SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References

Transport of DNA across bacterial membranes during natural transformation is a fascinating and elaborate process. It requires the functional integrity of huge multi-protein complexes present in the bacterial envelope at distinct loci. After successful mapping of essential gene products involved in natural transformation, current research focuses on the functional interplay of these components in order to understand the mechanisms how DNA enters the bacterium. Here, we discuss the model of a two-step DNA uptake process in competent Gram-negative and Gram-positive bacteria. The first step comprises the transfer of DNA from the bacterial surface to the cytoplasmic membrane. For this purpose, bacteria use a variety of machineries, mostly, but not necessarily, sharing key homologous components. The second step is the translocation of DNA across the cytoplasmic membrane, a tight barrier at which ion gradients are established for energization of the cell. Crossing the latter is mediated by a protein complex harbouring a highly conserved membrane channel. On the basis of current data, at least the first step is uncoupled from the second. This review intends to highlight mechanistic features of both steps of bacterial DNA uptake by the integrative interpretation of genetic, biochemical and biophysical data.

Natural transformation is regarded as the capability to take up naked DNA from the environment. It is widespread among bacteria and thought to provide material for DNA repair as well as for nutrition. But most importantly, it enables bacteria to acquire new genetic components suitable for adaptation to changing external conditions. The process of natural transformation comprises different stages: (i) DNA uptake, (ii) recombination of homologous DNA into the chromosome or reconstitution of plasmid DNA and (iii) eventually phenotypic expression of acquired genetic material. This review deals with the initial process of DNA uptake, regardless of the fate imported DNA faces in the interior of the bacterium. Considering experimental data, ‘DNA uptake’ is defined as the transport of DNA into a DNaseI-resistant state. The latter is established by transport of DNA across the outer membrane in Gram-negative bacteria, irrespective of further transport into the cytoplasm. For Gram-positive bacteria, lacking an outer membrane, DNA uptake is synonymous to DNA transport into the cytoplasm.

The characterization of several DNA uptake machineries revealed homologous functions as well as species-specific adaptations. Among these transport complexes, those of Bacillus subtilis and Neisseria gonorrhoeae are the most extensively studied. They serve as archetype DNA uptake machineries of Gram-positive or Gram-negative bacteria respectively. For comprehensive reviews on the phenotypes of transformation-deficient mutants the reader might refer to the following publications (Averhoff and Friedrich, 2003; Chen and Dubnau, 2004; Hamilton and Dillard, 2006; Claverys et al., 2009).

Data on the existence of DNA intermediates within different bacterial compartments were so far missing or based on indirect evidence. Recently, we directly visualized DNA intermediates during uptake in different bacterial compartments in the Gram-negative bacterium Helicobacter pylori and demonstrated that the bacterium transports DNA in steps across each membrane (Stingl et al., 2010). Moreover, the first step was mechanistically and temporally uncoupled from the second. Here we propose that DNA uptake proceeds in two steps in all transformable bacteria and present models how the macromolecule is pulled into the interior of the cell.

The first step – transfer of DNA from the surface to the cytoplasmic membrane

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References

The first step comprises the delivery of the DNA substrate to the cytoplasmic membrane, i.e. transport of DNA across the cell wall in Gram-positives or across the outer membrane in Gram-negatives. Although Gram-positive and Gram-negative bacteria differ enormously concerning the composition of the external surface, they mostly share related components for transport of free DNA towards the cytoplasmic membrane (Fig. 1). Except for H. pylori, in which a conjugation-like type IV secretion system mediates the first step in DNA uptake (Hofreuter et al., 2000; Stingl et al., 2010), all other studied transformable bacteria use components of type IV pili and type II secretion systems (Chen and Dubnau, 2004). The close relative of H. pylori, Campylobacter jejuni, is special in that both systems were identified to play a role in natural transformation. However, the plasmid-encoded type IV secretion system is only present in certain strains and therefore appears to play a minor role in DNA uptake of this pathogen (Bacon et al., 2000; Wiesner et al., 2003).

