictionary sessions. • Camilla N my dear friend from my undergraduate studies in biology, I hope we will see Gotska

Sandön some day. • Thomas S-P for being a great landlord and all fun during my first years in Uppsala • JohannesL the cave trip was so funny welcome back from Japan. • Anders A and Jenny, the secret behind my SM-gold is your pasta and hospitality., May the next SM

be in Gothenburgh • Otto and Laila for nice sushi-lunch • Sara for salsa dance and cinema • I want to thank all my original friends in Nässjö Elin, Tommy, Olof, Karin, Lena, Roger, Birgitta,

Tomas, Cattis, Svante, MarieA, FredrikP, AnnikaB ….it is great to know that you will always be there when I need you.

• Peder for all great adventures in Träslända • Niklas Graberg initiating my interest for butterflies, fishing, bugs, frogs, lizards and aquariums • Ann-Marie teacher in biology, thanks for your enthusiastic lectures • My father who initiated my interest for science in general. • Mina kära vänner och släktingar i Hamburgsund. Annie och Claerence, Yngve och Gunbritt. • My dear Monica I love you! • Finally I want to thank my dear Mother Kerstin for all love and köttbullar, My dear sister Karin who

knows the art to live a good life, my greatest raw model, Kimsalabim may your dreams come true.

Page 62: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

60 Jan Olsson

REFERENCES

d'Alencon E, Taghbalout A, Kern R, Kohiyama M. (1999) Replication cycle dependent association of

SeqA to the outer membrane fraction of E. coli. Biochimie. 81(8-9):841-6. Atlung, T., and Hansen, F. G. (1993) Three distinct chromosome replication states are induced by

increasing concentrations of DnaA protein in Escherichia coli, J Bacteriol 175: 6537-6545. Austin S, Nordström K. (1990) Partition-mediated incompatibility of bacterial plasmids. Cell. 60(3):351-4.

Review. Bach T, Krekling MA, Skarstad K. (2003) Excess SeqA prolongs sequestration of oriC and delays nucleoid

segregation and cell division. EMBO J. 22(2):315-23. Bakker, A., and Smith, D. W. (1989) Methylation of GATC sites is required for precise timing between

rounds of DNA replication in Escherichia coli. J Bacteriol 171: 5738-5742. Bernander, R., Dasgupta, S. and Nordström, K. (1991) The E. coli cell cycle and the plasmid R1 replication

cycle in the absence of the DnaA protein. Cell 64: 1145-1153. Bernander, R., Palm, J.E., and Svärd, S.G. (2001) Genome ploidy in different stages of the Giardia lamblia

life cycle. Cell Microbiol 3: 55-62. Bird RE, Louarn J, Martuscelli J, Caro L. (1972) Origin and sequence of chromosome replication in

Escherichia coli. J Mol Biol. 70(3):549-66. Blattner FR, Plunkett G 3rd, Bloch CA, Perna NT, Burland V, Riley M, Collado-Vides J, Glasner JD, Rode

CK, Mayhew GF, Gregor J, Davis NW, Kirkpatrick HA, Goeden MA, Rose DJ, Mau B, Shao Y. (1977) The complete genome sequence of Escherichia coli K-12. Science. ;277(5331):1453-74

Blomberg P, Wagner EG, Nordström K. (1990) Control of replication of plasmid R1: the duplex between the antisense RNA, CopA, and its target, CopT, is processed specifically in vivo and in vitro by RNase III. EMBO J. 9(7):2331-40.

Blomberg P, Engdahl HM, Malmgren C, Romby P, Wagner EG. (1994) Replication control of plasmid R1: disruption of an inhibitory RNA structure that sequesters the repA ribosome-binding site permits tap-independent RepA synthesis. Mol Microbiol. 12(1):49-60.

Bogan JA, Helmstetter CE. (1997) DNA sequestration and transcription in the oriC region of Escherichia coli. Mol Microbiol. 26(5):889-96.

Boye E, Løbner-Olesen A. (1990) The role of dam methyltransferase in the control of DNA replication in E. coli. Cell. 62(5):981-9

Boye E. (1991) The hemimethylated replication origin of Escherichia coli can be initiated in vitro. J Bacteriol.173(14):4537-9.

