Tải bản đầy đủ - 0 (trang)
2 Plasmid Segregation: Four Types of Modules

2 Plasmid Segregation: Four Types of Modules

Tải bản đầy đủ - 0trang

4  Extrachromosomal Components of the Nucleoid


segrosome, the nucleoprotein complex that directs plasmid partition (Hayes and

Barillà 2006a, b), which has provided key insights into the process. Segregation

cassettes have been classified into four types (I–IV) based on the genetic organization of the module and the evolutionary relationships among the encoded proteins

(Fig. 4.1) (Ebersbach and Gerdes 2005a; Hayes and Barillà 2006a, b; Schumacher

2007, 2008). The type I partition system is the most prevalent and, along with the

type II complex, has been most extensively studied. The most well-characterized

Segrosome assembly

Segrosome assembly

Transcriptional autoregulation

Transcriptional autoregulation


Segrosome assembly


Transcriptional autoregulation


Centromere binding

and autoregulation?



Fig.  4.1  Genetic organization and interactions in plasmid segregation cassettes. The widelydisseminated Type I partition modules possess genes for a ParA Walker-type ATPase and a CBF.

The genes for parA and the CBF, and the corresponding proteins are shown by open and filled

symbols, respectively. The centromere located either upstream or downstream of the genes is indicated

by a hatched box. The operator site upstream of the genes is shown schematically by small, inverted

arrows. During segrosome formation, the CBF binds to the centromere and recruits the ParA factor

which itself does not bind centromeric DNA. In some cases, the ParA protein is the principal

transcriptional autoregulator of the par operon, whereas the ParB CBF is a co-repressor that does

not interact directly with the operator site (top left). In other cases, the CBF binds to both the centromere

and regulatory region (top right). In these cases, the ATPase is recruited to the segrosome via proteinprotein interactions with the CBF, but does not participate in transcriptional autoregulation. In the

well-characterized type II module, a pair of genes specify an actin-like ATPase (ParM; shaded

ovoid), and a small DNA binding protein (ParR). The centromere and operator site located upstream

of the genes overlap. ParR is both a CBF, as well as a transcriptional autorepressor. ParM interacts

with ParR at the centromere, but is not involved in gene regulation. Type III partition cassettes are

found on certain large plasmids of the Bacillus cereus group and comprise two genes. The first gene

encodes the RepR protein that is a likely transcriptional autorepressor, whereas the following gene

(grey arrow) encodes a member of the tubulin family known as TubZ or RepX. Type IV segregation modules include a single gene (hatched arrow) whose product is predicted to possess both

DNA binding properties, as well a region of coiled-coil structure


F. Hayes and D. Barillà

examples of these two types are discussed here first, with a subsequent brief

description of type III and IV classes.

Type I cassettes are the most widespread segregation modules, and are found

very commonly on plasmids in a wide diversity of eubacteria, as well as in archaea.

The cassettes are disparate in composition and organization, but are unified by the

presence of genes specifying Walker-type ATPases that have a close evolutionary

origin (Fig.  4.2). These motor-like ATPases, commonly denoted ParA, possess

divergent Walker-like ATP binding motifs (Koonin 1993; Motallebi-Veshareh et al.

1990). The ParA proteins range in size from ~200 amino acids to ~450 amino acids.

Type I modules invariably contain a second gene downstream of, and transcribed in

the same direction as, the gene for ParA. The encoded product is a DNA binding

factor that recognizes the cis-acting centromere sequence located within the cassette, forming a pre-segrosome complex. These centromere binding factors (CBFs)

are more diverse than the ParA components, both in size and sequence (Fothergill

et al. 2005; Hayes and Barillà 2006a). However, one subset of CBFs, exemplified

by ParB of plasmid P1, is particularly common and is frequently specified by type

I cassettes that encode large ParA proteins. Many bacterial chromosomes also harbour

parAB cassettes (Bartosik and Jagura-Burdzy 2005; Thanbichler and Shapiro 2006).

