Gene regulation circuits control all aspects of the life of plasmids. This review gives an overview of the current orchestration of the circuits that control plasmid replication, plasmid transfer, plasmid segregation and plasmid maintenance.

Plasmids are extrachromosomal genetic elements that by definition are non-essential for survival of their host, but which often confer a genetic advantage. Plasmids occur in many systems but have been most intensively and extensively studied in bacteria and this paper will concentrate exclusively on bacterial plasmids, although many of the general points may be universally applicable. One of the main reasons for the interest in plasmids is because of the phenotypic traits that they confer, antibiotic resistance in clinical contexts and degradation of xenobiotics in an environmental context being the most well known examples. Both these traits provide their host with an advantage in a specific niche that is not the exclusive habitat for the host. Thus Pseudomonas aeruginosa is found both in clinics and in natural aquifers. Specific plasmid-encoded phenotypes may allow this species to occupy niches of these environments defined by the presence of harmful compounds or unusual nutrients.

Particular plasmids go through peaks of abundance when they are associated with a trait that is spreading due to its ability to confer a newly acquired beneficial trait. Ultimately however, plasmids are examples of selfish DNA that exist because they have evolved the ability to replicate faster than they are lost from populations of their hosts. That is, they are genetic elements that have learnt to replicate efficiently, be inherited stably, cause very little burden on their host and in some cases spread from one bacterium to another [1]. A great deal of this success comes from the ability to regulate their genes tightly [2], and the present paper will deal with examples of such regulation which are of general biological interest and where recent developments have occurred that are worth highlighting.

Regulation of replication

In contrast with classical subjects of gene regulation studies such as the lac operon or the SOS regulon, which respond to external factors such as the availability of nutrients or stress, plasmid replication control, at its most basic, senses past cellular events that affect current plasmid concentration– cell growth and replication itself. Plasmids do this via various kinds of negative feedback loops. Antisense RNA was first most extensively explored in bacterial plasmids and can have a number of effects: inhibiting transcription, inhibiting translation or inhibiting the activity of the target RNA in priming replication. A key feature of these circuits is the short half-life of the regulatory RNA, which allows the concentration of the repressor to change quickly in response to plasmid copy number. The regulatory RNAs interact with their target mRNA to either control translation directly (for example to block translation), or influence the direction of folding and thus modulate structures that promote translational initiation or transcriptional termination. The basic mechanisms of these are understood [3,4] and the rules for how the circuits function provide an interesting lesson in predictive biology, but many questions remain and as more systems are studied, variations on the theme and novel combinations of mechanism continue to be found [5].

The other key circuits are those autogenous or pseudo-autogenous circuits that limit the amount of Rep protein produced but which do not actually control copy number [6]. These are typically associated with systems in which Rep binds to two types of related sequences in or near the replication origin that may be bound by different versions of the protein, for example monomer versus dimer. Repression of rep gene expression is normally through binding of the dimer at or near the rep promoter, while activation of replication is dependent on Rep (often the monomeric version) binding to repeats, ‘iterons’ in the origin itself. Plasmid replication is then not controlled directly by the amount of Rep protein present but by inhibitory complexes that form as replication events occur – so-called handcuffing [7].

Regulation of Psk (post-segregational killing) systems

Although not unique to plasmids, Psk systems represent some of the most interesting regulatory phenomena encoded by plasmids. Two types are known: proteic systems in which the action of a potentially lethal protein is controlled by an antidote protein that decays faster than the toxin; and those systems controlled by antisense RNA such as the classic hok-sok system, where the killer mRNA becomes active after a delay but is then controlled by an antisense RNA that will have degraded in plasmid-free segregants [8]. An exciting recent discovery is that the kis-kid system, encoded by the IncFII plasmid R1, targets RNA molecules with sequence specificity, and is switched on when plasmid copy number decreases and before the plasmid is lost [9]. The result is that not only is cell division delayed but also the copA RNA that represses expression of rep is reduced, thus allowing increased replication. Since this Psk operon is autogenously controlled like many others [10], an important question is how the decrease in copy number is rapidly detected to result in increased and differential expression. This represents a very sophisticated degree of co-ordination of plasmid stable inheritance functions not previously suspected and will prompt a re-examination of other systems to see if similar links have been missed in Psk systems.