Boye E, Stokke T, Kleckner N, Skarstad K. (1996) Coordinating DNA replication initiation with cell growth: differential roles for DnaA and SeqA proteins. Proc Natl Acad Sci U S A. 93(22):12206-11.

Boye E, Løbner-Olesen A, Skarstad K. (2000) Limiting DNA replication to once and only once. EMBO Rep. 1(6):479-83. Review..

Bramhill D, Kornberg A. (1988) Duplex opening by DnaA protein at novel sequences in initiation of replication at the origin of the E. coli chromosome. Cell. Mar 11;52(5):743-55.

Bravo A, de Torrontegui G, Diaz R. (1987) Identification of components of a new stability system of plasmid R1, ParD, that is close to the origin of replication of this plasmid. Mol Gen Genet. 210(1):101-10.

Brendler T, Austin S. (1999) Binding of SeqA protein to DNA requires interaction between two or more complexes bound to separate hemimethylated GATC sequences. EMBO J. 18(8):2304-10

Brendler T, Sawitzke J, Sergueev K, Austin S. (2000) A case for sliding SeqA tracts at anchored replication forks during Escherichia coli chromosome replication and segregation. EMBO J. 19(22):6249-58.

Campbell JL, Kleckner N. (1990) E. coli oriC and the dnaA gene promoter are sequestered from dam methyltransferase following the passage of the chromosomal replication fork. Cell.;62(5):967-79.

Cooper S, Helmstetter CE. (1968) Chromosome replication and the division cycle of Escherichia coli B/r. J Mol Biol. 31(3):519-40.

Crooke E, Castuma CE, Kornberg A.(1992) The chromosome origin of Escherichia coli stabilizes DnaA protein during rejuvenation by phospholipids. J Biol Chem. 267(24):16779-82.

Crooke E, Thresher R, Hwang DS, Griffith J, Kornberg A. (1993) Replicatively active complexes of DnaA protein and the Escherichia coli chromosomal origin observed in the electron microscope. J Mol Biol. 233(1):16-24.

Page 63: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 61

Dam M, Gerdes K. (1994) Partitioning of plasmid R1. Ten direct repeats flanking the parA promoter constitute a centromere-like partition site parC, that expresses incompatibility. J Mol Biol. 236(5):1289-98.

Dasgupta S, Maisnier-Patin S, Nordström K. (2000) New genes with old modus operandi. The connection between supercoiling and partitioning of DNA in Escherichia coli. EMBO Rep. 1(4):323-7. Review.

DiNardo S, Voelkel KA, Sternglanz R, Reynolds AE, Wright A. (1982) Escherichia coli DNA topoisomerase I mutants have compensatory mutations in DNA gyrase genes. Cell. 31(1):43-51.

Dionisio F, Matic I, Radman M, Rodrigues OR, Taddei F. (2002)Plasmids spread very fast in heterogeneous bacterial communities. Genetics. 162(4):1525-32.

Donachie W. D. (1968) Relationship between cell size and time of initiation of DNA replication. Nature. 219:1077-1079

Ehrenberg M, Sverredal A. (1995) A model for copy number control of the plasmid R1. J Mol Biol. Mar 3;246(4):472-85.

Ehrenberg M. (1996) Hypothesis: hypersensitive plasmid copy number control for ColE1. Biophys J. 70(1):135-45.

Eliasson A, Bernander R, Dasgupta S, Nordström K. (1992) Direct visualisation of plasmid DNA in bacterial cells. Mol Microbiol. 6(2):165-70.

Eliasson A, Bernander R, Nordström K. (1996) Random initiation of replication of plasmids P1 and F (oriS) when integrated into the Escherichia coli chromosome. Mol Microbiol. 20(5):1025-32.

Eliasson A, Nordström K. (1997) Replication of minichromosomes in a host in which chromosome replication is random. Mol Microbiol. 23(6):1215-20.

Fang L, Davey MJ, O'Donnell M(1999) Replisome assembly at oriC, the replication origin of E. coli, reveals an explanation for initiation sites outside an origin. Mol Cell. 4(4):541-53.

Fuller RS, Funnell BE, Kornberg A. (1984) The DnaA protein complex with the E. coli chromosomal replication origin (oriC) and other DNA sites. Cell. 38(3):889-900.