ParA ( S. coelicolor)


IncC (RK2)


ParA ( C. crescentus)

Soj ( B. subtilis)

SopA (F)

Soj ( P. putida)

ParA1 ( V. cholerae)

ParA2 ( V. cholerae)

RepB (pAD1)

ParA (P7)


ParA ( C. difficile)

ParA (pVEF2)

ParA (P1)

PrgP (pGENT)

ParA (pB171)

ParA (pCF10)

ParA ( L. casei )

ParA (pTAR)

ParA (pVS1)


Rep63B (pNVH0597_60)

ParF (TP228)


Fig.  4.2  Phylogenetic tree of selected ParA family members. Plasmid-encoded proteins only are

shown, except for the Soj group of chromosomally-encoded homologues, and the ParA and ParA2

proteins specified by the S. coelicolor chromosome and chromosome II of V. cholerae, respectively

4  Extrachromosomal Components of the Nucleoid


By contrast, plasmid segregation modules that possess the shortest parA loci

include genes for CBFs that may have very little, if any, amino acid sequence similarity and which preferentially recognize the cognate centromeric sites (Fothergill

et al. 2005). Nevertheless, two of these disparate CBFs, ParG and w of plasmids

TP228 and pSM19035, respectively, are homodimers that possess ribbon-helixhelix (RHH) folds which bind DNA (Golovanov et  al. 2003; Murayama et  al.

2001), hinting that common structural motifs may be employed for centromere

recognition among smaller CBFs within the type I class.

Type II modules, typified by the partition cassette of plasmid R1, have a similar

genetic organization to type I operons (Fig.  4.1). Type II modules comprise two

genes, one of which encodes an ATPase (ParM) and the second of which specifies

a CBF (ParR). The centromere (parC) is situated upstream of the genes (Salje and

Lowe 2008). However, by contrast with type I cassettes, the motor-like ATPase

specified by type II systems is an ancestral homologue of eukaryotic actin (Bork

et al. 1992), and is unrelated evolutionarily to ParA proteins. A series of elegant

studies, described further in Section 4.5, have illuminated core aspects of the type

II partitioning mechanism.

Both type I and II modules are transcriptionally autoregulated (Fig. 4.1). Large

ParA homologues repress transcription via an N-terminal helix-turn-helix (HTH)

motif that binds to an operator site located 5¢ of the partition cassette. The CBF is

a co-repressor that decreases operon expression further by an unknown mechanism.

Autoregulation of the parAB operon of the P1 plasmid arguably is the most wellstudied example among segregation cassettes (Davey and Funnell 1994; Friedman

and Austin 1988; Hayes et al. 1994; Radnedge et al. 1998). The nucleotide bound

state of ParA modulates transcriptional regulation: ADP, but not ATP, promotes

ParA repression of the operon (Bouet and Funnell 1999). ADP may induce a conformational change in the protein so that it is more proficient than nucleotide-free ParA

in DNA binding. Instead, the ATP-bound form of ParA is required for segregation

(Davis et al. 1996). Most plausibly, ATP fulfils an analogous function in P1 segregation as it does in other type I complexes, by promoting ParA polymerization (see

Section 4.4). By contrast with P1 ParA and its immediate homologues, both short

ParA type I proteins and the ParM type II ATPase lack DNA-binding motifs and are

not implicated in transcriptional regulation of their operons. Instead, the CBFs

alone are transcriptional repressors of the operons, as well as integral components

of the segrosome (Fig. 4.1) (e.g., Carmelo et al. 2005; Dmowski et al. 2006; Jensen

et al. 1994; Ringaard et al. 2007a; Zampini et al. 2009).

4.3 Centromere Binding Factors: Underpinning the Segrosome

Centromeres in higher eukaryotes are DNA regions that direct kinetochore formation

and the cohesion of sister chromatids. The structural features of eukaryotic centromeres, which in some species extend for many thousands of kilobases, continue

to defy full understanding. Like centromeres in eukaryotes, plasmid centromeres


F. Hayes and D. Barillà

are remarkably diverse in sequence, in length, and in the numbers, orientation and

spacing of repeated sequences that they contain (Hayes and Barillà 2006a)

(Fig. 4.3). Nevertheless, plasmid centromeres typically are ~100–200 bp in length

P1 parS

10 bp

pTAR parS

pB171 parC1

R1 parC

pGENT cenE

pSM19035 P


F sopC

100 bp

Fig. 4.3  Plasmid centromere organization. Top, the orientation of repeat motifs in selected centromeres is shown by arrows. The lengths and sequences of these motifs differ between different

centromeres. Additionally, the P1 parS and pB171 parC1 sites each possess two different repeat

sequences highlighted by different arrow shadings. The binding site for IHF in P1 parS is denoted

by the rectangle. The Pd region is one of three loci on pSM19035 that possess centromere activity.

