Molecular Microbiology (2010) 78(1), 9–12 䊏
doi:10.1111/j.1365-2958.2010.07316.x
First published online 19 August 2010
MicroCommentary
What is the mechanism of ParA-mediated DNA movement?
mmi_7316 9..12
Martin Howard1 and Kenn Gerdes2*
1
Department of Computational and Systems Biology,
John Innes Centre, Colney Lane, Norwich NR4 7UH,
UK.
2
Centre for Bacterial Cell Biology, Institute for Cell and
Molecular Biosciences, Newcastle University, Newcastle
upon Tyne NE2 4AX, UK.
Summary
The stable maintenance of low-copy-number plasmids requires active partitioning, with the most
common mechanism in prokaryotes involving the
ATPase ParA. ParA proteins undergo intricate spatiotemporal relocations across the nucleoid, dynamics that function to position plasmids at equally
spaced intervals. This spacing naturally guarantees
equal partitioning of plasmids to each daughter cell.
However, the fundamental mechanism linking ParA
dynamics with regular plasmid positioning has
proved difficult to dissect. In this issue of Molecular
Microbiology, Vecchiarelli et al. report on a time-delay
mechanism that allows a slow cycling between the
nucleoid-bound and unbound forms of ParA. The
authors also propose a mechanism for plasmid movement that does not rely on ParA polymerization.
Segregation of replicated DNA prior to cell division is a
vital step in all cells. However, the underlying mechanisms
responsible for reliable segregation have proved surprisingly difficult to uncover. Due to its innate simplicity,
plasmid segregation in prokaryotes has emerged as an
attractive system in which to study this question (Gerdes
et al., 2010). Bacterial plasmids typically encode only
three components that are necessary and sufficient for
segregation. The type I plasmid partitioning (par) locus
encodes the P loop ATPase ParA, a DNA-binding protein
ParB that interacts with ParA and which can also bind to
one or more cis-acting DNA regions (e.g. in Escherichia
coli, parC from plasmid pB171 or parS from plasmid P1).
in vitro ParA associates cooperatively in a dimeric ATP
Accepted 16 July, 2010. *For correspondence. E-mail kenn.gerdes@
newcastle.ac.uk; Tel. +44 0191 208 3230; Fax +44 0191 222 7424.
© 2010 Blackwell Publishing Ltd
form with non-specific DNA (Leonard et al., 2005; Pratto
et al., 2008; Dunham et al., 2009), consistent with in vivo
nucleoid association. ParB also stimulates ParA ATPase
activity in vitro, potentially causing the detachment of
ParA from the nucleoid. Intriguingly, fluorescent labelling
of ParA from the type I locus of plasmid pB171 has
revealed complex spatiotemporal dynamics, with the
protein dynamically relocating across the nucleoid (see,
for example, Ebersbach and Gerdes, 2001; Ringgaard
et al., 2009) in a process requiring both ParB and parC
(Ebersbach and Gerdes, 2001). Originally the relocations
were believed to be simple oscillations between the two
ends of the nucleoid, but recent studies have revealed a
more complex pattern of relocation (Ringgaard et al.,
2009). It is believed that this dynamic ParA relocation is
somehow responsible for plasmid partitioning, as only
cells with a functional par locus are able to equidistantly
position plasmid foci (Ebersbach et al., 2006). However,
the mechanism linking ParA dynamics with plasmid positioning is still far from understood. Nevertheless, recent
evidence does favour a pulling-like mechanism where
plasmids trail retracting regions of high ParA concentration (Ringgaard et al., 2009).