Garner J, Crooke E.(1996) Membrane regulation of the chromosomal replication activity of E. coli DnaA requires a discrete site on the protein. EMBO J. 15(13):3477-85.

Garner J, Durrer P, Kitchen J, Brunner J, Crooke E. (1998) Membrane-mediated release of nucleotide from an initiator of chromosomal replication, Escherichia coli DnaA, occurs with insertion of a distinct region of the protein into the lipid bilayer. J Biol Chem. 273(9):5167-73.

Gerdes K, Bech FW, Jorgensen ST, Lobner-Olesen A, Rasmussen PB, Atlung T, Boe L, Karlstrom O, Molin S, von Meyenburg K.(1986a) Mechanism of postsegregational killing by the hok gene product of the parB system of plasmid R1 and its homology with the relF gene product of the E. coli relB operon. EMBO J. 5(8):2023-9.

Gerdes K, Rasmussen PB, Molin S. (1986b) Unique type of plasmid maintenance function: postsegregational killing of plasmid-free cells. Proc Natl Acad Sci U S A.;83(10):3116-20.

Griffith F. (1928) The significance of pneumococcal types. J. Hyg., 27:113-159 Gustafsson P, Nordström K, Perram JW. (1978) Selection and timing of replication of plasmids R1drd-19

and F'lac in Escherichia coli. Plasmid. 1(2):187-203. Gustafsson P, Nordström K. (1980) Control of plasmid R1 replication: kinetics of replication in shifts

between different copy number levels. J Bacteriol. 141(1):106-10. Hansen FG, Koefoed S, Sorensen L, Atlung T. (1987) Titration of DnaA protein by oriC DnaA-boxes

increases dnaA gene expression in Escherichia coli. EMBO J. 6(1):255-8. Hansen FG, Christensen BB, Atlung T. (1991) The initiator titration model: computer simulation of

chromosome and minichromosome control. Res Microbiol. 142(2-3):161-7. Hiraga S, Ichinose C, Niki H, Yamazoe M. (1998) Cell cycle-dependent duplication and bidirectional

migration of SeqA-associated DNA-protein complexes in E. coli. Mol Cell. 1(3):381-7. Hu KH, Liu E, Dean K, Gingras M, DeGraff W, Trun NJ. (1996) Overproduction of three genes leads to

camphor resistance and chromosome condensation in Escherichia coli. Genetics. 143(4):1521-32. Hwang DS, Kaguni JM. (1988) Interaction of dnaA46 protein with a stimulatory protein in replication from

the Escherichia coli chromosomal origin. J Biol Chem. 263(22):10633-40. Jacob F, Brenner S and Cuzin F. (1963) On the regulation DNA replication in Bacteria. Cold Spring

Harbor Symp Quant Biol 28: 329-348. Jensen RB, Lurz R, Gerdes K. (1998) Mechanism of DNA segregation in prokaryotes: replicon pairing by

parC of plasmid R1. Proc Natl Acad Sci U S A. 95(15):8550-5. Jensen RB, Gerdes K. (1999) Mechanism of DNA segregation in prokaryotes: ParM partitioning protein of

plasmid R1 co-localizes with its replicon during the cell cycle. EMBO J. 18(14):4076-84. Katayama T, Kubota T, Kurokawa K, Crooke E, Sekimizu K. (1998) The initiator function of DnaA protein

is negatively regulated by the sliding clamp of the E. coli chromosomal replicase. Cell. 94(1):61-71.

Page 64: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

62 Jan Olsson

Katayama T. (2001a) Feedback controls restrain the initiation of Escherichia coli chromosomal replication. Mol Microbiol. 41(1):9-17. Review.

Katayama T, Fujimitsu K, Ogawa T. (2001b) Multiple pathways regulating DnaA function in Escherichia coli: distinct roles for DnaA titration by the datA locus and the regulatory inactivation of DnaA. Biochimie. 83(1):13-7.

Kitagawa R, Mitsuki H, Okazaki T, Ogawa T. (1996) A novel DnaA protein-binding site at 94.7 min on the Escherichia coli chromosome. Mol Microbiol. (5):1137-47.

Kitagawa R, Ozaki T, Moriya S, Ogawa T. (1998) Negative control of replication initiation by a novel chromosomal locus exhibiting exceptional affinity for Escherichia coli DnaA protein. Genes Dev. 12(19):3032-43.