Regulation of plasmid partitioning operons

Plasmid partitioning involves one or more proteins that bind the plasmid DNA and achieve symmetrical placement and movement of the plasmid inside the cell so that there is better than random movement to either side of the cell division plane. The best-studied examples of gene sets that achieve this function encode two key proteins. One of these is the so-called centromere-binding protein that provides the handle to pull or push the plasmid and the other is the force-generating protein that may belong to either the actin-cytoskeleton protein family, or the Walker ATPase family, many of which also possess an N-terminal domain that allows DNA binding [11]. Since both partitioning proteins can frequently bind DNA, it is not surprising that the par operons are controlled by a variety of autogenous control circuits that can involve one or both proteins [12]. It is still not understood how the Walker ATPase ParA proteins balance their dual role of cytoskeleton polymer generation and DNA-binding autogenous repressor.

Regulation of plasmid transfer

In contrast with replication, a number of transfer systems seem to respond to physiological factors other than simple cell growth. The best studied example is the Ti plasmid family whose transfer system is quorum-regulated. In addition, expression responds to nutrients released by plants already transformed by cognate T-DNA (transfer DNA) that directs production of the opine molecules. Thus the plasmid is stimulated to spread when the donor bacteria are at high density and the conditions favour bacterial growth [13]. In contrast, apart from F itself, which is a mutant constitutive for expression of transfer genes, plasmids of the F family transfer at reduced frequency or not at all at high cell density and in stationary phase. Recent evidence has indicated that this is due to increase in the level of the nucleoid-associated protein H-NS (histone-like nucleoid structuring protein) binding to bent DNA around the oriT traJ, traM region to inhibit activation of the tra promoters [14]. Activation of expression of the tra genes when it happens is due to TraJ blocking the action of H-NS.

The TraJ level is regulated by a number of factors including the FinO–FinP protein–RNA antisense repressor complex that blocks translation. The stability of RNA is also a key stage of control and this is illustrated by the role of Hfq in traJ [15]. Hfq binds to an AU-rich spacer region between RNA stem–loop structures. Inactivation of hfq increases stability, thus indicating that Hfq may be important in limiting the level of key transcripts so that other factors could have a significant effect on the overall translational rate (if RNA level was not limiting then minor changes might not have much effect).

In the symbiotic megaplasmids of the Rhizobiaceae, a new system has been discovered recently in which the transfer genes seem to be largely repressed by the presence of a small plasmid-encoded repressor (RctA), which autogenously activates its own expression. This repression is counter-balanced by the action of RctB, but the signal to which RctB responds is unknown [16]. The formal features of this circuitry are reminiscent of the F-type circuitry (i.e. active repression counter-balanced by activation) and may indicate that the disadvantage of the basal expression of a transfer system (conferring sensitivity on male-specific bacteriophages for example) is such that it is advantageous to ensure that the system is completely off when it is not activated.

Other transfer systems appear to be expressed at all times in an autogenously regulated way. For the IncP-9 and IncW plasmid families, which have closely regulated transfer systems, this occurs through feedback repression by open reading frames encoded in each of its main transfer gene operons [17]. For the IncP-1 plasmids, there is additional control through the global repressor KorB encoded in the central control operon as described below.

Global regulation of plasmid backbone functions

While most of the basic modules of which plasmids are composed encode their own regulatory circuits, there are few well-documented systems in which there is over-arching co-ordination of the expression of these functions. One such system is that of plasmid pSM19035 that is endogenous to Streptococcus pyogenes [18]. A small repressor protein called ω regulates transcription of genes for plasmid copy number control, active partitioning and Psk. The sequences to which this protein binds are found in three promoter regions and consist of arrays of 7 bp sequences that can be present as a mixture of either direct or inverted repeats [19].