In this issue, work by Vecchiarelli et al. (2010) casts
light on this subject by focusing on how ATP promotes and
regulates the DNA binding of ParA. The authors utilized
total internal reflection microscopy to probe the interaction
between phage l-DNA and functional ParA–GFP (green
fluorescent protein) from plasmid P1. Importantly, only
the ATP-bound form of ParA was able to support DNA
binding, in contrast to all the other nucleotide cofactors
tested. Vecchiarelli et al. then investigated the reason for
this specificity by focusing on the structural effects of
nucleotide binding (as assayed by changes in tryptophan
fluorescence). Steady-state tryptophan fluorescence
measurements revealed the presence of an ATP-specific
conformational change, with a ~20% fluorescence reduction for ATP as compared with a slight increase for ADP
and ATPgS, and little change for AMP. Adding DNA along
with ATP made little difference as compared with ATP
alone, indicating that ParA did not undergo additional
structural changes upon DNA binding that were detectable with this technique. These results were then
extended by dynamic measurements. When ATP was
10 M. Howard and K. Gerdes 䊏
added to ParA, the tryptophan fluorescence underwent
two distinct phases: an initial (~1 min) increase (similar to
that also induced by ADP and ATP analogues), followed
by a slow (~10 min) decrease in fluorescence before
reaching steady state. This second phase was unique to
ATP. Vecchiarelli et al. then used tandem size-exclusion
chromatography/multi-angle light scattering to probe the
oligomeric state of ParA in the presence of ADP or ATP.
Even at relatively high concentrations (30 mM) only dimers
were detected.
Vecchiarelli et al. interpret these results in the following way: ParA–ATP dimerizes to form ParA2:ATP2, after
which it undergoes a slow conformational change to a
ParA*2:ATP2 form. Although the ParA–ADP form (for
example) also dimerizes, only the ATP state is capable
of undergoing the second, slow transition. The authors
further suggest that only the ParA*2:ATP2 form is competent to bind non-specific DNA. An important consequence of this sequence of events is an effective time
delay between the release of ParA from the nucleoid via
ATP hydrolysis and its reacquisition of the nucleoidbinding conformation ParA*2:ATP2. Provided this delay is
long compared with the time to diffuse across the cell
(~1 s), diffusion will completely disperse the ParA*2:ATP2
allowing nucleoid re-binding with an equal probability in
all locations. In a final interesting result, Vecchiarelli
et al. also found little evidence for ATP-dependent filament formation, even at concentrations substantially
higher than the estimated in vivo value. This conclusion,
if confirmed, is also likely to be important, although it is
at odds with previous results from a variety of groups
(see below).
Using the above results, Vecchiarelli et al. constructed
an outline ‘diffusion-ratchet’ model for ParA-mediated
plasmid partitioning (see Fig. 1A). In this model, since
plasmid-bound ParB interacts directly with ParA, the
plasmid will have a higher affinity for high-density ParA
regions, due to the larger number of ParA–ParB bonds
that can be formed there. However, ParB stimulates the
ATPase activity of ParA causing ParA to fall off the
nucleoid. The disassociated ParA then experiences the
time delay discussed above, before being able to re-bind
in an effectively random location on the nucleoid. These
processes deplete the region around the plasmid of ParA:
the plasmid will therefore experience an effective force
towards regions of higher ParA concentration with a larger
number of ParA–ParB bonds. In this way long-ranged
plasmid movement can be achieved through the relocation dynamics of ParA. Equidistant plasmid positioning
could potentially result from such a mechanism, although
more rigorous mathematical modelling would be required
to confirm this conclusion.
Previous experimental and mathematical modelling
work has suggested a filament-pulling model where plas-
mids are pulled by retracting ParA cytoskeletal filaments
(see Fig. 1B), but with a length-dependent plasmidfilament detachment rate (Ringgaard et al., 2009). Similar
conclusions have very recently been reached in Caulobacter crescentus using super-resolution fluorescence
imaging, where discrete linear ParA structures were
observed (Ptacin et al., 2010). Simple stochastic simulations of such a model have demonstrated very good
agreement with experimentally measured plasmid distributions (Ringgaard et al., 2009). Interestingly, models
with length-dependent rates have recently been introduced and validated in microtubule length regulation
(Varga et al., 2009). It will be interesting to see if related
ideas could apply to prokaryotes. The Ringgaard et al.
model also assumes that the cytoplasmic ParA is well
mixed, consistent with the time delay discussed above.
Clearly, the time-delay feature cannot on its own be used
to differentiate between possible mechanisms.