Kolb FA, Malmgren C, Westhof E, Ehresmann C, Ehresmann B, Wagner EG, Romby P. (2000) An unusual structure formed by antisense-target RNA binding involves an extended kissing complex with a four-way junction and a side-by-side helical alignment. RNA. 6(3):311-24.

Kollek R, Oertel W, Goebel W. (1978) Isolation and characterization of the minimal fragment required for autonomous replication ("basic replicon") of a copy mutant (pKN102) of the antibiotic resistance factor R1. Mol Gen Genet. 162(1):51-7.

Koppes L, Nordström K. (1986) Insertion of an R1 plasmid into the origin of replication of the E. coli chromosome: random timing of replication of the hybrid chromosome. Cell. 44(1):117-24.

Koppes LJ, von Meyenburg K. (1987) Nonrandom minichromosome replication in Escherichia coli K-12. J Bacteriol. 169(1):430-3.

Krabbe M, Zabielski J, Bernander R, Nordström K. (1997) Inactivation of the replication-termination system affects the replication mode and causes unstable maintenance of plasmid R1. Mol Microbiol. 24(4):723-35.

Kücherer C, Lother H, Kolling R, Schauzu MA, Messer W. (1986) Regulation of transcription of the chromosomal dnaA gene of Escherichia coli. Mol Gen Genet. 205(1):115-21

Landoulsi A, Malki A, Kern R, Kohiyama M, Hughes P. (1990) The E. coli cell surface specifically prevents the initiation of DNA replication at oriC on hemimethylated DNA templates. Cell. 63(5):1053-60.

Langer U, Richter S, Roth A, Weigel C, Messer W. (1996) A comprehensive set of DnaA-box mutations in the replication origin, oriC, of Escherichia coli. Mol Microbiol. 21(2):301-11.

Larsen JE, Gerdes K, Light J, Molin S. (1984) Low-copy-number plasmid-cloning vectors amplifiable by derepression of an inserted foreign promoter. Gene. 28(1):45-54.

Lederberg J. (1952) Cell genetics and hereditary symbiosis. Physiol. Rev., 32:403-430 Leonard AC, Helmstetter CE. (1986) Cell cycle-specific replication of Escherichia coli

minichromosomes. Proc Natl Acad Sci U S A. 83(14):5101-5. Light J, Molin S. (1981) Replication control functions of plasmid R1 act as inhibitors of expression of a

gene required for replication. Mol Gen Genet. 184(1):56-61. Light J, Molin S. (1982a) Expression of a copy number control gene (copB) of plasmid R1 is constitutive

and growth rate dependent. J Bacteriol. 151(3):1129-35. Light J, Molin S. (1982b) The sites of action of the two copy number control functions of plasmid R1. Mol

Gen Genet. 187(3):486-93. Light J, Riise E, Molin S. (1985) Transcription and its regulation in the basic replicon region of plasmid

R1. Mol Gen Genet. 198(3):503-8. Løbner-Olesen A, Skarstad K, Hansen FG, von Meyenburg K, Boye E. (1989) The DnaA protein

determines the initiation mass of Escherichia coli K-12. Cell. 57(5):881-9. Løbner-Olesen A. (1999) Distribution of minichromosomes in individual Escherichia coli cells:

implications for replication control. EMBO J. 18(6):1712-21. Lu M, Campbell JL, Boye E, Kleckner N. (1994) SeqA: a negative modulator of replication initiation in E.

coli. Cell. 77(3):413-26. Lu M, Kleckner N. (1994) Molecular cloning and characterization of the pgm gene encoding

phosphoglucomutase of Escherichia coli. J Bacteriol. 176(18):5847-51. Malmgren C, Engdahl HM, Romby P, Wagner EG. (1996) An antisense/target RNA duplex or a strong

intramolecular RNA structure 5' of a translation initiation signal blocks ribosome binding: the case of plasmid R1.RNA. 2(10):1022-32.

Marsh RC, Worcel A. (1977) A DNA fragment containing the origin of replication of the Escherichia coli chromosome. Proc Natl Acad Sci U S A. 74(7):2720-4.

Masai H, Kaziro Y, Arai K. (1983) Definition of oriR, the minimum DNA segment essential for initiation of R1 plasmid replication in vitro. Proc Natl Acad Sci U S A.;80(22):6814-8.