Similarly, a second set of repeat motifs in pB171 is implicated in centromere function. A 10-bp

scale bar is shown, except for F sopC for which a 100-bp scale is shown. Bottom, schematic representation of co-crystal structures of ParR dimers (oval) of plasmid pSK41 bound to one repeat

motif (arrow) from its cognate centromere (Schumacher et al. 2007a). A pair of ParRSK41 dimers

binds to a single repeat. Interactions between dimers bound to separate DNA fragments mediate

the assembly of a nucleoprotein superstructure containing a ‘pseudo-centromere’-like sequence

wrapped around a ParRSK41 protein core. The ParR protein of pSK41 is homologous to the ParR

proteins encoded by plasmids R1 and pB171. Dimers of the latter also coalesce into ring-like

structures both in the absence and presence of centromere DNA (Moller-Jensen et al. 2007)

4  Extrachromosomal Components of the Nucleoid


(Fig.  4.3), and occupy defined locations either upstream or downstream of the

partition genes (Fig. 4.1). Despite their diversity, centromeres uniformly provide a

scaffold for loading of the cognate CBF (Fig.  4.4a). Elucidation of the tertiary

structures of a number of CBFs, in some instances complexed with centromeric

DNA within the pre-segrosomal complex, recently have begun to provide fascinating

glimpses into assembly of the mature segrosome (de la Hoz et al. 2004; Delbrück

et al. 2002; Golovanov et al. 2003; Khare et al. 2004; Moller-Jensen et al. 2007;

Murayama et al. 2001; Schumacher 2007, 2008; Schumacher et al. 2007b).

The parS site of P1 is among the most well-studied plasmid centromeres. The

site comprises two sets of repeat motifs separated by a central region that is bound

by the DNA bending protein, integration host factor (IHF). The repeats are of two

distinct types, either Box A (heptameric) or Box B (hexameric) (Fig.  4.3). The

dimeric ParB CBF recognizes the Box A and Box B motifs using separate protein

domains. Amino terminal domains of ParB recognize the Box A repeats via HTH

motifs, whereas a dimerized DNA-binding module composed of a six-stranded

b-sheet coiled-coil interacts with the hexamer boxes near to the ends of the site

(Schumacher and Funnell 2005; Vecchiarelli et al. 2007). A HTH motif has also

been implicated in DNA binding by the ParB homologue, KorB, of plasmid RK2

(Khare et al. 2004). The DNA-binding domains within ParB are connected by flexible linkers around which the domains can swivel freely to permit multiple arrangements of ParB-parS contacts. Indeed, the ParB dimer simultaneously can make in

trans contacts with parS sites on different DNA molecules (Schumacher and

Funnell 2005; Schumacher et al. 2007a), supporting observations that ParB can pair

plasmids that harbour the site (Edgar et al. 2001). The role of IHF is entirely architectural, serving to bend the arms of parS to accommodate ParB dimers that span

the two arms (Funnell 1988, 1991; Hayes and Austin 1994). The parS locus is a

nucleation point for the spreading of ParB many kilobases away from the site

(Rodionov et al. 1999), as is the sopC centromere for the ParB homologue encoded

by the F plasmid (Lynch and Wang 1995). It is tempting to speculate that this

reflects the formation of a nucleoprotein superstructure in the vicinity of the centromere, but the purpose, if any, of this spreading during segregation remains unresolved (Rodionov and Yarmolinsky 2004).