Another well-studied system is the IncP-1 family, of which the archetype is RK2/RP4. The key part of this is the socalled central control operon that encodes homologues of the ParA and ParB active partitioning genes, IncC and KorB respectively [20,21]. IncC is an ATPase that interacts with KorB and directs the symmetrical distribution of the plasmid DNA–KorB complex in the cell [22]. KorB binds 12 operators, distributed across the plasmid genome, and these, individually or collectively, act as the centromere-like sequence for active partitioning, the handle that the KorB uses to hold the plasmid. Six of these sites are at or near the promoters for genes required for replication (trfA), stable inheritance (kfrA and kle) and conjugative transfer (trbB and traG) [23]. Through these sites, KorB can repress transcription of the cognate genes. In doing this, KorB normally works co-operatively with either KorA or TrbA [24], repressor proteins encoded in the central control operon and the trb region respectively. KorA and TrbA are composed of a closely related C-terminal dimerization domain (CTD), but different DNA-binding domains [25]. The CTD is important for the co-operative interaction with KorB [24,26]. Co-operativity allows very high levels of repression without a need for great increases in repressor concentration and the system is finely balanced [27]. This appears to lie at the heart of one of the IncP-1 strategies that promote broad host range. Strong universal promoters exist for all operons encoding important maintenance and transfer functions and tight feedback regulation ensures that expression is shut down to basal maintenance levels once the plasmid is established in a new host and repressor levels have built up to a steady state [2].

Conclusions

All of the genetic modules mentioned in this paper contribute to the overall success (replication and spread) of the plasmids that carry them. In most cases, they have their own regulatory circuits, but it makes sense that where circuits evolve that co-ordinate expression of different modules this may provide an advantage and these will be selected over the system without such co-ordination. Once such a circuitry exists, it will also impose a certain amount of extra stability on the system which will counteract genetic changes or rearrangements that would disrupt the circuits. Not surprisingly, the IncP-1 backbone with its complex circuits is one of the most conserved systems found in plasmids to date as evidenced from more than 17 genome sequences.

Molecular Basis of Transcription: A Focus Topic at BioScience2006, held at SECC Glasgow, U.K., 23–27 July 2006. Edited by S. Busby (Birmingham, U.K.), R. Weinzierl (Imperial College London, U.K.) and R. White (Glasgow, U.K.).

Abbreviations

     
  • CTD

    C-terminal dimerization domain

  •  
  • H-NS

    histone-like nucleoid structuring protein

  •  
  • Psk

    post-segregational killing

The work on plasmid gene regulation in this laboratory has been funded by project grants from The Wellcome Trust.