Masai H, Arai K. (1987) RepA and DnaA proteins are required for initiation of R1 plasmid replication in vitro and interact with the oriR sequence. Proc Natl Acad Sci U S A. 84(14):4781-5

Matsui M, Oka A, Takanami M, Yasuda S, Hirota Y. (1985) Sites of DnaA protein-binding in the replication origin of the Escherichia coli K-12 chromosome. J Mol Biol. 184(3):529-33.

Page 65: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 63

McAdams HH, Arkin A. (1997) Stochastic mechanisms in gene expression. Proc Natl Acad Sci U S A. 94(3):814-9.

Meijer M, Beck E, Hansen FG, Bergmans HE, Messer W, von Meyenburg K, Schaller H. (1979) Nucleotide sequence of the origin of replication of the Escherichia coli K-12 chromosome. Proc Natl Acad Sci U S A. 76(2):580-4.

Menzel R, Gellert M. (1983) Regulation of the genes for E. coli DNA gyrase: homeostatic control of DNA supercoiling. Cell. 34(1):105-13.

Meselson M and Stahl FW. (1958) The replication of DNA in Escherichia coli. Proc Natl Acad Sci U S A. 44:671-682

Messer W, Bergmans HE, Meijer M, Womack JE, Hansen FG, von Meyenburg K. (1978) Mini-chromosomes: plasmids which carry the E. coli replication origin. Mol Gen Genet. 4;162(3):269-75.

Messer W, Meijer M, Bergmans HE, Hansen FG, von Meyenburg K, Beck E, Schaller H. (1979) Origin of replication, oriC, of the Escherichia coli K12 chromosome: nucleotide sequence. Cold Spring Harb Symp Quant Biol. 43 Pt 1:139-45.

Messer W, Bellekes U, Lother H. (1985) Effect of dam methylation on the activity of the E. coli replication origin, oriC. EMBO J. 4(5):1327-32.

Messer W and Weigel C (1996) Initiation of chromosome replication. In Escherichia coli and Salmonella: Cellular and molecular Biology, 2nd edn, Vol2. Neidhardt FC, Cutriss R, III, Ingraham JL, Lin ECC, Low KB and Magasanik B et al. (eds). Washington, DC: American Society for Microbiology Press, pp. 1579-1601

Messer W, Blaesing F, Jakimowicz D, Krause M, Majka J, Nardmann J, Schaper S, Seitz H, Speck C, Weigel C, Wegrzyn G, Welzeck M, Zakrzewska-Czerwinska J. (2001) Bacterial replication initiator DnaA. Rules for DnaA binding and roles of DnaA in origin unwinding and helicase loading. Biochimie. 83(1):5-12. Review.

Meynell E, Datta N. (1967) Mutant drug resistant factors of high transmissibility. Nature. 214(91):885-7. Min YN, Tabuchi A, Womble DD, Rownd RH. (1991) Transcription of the stability operon of IncFII

plasmid NR1. J Bacteriol. 173(7):2378-84. Molin S, Stougaard P, Uhlin BE, Gustafsson P, Nordström K. (1979)Clustering of genes involved in

replication, copy number control, incompatibility, and stable maintenance of the resistance plasmid R1drd-19. J Bacteriol. 138(1):70-9.

Møller-Jensen J, Jensen RB, Lowe J, Gerdes K. (2002) Prokaryotic DNA segregation by an actin-like filament. EMBO J. 21(12):3119-27.

Neidhardt F.C., Ingraham J.L. and Schaecter M., (1990) Physiology of the Bacterial Cell. Sinauer Associates, Inc. Publishers Sunderland, Massachusettes

Neuwald AF, Aravind L, Spouge JL, Koonin EV. (1999) AAA+: A class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res.9(1):27-43.

Newman G, Crooke E. (2000) DnaA, the initiator of Escherichia coli chromosomal replication, is located at the cell membrane. J Bacteriol. 182(9):2604-10

Niki H, Jaffe A, Imamura R, Ogura T, Hiraga S. (1991) The new gene mukB codes for a 177 kd protein with coiled-coil domains involved in chromosome partitioning of E. coli. EMBO J. 10(1):183-93.