Variations in the Box B repeats in the parS site and corresponding substitutions

in the dimerized DNA-binding module of ParB provide species specificity among

segrosomes that are closely-related to the P1 partition complex (Hayes and Austin

1993; Hayes et al. 1993; Radnedge et al. 1996, 1998; Sergueev et al. 2005). Indeed,

a single alteration in the Box B repeats and a matching substitution in the one of

the ParB residues that contacts this position can induce a comprehensive switch in

the specificity of the interactions (Dabrazhynetskaya et al. 2005, 2009). These natural variations potentially permit plasmids with closely-related partition modules to

co-exist without cross-interference in segrosome assembly. An alternative evolutionary strategy to this end is the acquisition by plasmids of CBFs that recognize

centromeres which differ majorly in the number, arrangement and sequences of the

repeat motifs that they contain. The resulting CBF-centromere interactions are

exclusive, conferring a layer of specificity to the macromolecular interactions that

mediate plasmid segregation (Fothergill et al. 2005).


F. Hayes and D. Barillà






Fig. 4.4  (a) Atomic force microscopy of the PrgP CBF of plasmid pGENT loaded on the cenE

centromere located within a linear DNA fragment (Derome et al. 2008; M. Bussiek, A. Derome,

C. Hoischen, S. Diekmann, D. Barillà, and F. Hayes, unpublished data). PrgP recognizes two

arrays of seven TATA boxes (CESI and CESIII) separated by the CESII spacer in cenE (Fig. 4.3).

The PrgP foci at these two arrays are highlighted with arrows. (b) Electron microscopy reverse

contrast images of multistranded ParF polymers (inset) and ParF fibres assembled in the presence

of ParG (main). Bar = 210 nm (main) and 500 nm (inset)

The ParR protein of plasmid R1 binds to two arrays of five 11-bp direct repeats

in the parC centromere that sandwich a spacer region containing the parMR

promoter (Fig. 4.3) (Breüner et al. 1996; Dam and Gerdes 1994; Hoischen et al.

2008; Salje and Lowe 2008). Whereas IHF bends the parS site of the P1 plasmid, a

4  Extrachromosomal Components of the Nucleoid


combination of intrinsic curvature and ParR-induced bending severely distorts parC

into a U-shaped structure (Hoischen et  al. 2004, 2008). ParR of plasmid R1 has

proven refractory to detailed structural studies, but dimeric sequence homologues

encoded by pB171 of E. coli and staphylococcal plasmid pSK14 possess RHH

folds that mediate binding to their cognate centromeres. Intriguingly, the crystal

structure of ParRB171 is arranged into a continuous helical assembly comprising 12

dimers per 360° turn. These closed rings are also visible by electron microscopy in

the presence of centromere DNA (Moller-Jensen et al. 2007). Analogously, pairs of

ParRSK41 dimers bound to a minimal centromere subsite coalesce into an extended

looped structure held in place by protein-protein interactions (Fig. 4.3) (Schumacher

et al. 2007a). The DNA binding domains of both ParRB171 and ParRSK41 within these

ring-shaped superstructures point away from the central cavity, suggesting that the

centromeric DNA wraps around the external surface of the rings (Moller-Jensen

et al. 2007; Schumacher et al. 2007a). It has been hypothesized that the extended

ParM filament (see Section  4.5) may be tethered within the central cavity of the

ParR ring in the mature segrosome (Moller-Jensen et al. 2007).

Like yeast centromeres, the cenE site of enterococcal plasmid pGENT is

intrinsically curved, albeit less than the R1 parC site (Derome et al. 2008). The

cenE centromere comprises three subregions: the CESI and CESIII subsites are

separated by the CESII spacer (Fig.  4.3). CESI and CESIII each contain seven

TATA boxes spaced by half-helical turns. The PrgO CBF independently binds

CESI and CESIII, but with different avidities, whereas the protein does not

occupy the CESII subsite. The function of the P1 parS site is unimpaired by the

insertion of integral helical turns between the site’s arms (Hayes and Austin

1994), reflecting the relative flexibility of the DNA binding domains in the ParB

dimers that span the arms (Schumacher and Funnell 2005). However, parS

activity is abolished by the insertion of non-integral helical turns in the centre of

the site, indicating that the arms must be positioned with specific faces of the two

helices facing each other (Hayes and Austin 1994). By contrast, cenE is tolerant

to insertions of both integral and non-integral turns in the CESII subsite suggesting that the architecture of the PrgO-cenE pre-segrosome complex differs from

that of ParB-parS (Derome et al. 2008).