References

References
1
Thomas
C.M.
Funnell
B.E.
Phillips
G.J.
Plasmid Biology
2004
Washington, DC
American Society for Microbiology
(pg. 
509
-
528
)
2
Bingle
L.E.H.
Thomas
C.M.
Curr. Opin. Microbiol.
2001
, vol. 
4
 (pg. 
194
-
200
)
3
del Solar
G.
Espinosa
M.
Mol. Microbiol.
2000
, vol. 
37
 (pg. 
492
-
500
)
4
Brantl
S.
Funnell
B.E.
Phillips
G.J.
Plasmid Biology
2004
Washington, DC
American Society for Microbiology
(pg. 
47
-
62
)
5
Kwong
S.M.
Skurray
R.A.
Firth
N.
J. Bacteriol.
2006
, vol. 
188
 (pg. 
4404
-
4412
)
6
Kruger
R.
Rakowski
S.A.
Filutowicz
M.
Funnell
B.E.
Phillips
G.J.
Plasmid Biology
2004
Washington, DC
American Society for Microbiology
(pg. 
25
-
45
)
7
Chattoraj
D.K.
Mol. Microbiol.
2000
, vol. 
37
 (pg. 
467
-
476
)
8
Gerdes
K.
Ayora
S.
Canosa
I.
Ceglowski
P.
Diaz-Orejas
R.
Franch
T.
Gultyaev
A.P.
Bugge Jensen
R.
Kobayashi
I.
Macpherson
C.
, et al. 
Thomas
C.M.
The Horizontal Gene Pool: Bacterial Plasmids and Gene Spread
2000
Amsterdam
Harwoord Academic Press
(pg. 
49
-
85
)
9
Pimentel
B.
Madine
M.A.
de la Cueva-Mendez
G.
EMBO J.
2005
, vol. 
24
 (pg. 
3459
-
3469
)
10
Ruiz-Echevarria
M.J.
Berzalherranz
A.
Gerdes
K.
Diaz-Orejas
R.
Mol. Microbiol.
1991
, vol. 
5
 (pg. 
2685
-
2693
)
11
Hayes
F.
Barillà
D.
Nat. Rev. Microbiol.
2006
, vol. 
4
 (pg. 
133
-
143
)
12
Williams
D.R.
Thomas
C.M.
J. Gen. Microbiol.
1992
, vol. 
138
 (pg. 
1
-
16
)
13
Oger
P.
Farrand
S.K.
Mol. Microbiol.
2001
, vol. 
41
 (pg. 
1173
-
1185
)
14
Will
W.R.
Frost
L.S.
J. Bacteriol.
2006
, vol. 
188
 (pg. 
507
-
514
)
15
Will
W.R.
Frost
L.S.
J. Bacteriol.
2006
, vol. 
188
 (pg. 
124
-
131
)
16
Perez-Mendoza
D.
Sepulveda
E.
Pando
V.
Munoz
S.
Nogales
J.
Olivares
J.
Soto
M.J.
Herrera-Cervera
J.A.
Romero
D.
Brom
S.
Sanjuan
J.
J. Bacteriol.
2005
, vol. 
187
 (pg. 
7341
-
7350
)
17
Lambertsen
L.M.
Molin
S.
Kroer
N.
Thomas
C.M.
Plasmid
2004
, vol. 
52
 (pg. 
169
-
181
)
18
de la Hoz
A.B.
Ayora
S.
Sitkiewicz
I.
Fernandez
S.
Pankiewicz
R.
Alonso
J.C.
Ceglowski
P.
Proc. Natl. Acad. Sci. U.S.A.
2000
, vol. 
97
 (pg. 
728
-
733
)
19
Weihofen
W.A.
Cicek
A.
Pratto
F.
Alonso
J.C.
Saenger
W.
Nucleic Acids Res.
2006
, vol. 
34
 (pg. 
1450
-
1458
)
20
Motallebi-Veshareh
M.
Rouch
D.A.
Thomas
C.M.
Mol. Microbiol.
1990
, vol. 
4
 (pg. 
1455
-
1463
)
21
Siddique
A.
Figurski
D.H.
J. Bacteriol.
2002
, vol. 
184
 (pg. 
1788
-
1793
)
22
Bignell
C.R.
Haines
A.S.
Khare
D.
Thomas
C.M.
Mol. Microbiol.
1999
, vol. 
34
 (pg. 
205
-
216
)
23
Kostelidou
K.
Thomas
C.M.
J. Mol. Biol.
2000
, vol. 
295
 (pg. 
411
-
422
)
24
Kostelidou
K.
Jones
A.C.
Thomas
C.M.
J. Mol. Biol.
1999
, vol. 
289
 (pg. 
211
-
221
)
25
Bhattacharyya
A.
Figurski
D.H.
J. Mol. Biol.
2001
, vol. 
310
 (pg. 
51
-
67
)
26
Zatyka
M.
Bingle
L.
Jones
A.C.
Thomas
C.M.
J. Bacteriol.
2001
, vol. 
183
 (pg. 
1022
-
1031
)
27
Bingle
L.E.H.
Zatyka
M.
Manzoor
S.E.
Thomas
C.M.
Mol. Microbiol.
2003
, vol. 
49
 (pg. 
1095
-
1108
)