Nordström K, Molin S, Aagaard-Hansen H. (1980) Partitioning of plasmid R1 in Escherichia coli. I. Kinetics of loss of plasmid derivatives deleted of the par region. Plasmid. 4(2):215-27.

Nordström K, Aagaard-Hansen H. (1984) Maintenance of bacterial plasmids: comparison of theoretical calculations and experiments with plasmid R1 in Escherichia coli. Mol Gen Genet. 197(1):1-7.

Nordström K, Molin S, Light J. (1984) Control of replication of bacterial plasmids: genetics, molecular biology, and physiology of the plasmid R1 system. Plasmid. 12(2):71-90. Review.

Nordström K. (1990) Control of Plasmid replication--how Do DNA Iterons Set the Replication Frequency? Cell 63: 1121.

Nordström, K. And Wagner, E.G.H. (1994) Kinetic aspects of control of plasmid replication by antisense RNA. Trends Biochem Sci 19:294-300.

Nordström K, Uhlin BE. (1992) Runaway-replication plasmids as tools to produce large quantities of proteins from cloned genes in bacteria. Biotechnology (N Y). 10(6):661-6. Review

Ogden GB, Pratt MJ, Schaechter M. (1988) The replicative origin of the E. coli chromosome binds to cell membranes only when hemimethylated. Cell. 54(1):127-35.

Oka A, Sugimoto K, Takanami M, Hirota Y. (1980) Replication origin of the Escherichia coli K-12 chromosome: the size and structure of the minimum DNA segment carrying the information for autonomous replication. Mol Gen Genet. 178(1):9-20.

Olsson, J., Dasgupta, S., Berg, OG., Nordström, K. (2002) Eclipse period without sequestration in Escherichia coli. Mol Microbiol 44(6):1429-40

Page 66: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

64 Jan Olsson

Onogi T, Niki H, Yamazoe M, Hiraga S. (1999) The assembly and migration of SeqA-Gfp fusion in living cells of Escherichia coli. Mol Microbiol. 31(6):1775-82.

Paulsson J, Ehrenberg M. (1998) Trade-off between segregational stability and metabolic burden: a mathematical model of plasmid ColE1 replication control. J Mol Biol. 279(1):73-88.

Paulsson J, Ehrenberg M. (2000) Molecular clocks reduce plasmid loss rates: the R1 case. J Mol Biol. 297(1):179-92.

Pogliano J. (2002) Dynamic cellular location of bacterial plasmids. Curr Opin Microbiol. Dec;5(6):586-90. Polaczek P, Wright A. (1990) Regulation of expression of the dnaA gene in Escherichia coli: role of the

two promoters and the DnaA box. New Biol. 2(6):574-82. Proctor GN. (1994) Mathematics of microbial plasmid instability and subsequent differential growth of

plasmid-free and plasmid-containing cells, relevant to the analysis of experimental colony number data. Plasmid. 32(2):101-30. Review.

Rasmussen PB, Gerdes K, Molin S. (1987)Genetic analysis of the parB+ locus of plasmid R1. Mol Gen Genet. 209(1):122-8.

Riise E, Stougaard P, Bindslev B, Nordström K, Molin S. (1982) Molecular cloning and functional characterisation of a copy number control gene (copB) of plasmid R1. J Bacteriol. 151(3):1136-45.

Riise E, Molin S. (1986) Purification and characterisation of the CopB replication control protein, and precise mapping of its target site in the R1 plasmid. Plasmid. 15(3):163-71

Roth A, Messer W. (1995) The DNA binding domain of the initiator protein DnaA. EMBO J. 14(9):2106-11

Ruiz-Echevarria MJ, Berzal-Herranz A, Gerdes K, Diaz-Orejas R. (1991) The kis and kid genes of the parD maintenance system of plasmid R1 form an operon that is autoregulated at the level of transcription by the co-ordinated action of the Kis and Kid proteins. 5(11):2685-93.