Like ParR, the w CBF possesses a RHH fold that binds DNA (Murayama et al.

2001). The w protein was first established as a global transcriptional repressor of

pSM19035 genes (de la Hoz et  al. 2000), before its role as a CBF was clarified

more recently (Dmowski et  al. 2006; Pratto et  al. 2008). The protein recognizes

multiple heptad motifs in its binding sites; at least three of these sites may act as

centromeric sequences (Fig. 4.3). Co-crystal structures and footprinting data of w

on these sites suggest an elongated protein superstructure in which w dimers spiral

as a left-handed helix that enwraps the centromeric DNA (de la Hoz et al. 2004;

Weihofen et al. 2006). Thus, an emerging theme in the interaction of CBFs with

their binding sites is the formation of intricate nucleoprotein superstructures that

underpin the segrosome. The specific conformation of these structures is likely to

be crucial for interaction with the motor-like ParA or ParM polymeric protein and

the assembly of the functional segrosome.


F. Hayes and D. Barillà

4.4 Plasmid-Encoded ParA Proteins are Polymerizing

Motor-Like ATPases

Although it has been recognized for some time that ParA proteins encoded by type

I modules possess Walker box ATP binding and hydrolysis motifs (Koonin 1993;

Motallebi-Veshareh et al. 1990), and that disruption of these motifs abolished plasmid

partitioning in vivo (Davis et al. 1996), the purpose of nucleotide binding and cleavage was uncertain. Was ATP hydrolysis the driving force for plasmid movement? If

so, how might the comparatively weak ATPase activity of ParA proteins provide

sufficient energy to propel plasmids through the dense cytoplasm? Recent studies

demonstrating that plasmid-encoded ParA proteins polymerize in vivo and in vitro

in response to ATP binding have begun to answer these conundrums (Adachi et al.

2006; Barillà et  al. 2005, 2007; Bouet et  al. 2007; Ebersbach and Gerdes 2001,

2005b; Machón et al. 2007; Pratto et al. 2008). The ParA homologue, ParF, encoded

by the TP228 multiresistance plasmid (Barillà and Hayes 2003; Hayes 2000) assembles into extensive multistranded filaments in response to ATP binding in  vitro.

Nucleotide hydrolysis is not required for filamentation. Moreover, mutations within

the Walker box motifs of ParF perturb polymerization, and ADP blocks polymerization, confirming the role of nucleotide binding in filamentation (Barillà et al. 2005).

ParF polymerization is modulated by the ParG CBF, both in the presence and

absence of nucleotide (Fig.  4.4b). More specifically, an unstructured N-terminal

region of ParG is required for stimulation of ParF polymerization (Barillà et  al.

2007). Although the mechanism by which the ParG flexible tail affects ParF polymerization has yet to be revealed, it might act analogously to the mobile ‘tentacles’ of

actin capping protein that regulate polymerization by binding the barbed ends of

actin filaments (Cooper and Sept 2008). Interestingly, CBFs with unrelated primary

sequences can also promote ParF polymerization in vitro suggesting that a common

mechanism may exist by which diverse CBFs modulate the polymerization of

homologous ParA proteins (Machón et al. 2007). Elucidation of the tertiary structures of additional CBFs will reveal whether they also possess flexible regions that

interact with ParA proteins and modulate their polymerization kinetics.

In common with numerous other bacterial cytoskeletal proteins (Pogliano 2008),

the ParA homologues of plasmids pB171 (ParAB171), F (SopA) and pSM19035 (d)

display distinct subcellular localization patterns, as well as polymerizing in vitro in

response to ATP binding. ParAB171 oscillates along a spiral structure that is distributed over the nucleoid. Oscillation requires both other components of the segrosome, the ParBB171 CBF and the cognate centromere, parC, as well as intact

Walker box motifs within ParAB171. The three elements coordinately position plasmids at regular intervals along the main axis of the cell (Ebersbach and Gerdes

2001; Ebersbach et al. 2006). The oscillation of ParAB171 mimics the behaviour of

MinD proteins that form a discrete branch of the ParA superfamily. MinD, in

concert with MinC, inhibits random placement of the bacterial cell division septum.