The fundamental difference between the Ringgaard
et al. and Vecchiarelli et al. models lies instead in the
presence or absence of ParA polymerization into filaments, and, critically, whether depolymerization of filaments can provide the force for plasmid movement. As
mentioned above, Vecchiarelli et al. failed to find evidence
for self-supporting ParA polymerization even at high
concentrations. However, previous work in vitro and in
vivo has produced evidence in the opposite direction:
using electron microscopy (EM), ParA2 of Vibrio cholerae
chromosome II was observed to form helical filaments on
double-stranded DNA at roughly physiological concentrations (Hui et al., 2010). Similarly, ParA from plasmid P1
was shown by EM to form filaments in vitro, although
at high concentrations (~40 mM ParA) (Dunham et al.,
2009). Fluorescent microscopy of the same system also
showed ParA filaments connecting ParB-covered beads,
although at similarly high ParA concentrations. The
Dunham et al. experiments did, however, lack DNA which
is potentially an important aid for polymerization, and this
absence may explain the high concentrations that were
required before ParA polymerization was observed. in
vivo imaging of fluorescently labelled ParA at close to
wild-type concentration levels has also pointed strongly
towards helical filaments wrapping the nucleoid (Ringgaard et al., 2009). Vecchiarelli et al. propose that the
helical structures imaged in vivo are the result of an
underlying helical structure of the nucleoid. Whether that
is plausible remains to be tested, but certainly some kind
of helical structure is required by the ‘diffusion-ratchet’
mechanism to bias plasmid movement along the long axis
of the nucleoid and to prevent transverse plasmid motion
that would not contribute to partitioning. In any case, it is
important to establish whether ParA really does polymerize (and hence whether the above evidence for polymerization is simply an in vitro artefact) and, furthermore, if
© 2010 Blackwell Publishing Ltd, Molecular Microbiology, 78, 9–12
Models for DNA Movement by ParA 11
Fig. 1. A. Schematic representation of the ‘diffusion-ratchet’ model proposed by Vecchiarelli et al. The figure illustrates two plasmids being
pulled towards each other across the nucleoid surface. The pulling force is directed towards regions of high ParA concentration where a
greater number of ParA–ParB bonds can be formed. The clock symbol represents a time delay induced by a slow conformational change of
ParA to its DNA binding form.
B. Schematic representation of the filament-pulling model proposed by Ringgaard et al., also illustrating two plasmids being pulled towards
each other. The model in (B) relies on a pulling force generated by ParA filament depolymerization. The same time delay as in (A) is shown.
The length-dependent plasmid detachment rate is not illustrated.
such polymerization is present, whether depolymerization
is critical for plasmid segregation. Reliable information
about ParA polymerization is vital, as the presence or
absence of ParA filaments will inform the biophysical
mechanism of pulling force generation. In vitro reconstitution would also be extremely useful, as it would stringently test minimal models for ParA dynamics/plasmid
partitioning. Reconstitution has already been achieved for
the closely related Min system regulating cell division
positioning in E. coli (Loose et al., 2008; Ivanov and Mizuuchi, 2010), a system where mathematical modelling
has also been very useful (Kruse et al., 2007). However,
the difficulty of working with ParA/ParB in vitro makes
reconstitution problematic and may also partly explain the
© 2010 Blackwell Publishing Ltd, Molecular Microbiology, 78, 9–12
lack of consensus for the polymerizing properties of the
ParA protein.
Segregation of DNA mediated by ParA is now a rapidly
advancing topic of research. The connection between the
complex spatiotemporal ParA dynamics and the mechanism of partitioning is subtle: certainly the connection is
too complex to intuit with simple cartoon-like models
(despite the best efforts of Fig. 1). Progress is, however,
still being made by employing a wide spectrum of methodologies, ranging from fluorescent microscopy and
EM studies to mathematical modelling and in vitro
reconstitution. ParA-mediated partitioning is therefore an
excellent example of an intricate biological problem that
will only be solved by a truly interdisciplinary approach.