Russell DW, Zinder ND. (1987) Hemimethylation prevents DNA replication in E. coli. Cell. 50(7):1071-9. Schaper S, Messer W. (1995) Interaction of the initiator protein DnaA of Escherichia coli with its DNA

target. J Biol Chem. 270(29):17622-6. Sekimizu K, Bramhill D, Kornberg A. (1987) ATP activates DnaA protein in initiating replication of

plasmids bearing the origin of the E. coli chromosome. Cell. 50(2):259-65. Sekimizu K, Kornberg A. (1988) Cardiolipin activation of DnaA protein, the initiation protein of

replication in Escherichia coli. J Biol Chem. 263(15):7131-5 Sekimizu K, Yung BY, Kornberg A. (1988) The dnaA protein of Escherichia coli. Abundance, improved

purification, and membrane binding. J Biol Chem. 263(15):7136-40. Shakibai N, Ishidate K, Reshetnyak E, Gunji S, Kohiyama M, Rothfield L. (1998) High-affinity binding of

hemimethylated oriC by Escherichia coli membranes is mediated by a multiprotein system that includes SeqA and a newly identified factor, SeqB. Proc Natl Acad Sci U S A. 95(19):11117-21.

Skarstad K, Steen HB, Boye E. (1983) Cell cycle parameters of slowly growing Escherichia coli B/r studied by flow cytometry. J Bacteriol. 154(2):656-62.

Skarstad K, Steen HB, Boye E. (1985) Escherichia coli DNA distributions measured by flow cytometry and compared with theoretical computer simulations.J Bacteriol. 163(2):661-8.

Skarstad K, Boye E, Steen HB. (1986) Timing of initiation of chromosome replication in individual Escherichia coli cells. EMBO J. 5(7):1711-7.

Skarstad K, Boye E. (1988) Perturbed chromosomal replication in recA mutants of Escherichia coli. J Bacteriol. 170(6):2549-54.

Skarstad K, von Meyenburg K, Hansen FG, Boye E. (1988) Co-ordination of chromosome replication initiation in Escherichia coli: effects of different dnaA alleles. J Bacteriol. 170(2):852-8.

Skarstad K, Løbner-Olesen A, Atlung T, von Meyenburg K, Boye E. (1989) Initiation of DNA replication in Escherichia coli after overproduction of the DnaA protein. Mol Gen Genet. 218(1):50-6.

Skarstad K, Boye E. (1994) The initiator protein DnaA: evolution, properties and function. Biochim Biophys Acta. 1217(2):111-30. Review.

Skarstad, K., Bernander, R., Wold, S., Steen, H.B., and Boye, E. (1995a) in “Flow cytometry Applications

in Cell Culture” Al-Rubeai, M., and. Emery, A. N. eds. Mercel Dekker, New York. pp. 241-255. Skarstad, K., Bernander, R., and Boye, E. (1995b) Analysis of DNA replication in vivo by flow cytometry.

Methods in Enzymology 362: 604-613. Skarstad K, Lueder G, Lurz R, Speck C, Messer W. (2000) The Escherichia coli SeqA protein binds

specifically and co-operatively to two sites in hemimethylated and fully methylated oriC. Mol Microbiol. 36(6):1319-26.

Slater S, Wold S, Lu M, Boye E, Skarstad K, Kleckner N. (1995) E. coli SeqA protein binds oriC in two different methyl-modulated reactions appropriate to its roles in DNA replication initiation and origin sequestration. Cell. 82(6):927-36.

Page 67: Control of Chromosome and Plasmid Replication in ...162964/FULLTEXT01.pdf · DNA molecules duplicate by semi-conservative replication in which each new DNA molecule inherits one strand

Control of chromosome and plasmid replication in Escherichia coli 65

Smith DW, Garland AM, Herman G, Enns RE, Baker TA, Zyskind JW. (1985) Importance of state of methylation of oriC GATC sites in initiation of DNA replication in Escherichia coli. EMBO J. 4(5):1319-26.

Smith DW, Stine WB, Svitil AL, Bakker A, Zyskind JW. (1992) Escherichia coli cells lacking methylation-blocking factor (leucine-responsive regulatory protein) have precise timing of initiation of DNA replication in the cell cycle. J Bacteriol. 174(9):3078-82.

Söderbom F, Binnie U, Masters M, Wagner EG. (1997) Regulation of plasmid R1 replication: PcnB and RNase E expedite the decay of the antisense RNA, CopA. Mol Microbiol. 26(3):493-504.

Söderbom F, Wagner EG. (1998) Degradation pathway of CopA, the antisense RNA that controls replication of plasmid R1. Microbiology. 144 ( Pt 7):1907-17.