Inhibition is relaxed specifically at the cell centre by the MinE factor, thereby

permitting assembly of the cell division apparatus only at the correct location

4  Extrachromosomal Components of the Nucleoid


(Rothfield et al. 2005). Nevertheless, despite their common evolutionary origin and

related subcellular localization patterns, ParAB171 appears to function independently

of MinD, as well as of other components of the cell division complex (Ebersbach

and Gerdes 2001).

Like ParAB171, the ParA homologue (SopA) encoded by the F plasmid oscillates

between the one- and three-quarter cell length positions in E. coli with a periodicity

of ~20 min. This localization pattern requires the SopB CBF and the sopC centromere and likely reflects, first, the formation of SopA helical structures nucleated by

SopB loaded at sopC, followed by the disassembly of the SopA polymers (Adachi

et al. 2006; Hatano et al. 2007; Lim et al. 2005). However, SopA filaments also have

been reported to form independently of SopB. Moreover, SopB also aggregates into

helical structures which partially overlap with SopA spirals, and which require SopA

for their formation in vivo (Adachi et al. 2006). SopA also polymerizes in vitro into

filaments with ultrastructures that resemble those of ParF, and whose formation is

influenced both by SopB and by DNA (Bouet et al. 2007; Lim et al. 2005).

The d protein of plasmid pSM19035 of Streptococcus pyogenes associates with

the nucleoid. This pattern is altered in the presence of the w CBF and centromere

DNA, under which circumstances d oscillates between the nucleoid and the cell

poles, forming spiral-like structures. Accordingly, the protein polymerizes in vitro

in response to ATP binding, and this polymerization is fully dependent on the presence of both DNA bearing the cognate centromere and the w protein with which d

interacts (Pratto et al. 2008). Thus, the behaviour of d may resemble that of Soj, a

ParA homologue encoded by the Bacillus subtilis chromosome, that also polymerizes on DNA (Leonard et al. 2005). Soj binds to DNA non-specifically in vitro, and

also associates with the nucleoid in  vivo (Leonard et  al. 2005; Marston and

Errington 1999; Quisel et al. 1999). However, whether non-specific DNA binding

by ParA proteins is universal is unclear, and the role that this non-specific binding

might fulfil during DNA segregation remains uncertain (Castaing et al. 2008). ATP

hydrolysis induces depolymerization of d filaments indicating that the polymerization–depolymerization equilibrium is modulated by the nucleotide bound state of

the protein. The dimeric crystal structure of d suggests that ATP:ADP exchange can

occur within the dimer without requiring dissociation into monomers (Pratto et al.

2008). By contrast, polymers of ParF and closely-related homologues (Fig. 4.2) are

less prone to depolymerization in vitro and do not require DNA for filamentation

(Barillà et al. 2005; Machón et al. 2007) suggesting that subfamilies of ParA proteins may possess discrete polymerization characteristics.

The ATPase activity of ParA proteins is stimulated by the cognate CBFs, as well

as by non-specific DNA (Barillà et  al. 2005; Davis et  al. 1992; Fung et  al. 2001;

Libante et al. 2001; Pratto et al. 2008; Watanabe et al. 1992). A likely mechanism for

this enhancement has emerged from studies of the N-terminal flexible domain of

ParG and its interaction with ParF. Distinctly from its role in stimulation of ParF

polymerization, the ParG tail possesses an arginine finger-like motif that promotes the

ATPase activity of ParF by ~30-fold (Barillà et al. 2007). The motif may be part of a

semi-flexible loop that intercalates into the ParF nucleotide binding pocket, analogous to arginine fingers in proteins such as human Ras-GAPs (Ahmadian et al. 1997;


F. Hayes and D. Barillà

Nadanaciva et  al. 1999). Arginine finger loops stabilize the transition state during

nucleotide hydrolysis by their partner proteins (Bos et al. 2007), and the same may

be the case with ParF-ParG. Moreover, activation of nucleotide hydrolysis via an

arginine finger loop may be a universally conserved, regulatory mechanism of ParA

family members and their partner proteins (Barillà et  al. 2007). This contention is

supported by the observation that the mobile N-terminal tail in the w CBF also is

necessary for stimulation of ATPase hydrolysis by the d protein (Pratto et al. 2008).