12 M. Howard and K. Gerdes 䊏
Still, our current inability to properly comprehend a system
with such a small number of components (five: ParA,
ParB, parC or parS, ATP, DNA) is sobering. Systems
biologists who would claim to reliably predict the network
properties of hundreds of interacting proteins should
perhaps take note.
Acknowledgements
Work in the M.H. and K.G. groups is supported by the Biotechnology and Biological Sciences Research Council of the
UK. M.H. is supported financially by The Royal Society and
K.G. by the Wellcome Trust.
References
Dunham, T.D., Xu, W., Funnell, B.E., and Schumacher,
M.A. (2009) Structural basis for ADP-mediated transcriptional regulation by P1 and P7 ParA. EMBO J 28: 1792–
1802.
Ebersbach, G., and Gerdes, K. (2001) The double par locus
of virulence factor pB171: DNA segregation is correlated
with oscillation of ParA. Proc Natl Acad Sci USA 98:
15078–15083.
Ebersbach, G., Ringgaard, S., Møller-Jensen, J., Wang, Q.,
Sherratt, D.J., and Gerdes, K. (2006) Regular cellular distribution of plasmids by oscillating and filament-forming
ParA ATPase of plasmid pB171. Mol Microbiol 61: 1428–
1442.
Gerdes, K., Howard, M., and Szardenings, F. (2010) Pushing
and pulling in prokaryotic DNA segregation. Cell 141: 927–
942.
Hui, M.P., Galkin, V.E., Yu, X., Stasiak, A.Z., Stasiak, A.,
Waldor, M.K., and Egelman, E.H. (2010) ParA2, a Vibrio
cholerae chromosome partitioning protein, forms left-
handed helical filaments on DNA. Proc Natl Acad Sci USA
107: 4590–4595.
Ivanov, V., and Mizuuchi, K. (2010) Multiple modes of interconverting dynamic pattern formation by bacterial cell division proteins. Proc Natl Acad Sci USA 107: 8071–8078.
Kruse, K., Howard, M., and Margolin, W. (2007) An experimentalist’s guide to computational modelling of the Min
system. Mol Microbiol 63: 1279–1284.
Leonard, T.A., Butler, P.J., and Löwe, J. (2005) Bacterial
chromosome segregation: structure and DNA binding of
the Soj dimer – a conserved biological switch. EMBO J 24:
270–282.
Loose, M., Fischer-Friedrich, E., Ries, J., Kruse, K., and
Schwille, P. (2008) Spatial regulators for bacterial cell division self-organize into surface waves in vitro. Science 320:
789–792.
Pratto, F., Cicek, A., Weihofen, W.A., Lurz, R., Saenger, W.,
and Alonso, J.C. (2008) Streptococcus pyogenes
pSM19035 requires dynamic assembly of ATP-bound ParA
and ParB on parS DNA during plasmid segregation.
Nucleic Acids Res 36: 3676–3689.
Ptacin, J.L., Lee, S.F., Garner, E.C., Toro, E., Eckart, M.,
Comolli, L.R., Moerner, W.E., and Shapiro, L. (2010) A
spindle-like apparatus guides bacterial chromosome
segregation. Nat Cell Biol doi: 10.1038/ncb2083
Ringgaard, S., van Zon, J., Howard, M., and Gerdes, K.
(2009) Movement and equipositioning of plasmids by ParA
filament disassembly. Proc Natl Acad Sci USA 106:
19369–19374.
Varga, V., Leduc, C., Bormuth, V., Diez, S., and Howard, J.
(2009) Kinesin-8 motors act cooperatively to mediate
length-dependent microtubule depolymerization. Cell 138:
1174–1183.
Vecchiarelli, A.G., Han, Y.-W., Tan, X., Mizuuchi, M.,
Ghirlando, R., Biertümpfel, C., et al. (2010) ATP control of
dynamic P1 ParA–DNA interactions: a key role for the
nucleoid in plasmid partition. Mol Microbiol 78: 78–91.
© 2010 Blackwell Publishing Ltd, Molecular Microbiology, 78, 9–12