Speck C, Weigel C, Messer W. (1999) ATP- and ADP-dnaA protein, a molecular switch in gene regulation. EMBO J. 18(21):6169-76.

Sternglanz R, DiNardo S, Voelkel KA, Nishimura Y, Hirota Y, Becherer K, Zumstein L, Wang JC. (1981) Mutations in the gene coding for Escherichia coli DNA topoisomerase I affect transcription and transposition. Proc Natl Acad Sci U S A. 78(5):2747-51.

Stougaard P, Molin S, Nordström K, Hansen FG. (1981) The nucleotide sequence of the replication control region of the resistance plasmid R1drd-19. Mol Gen Genet. 181(1):116-22.

Stougaard P, Light J, Molin S (1982). Convergent transcription interferes wit expression of the copy number control gene, copA, from plasmid R1. EMBO J. 1, 323-328.

Suzuki E, Kondo T, Makise M, Mima S, Sakamoto K, Tsuchiya T, Mizushima T. (1998) Alteration in levels of unsaturated fatty acids in mutants of Escherichia coli defective in DNA replication. Biol Pharm Bull. 21(7):657-61.

Torheim NK, Skarstad K. (1999) Escherichia coli SeqA protein affects DNA topology and inhibits open complex formation at oriC. EMBO J. 18(17):4882-8.

Uhlin BE, Nordström K. (1985) Preferential inhibition of plasmid replication in vivo by altered DNA gyrase activity in Escherichia coli. J Bacteriol. 162(2):855-7.

von Freiesleben U, Krekling MA, Hansen FG, Løbner-Olesen A. (2000) The eclipse period of Escherichia coli. EMBO J. 15;19(22):6240-8.

Wang JC. (1971) Interaction between DNA and an Escherichia coli protein omega. J Mol Biol. 55(3):523-33.

Weitao T, Nordström K, Dasgupta S. (1999) Mutual suppression of mukB and seqA phenotypes might arise from their opposing influences on the Escherichia coli nucleoid structure. Mol Microbiol. 34(1):157-68.

Weitao T, Dasgupta S, Nordström K. (2000a) Role of the mukB gene in chromosome and plasmid partition in Escherichia coli. Mol Microbiol. 38(2):392-400.

Weitao T, Nordström K, Dasgupta S. (2000b) Escherichia coli cell cycle control genes affect chromosome superhelicity. EMBO Rep. 1(6):494-9.

Weigel C, Schmidt A, Ruckert B, Lurz R, Messer W. (1997) DnaA protein binding to individual DnaA boxes in the Escherichia coli replication origin, oriC. EMBO J. 16(21):6574-83.

Wold S, Boye E, Slater S, Kleckner N, Skarstad K. (1998) Effects of purified SeqA protein on oriC-dependent DNA replication in vitro. EMBO J. 17(14):4158-65.

Womble DD, Sampathkumar P, Easton AM, Luckow VA, Rownd RH. (1985) Transcription of the replication control region of the IncFII R-plasmid NR1 in vitro and in vivo. J Mol Biol. 181(3):395-410

Womble DD, Rownd RH. (1986) Regulation of IncFII plasmid DNA replication. A quantitative model for control of plasmid NR1 replication in the bacterial cell division cycle. J Mol Biol. 192(3):529-47

Wouters JT, Driehuis FL, Polaczek PJ, van Oppenraay ML, van Andel JG. (1980) Persistence of the pBR 322 plasmid in Escherichia coli K 12 grown in chemostat cultures. Antonie Van Leeuwenhoek. 46(4):353-62.

Xia W, Dowhan W. (1995) In vivo evidence for the involvement of anionic phospholipids in initiation of DNA replication in Escherichia coli. Proc Natl Acad Sci U S A. 92(3):783-7.

Zheng W, Li Z, Skarstad K, Crooke E. (2001) Mutations in DnaA protein suppress the growth arrest of acidic phospholipid-deficient Escherichia coli cells. EMBO J. 20(5):1164-72.

Zyskind JW, Smith DW. (1986) The bacterial origin of replication, oriC.Cell. Aug 15;46(4):489-90. Öhman M, Wagner EG. (1991) Regulation of replication of plasmid R1: an analysis of the intergenic region

between copA and repA. Mol Gen Genet. Er 230(1-2):321-8.