Further study of the interaction between ParA homologues and their corresponding

CBFs will provide crucial insights into the assembly, turnover, and mode of action of

ParA filaments.

How might the dynamic polymerization of ParA proteins mediate plasmid segregation? Two plausible mechanisms focus on the motor-like action of ParA within

the segrosome complex tethered at the centromere site (Barillà et al. 2005). Plasmid

pairing at the mid-cell mediated by the segrosome is considered an initial step in

the segregation process (Edgar et al. 2001; Funnell 2005; Jensen et al. 1998; Pratto

et al. 2008; Ringaard et al. 2007b). One possibility is that, as with type II partitioning complexes (see Section 4.5), ATP-induced elongation of ParA polymers from

paired segrosomes separates and pushes plasmids towards opposite cell poles.

Alternatively, the coupled segrosomes that pair plasmids prior to segregation may

remain relatively static until retraction of ParA polymers pulls the plasmids in different directions (Fig. 4.5) (Barillà et al. 2005). Support for the latter mechanism

comes from observations that ParA polymers specified by one of the two chromosomes of Vibrio cholerae (Fig. 4.2) appear to pull the attached chromosome towards

the cell pole in a process that superficially may mimic chromosome movement during mitosis in eukaryotic cells (Fogel and Waldor 2006). In either pushing or pulling scenarios, the ParA polymerization-depolymerization cycle likely requires an

optimal balance between ATP-induced and ADP-mediated inhibition of polymerization (Barillà et al. 2007). To this end, the intrinsic ATPase activity of ParA must

be stimulated at an appropriate rate and timepoint in the cell cycle by the cognate

CBF. Additionally, the direct modulation of ParA filamentation by the CBF likely

plays a fundamental role in the polymerization-depolymerization events. Indeed,

there may be intriguing parallels between the effects of the CBFs on ParA polymerization and the action of microtubule associated proteins, formins, and other

auxiliary protein factors that influence eukaryotic cytoskeletal dynamics (Machón

et al. 2007).

4.5 An Actin-Like Polymerizing Protein in Type II Segrosomes

Although the genetic organizations of type I and II partition cassettes are very similar, the latter is distinguished particularly by the presence of a gene for a motor-like

protein, ParM, that is structurally and functionally related to eukaryotic actin (Fig. 4.1)

(Bork et  al. 1992). Like actin, ParM is an ATPase (Jensen and Gerdes 1997).

Moreover, ATP induces ParM polymerization in  vitro (Moller-Jensen et  al. 2002),

4  Extrachromosomal Components of the Nucleoid







Fig. 4.5  Model for plasmid segregation mediated by ParA polymer retraction. (Step 1) The CBF

(filled ovals) binds to the centromere (shaded box) generating a pre-segrosome with specific topology and causing plasmid pairing. (Step 2) apo-ParA (open oval) enters the mature segrosome by

interaction with the CBF. (Step 3) ATP-induced polymerization of ParA causes bipolar filamentation of the protein. This filamentation may be modulated directly by the CBF. Stimulation of the

ATPase activity of ParA by the CBF using an arginine finger motif provokes the conversion of

ParA-ATP to an ADP-bound form which inhibits further polymerization. (Step 4) Depolymerization

of the ParA polymers draws the plasmid pairs apart in opposite poleward directions. (Step 5)

Septal closure traps the segregated plasmids in the daughter cells

although polymerization is even more proficient with GTP (Popp et al. 2008). ParM

is also a more effective GTPase than ATPase (Popp et al. 2008). The crystal structures of ParM bound to ADP, GDP or non-hydrolyzable GMPPNP are very similar

(Popp et al. 2008; van den Ent et al. 2002). Considering the properties of ParMGTP mentioned above, this has lead to a suggestion that GTP is the true functional

ligand for ParM activity (Popp et al. 2008).

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

2 Plasmid Segregation: Four Types of Modules

Tải bản đầy đủ ngay(0 tr)