image

Figure 1. Model of DNA uptake in Gram-positive (B. subtilis) and Gram-negative bacteria (N. gonorrhoeae, H. pylori). In dark blue, species-specific components of the DNA uptake machinery; same colour except dark blue, homologous components. The first step of DNA transfer from the surface to the cytoplasmic membrane is mediated by species-specific DNA uptake machineries, eventually sharing some homologous components. The second step of DNA translocation across the cytoplasmic membrane is highly conserved. The latter is mediated by a homologue of the ComEC channel (named ComA in Neisseria). ComEA or its soluble homologue ComE in Neisseria act as DNA binding proteins feeding the ComEC channel. The helicase ComFA or its close homologue PriA in Neisseria and H. pylori may pull ssDNA into the cytoplasm on the expense of ATP. Incoming ssDNA is protected by binding proteins. Species-specific adaptations for transfer of DNA to the cytoplasmic membrane are the following: A. In B. subtilis, a pseudopilus composed of ComG proteins (mainly ComGC, with contributions of comGD, GE, GG) spans the cell wall. Assembly of the pseudopilus is mediated by the integral membrane protein ComGB, a peptidase for processing of pilin subunits (ComC) and the assembly ATPase ComGA. Disassembly/retraction of the pseudopilus opens a cell wall hole that enables DNA to diffuse ∼ 55 nm from the surface to the DNA receptor ComEA, which is anchored in the membrane. B. In N. gonorrhoeae, the pilus (mainly composed of PilE proteins) is assembled by a complex, which comprises at least the ATPase PilF, the integral membrane protein PilG and a peptidase for pilin processing (PilD). In our model, the pilus fibre inserted into PilQ leads to maturation of the pore. The retraction of this ‘placeholder’ by the dissassembly ATPase PilT opens PilQ for DNA to bind at the central cavity of the pore. This might lead to transient exposure of the putative DUS receptor (DUS-R), probably stabilized by ComP (not depicted here), at the surface of the bacterium. Upon binding of DUS-containing DNA to its receptor, the latter relocalizes, thereby transferring a small DNA loop into the periplasm. ‘Receptor shuttling’ might depend on conformational changes in the complex after specific DNA binding and/or on active pilus retraction. In the periplasm the DNA loop is bound by ComE proteins that exert force on the macromolecule, thus, driving import of DNA across the outer membrane. DNA packed by ComE diffuses to ComA. C. In H. pylori, a type IV secretion system (ComB) mediates DNA translocation across the outer membrane. ComB is probably composed of an outer membrane complex (ComB7, B9 and B10) and an inner membrane complex (ComB3, B4, B6, B8 and also B10). Eventually, the pilin subunit ComB2 binds to DNA at the surface of the bacterium (pseudopilus). Its disassembly and/or rotation by the action of the ATPase ComB4 may lead to movement of bound DNA into the periplasm. A ComE homologue is lacking, so that the interaction partner of the ComEC channel remains to be identified.

Download figure to PowerPoint

Data obtained from different bacterial species revealed that DNA can be translocated as double strand (ds) across the outer membrane or the cell wall. In B. subtilis degradation of the second strand was observed, when the first strand concomitantly entered into a DNaseI-resistant state (Provvedi et al., 2001). Because DNaseI resistance is established in the cytoplasm in Gram-positive bacteria, these results indicate that dsDNA crossed the cell wall in the latter experiments. From Neisseria, linearized transformed plasmids were quantitatively reisolated from mutants deficient in DNA uptake into the cytoplasm (Facius et al., 1996). For H. pylori, we directly monitored entry of dsDNA into the periplasm by staining dsDNA with a dsDNA-specific intercalator dye (Stingl et al., 2010). However, also free ssDNA can serve as substrate for natural transformation in competent bacteria (Stein, 1991; Levine et al., 2007), implicating that the first step has to be flexible in terms of substrate recognition and transport.

Which bacterial structures get in touch with DNA first? And how is a dsDNA tether of approximately 2.4 nm diameter translocated from the surface to the cytoplasmic membrane? In B. subtilis, a pseudopilus composed of ComG proteins (ComGC, GD, GE, GG) was shown to be involved in access of DNA to the DNA binding protein ComEA localized at the cytoplasmic membrane (Provvedi and Dubnau, 1999) (Fig. 1A). The pseudopilus was dispensable for DNA binding to ComEA in Bacillus protoplasts lacking a cell wall. Homologous pili fibres in other organisms are able to retract and elongate. Hence, it was suggested that the Bacillus pseudopilus might promote DNA movement towards the cytoplasmic membrane by retraction. However, ComG proteins themselves did not bind DNA, leading to the proposition that the pseudopilus causes rearrangements in the cell wall (Provvedi and Dubnau, 1999). In Gram-positive bacteria, diffusion of DNA through cell wall holes induced by the ComG pseudopilus could putatively cause efficient DNA translocation from the surface to the cytoplasmic membrane, over a distance of approximately ∼ 55 nm (Matias and Beveridge, 2005). With respect to the putative role of pili in Gram-negative bacteria (see below) the pseudopilus, which is thought to extend from the cytoplasmic membrane through the cell wall to the bacterial surface, might even serve as a placeholder. Its retraction/disassembly would open a hole in the cell wall for diffusion of DNA to the membrane, where DNA is bound by ComEA (Fig. 1A).

In Gram-negative bacteria the outer membrane presents the first barrier for transforming DNA. In Neisseria, the secretin PilQ was shown to bind both ssDNA and dsDNA in a sequence-independent manner, with the binding site located in the central cavity (Assalkhou et al., 2007) (Fig. 1B). Although functional type IV pili are essential for DNA uptake, their central cavity of 1.2 nm is not sufficient to harbour the DNA during transport (Forest and Tainer, 1997). Type IV pili are retractable fibres that mediate bacterial motility and are involved in host–pathogen interaction beyond their role in natural transformation (Pelicic, 2008). Their proteinaceous components are thought to be recycled at the cytoplasmic membrane by the antagonistic action of two ATPases (PilF for assembly, PilT for disassembly). The precise mechanism of assembly and disassembly is currently under investigation. For natural transformation, the secretin PilQ is essential as well as pilus biogenesis and the pilus disassembly ATPase PilT. The narrowest diameter of the PilQ secretin exhibits 6.5 nm (Collins et al., 2004), suitable to translocate pili through the outer membrane and/or let DNA pass through. One simple hypothesis, which is often discussed, is that DNA may be taken up across the outer membrane as a result of DNA hitchhiking by virtue of its attaching to the retracting pilus. However, like their homologous counterparts in B. subtilis, pili only marginally interact with DNA (Assalkhou et al., 2007). In addition, low expression of pili or expression of genetically modified pilin subunits (leading to less piliation) did not lead to a significant defect in natural transformation (Long et al., 2003; Aas et al., 2007). In addition, the pilus disassembly ATPase PilT was not essential in the presence of His-tagged pilin subunits in Pseudomonas (Graupner et al., 2001). Therefore, the hypothesis arose that an alternative transformation machine shares and competes with the pilus assembly/disassembly system for basic functional components. The importance of PilT for DNA uptake might be understood by blockage of the PilQ secretin caused by a static pilus fibre in the absence of PilT (Graupner et al., 2001). In contrast, complete absence of pilin subunits abrogates natural transformation as well. The contact between pilin subunits and PilQ might be essential for opening the secretin for incoming DNA.

Furthermore, only few data are available concerning the need for punctures in the peptidoglycane layer for DNA access to the periplasmic space. This might be mediated by several additional components, like e.g. the periplasmatically located lipoproteins ComL and Tpc (Facius et al., 1996).

In Neisseria, the DNA binding protein ComE, a periplasmic soluble homologue of ComEA, is encoded by four identical comE copies on the chromosome. Their sequential knock-out led to an additive diminishing effect on DNA transformation, in particular on DNA uptake (Chen and Gotschlich, 2001). We propose that binding of periplasmic ComE to DNA might exert force on the macromolecule, packaging incoming DNA and driving its translocation across the outer membrane. A role for ComE in uptake across the outer membrane was already proposed, however in conjunction with a retractable competence pilus (Chen and Dubnau, 2004). As shown for Agrobacterium tumefaciens binding of a ssDNA binding protein (VirE2) to DNA was sufficient to exert forces up to 50 pN, proposed to drive the directional transfer of the T-plasmid into plants (Grange et al., 2008). In analogy, ComE could be sufficient to establish a driving force for DNA translocation across the outer membrane. B. subtilis ComEA, anchored to the cytoplasmic membrane facing the external environment, is restricted in its movement in only two dimensions. It binds DNA but is probably devoid of packaging the macromolecule as efficiently as its soluble counterpart in Gram-negative bacteria (thereby leaving DNA accessible to DNaseI degradation). The latter feature is probably dispensable in the absence of an outer membrane, which presents an additional barrier to be overcome by incoming DNA in Gram-negative bacteria.

We speculate that the pilus fibre inserted into PilQ leads to maturation and, thus, functionality of PilQ. In this scenario, the pilus would serve as ‘placeholder’ and its retraction opens PilQ for DNA to bind at the central cavity of the pore. After specific selection of DNA by sequence (see below), periplasmic binding of ComE to DNA would pull the macromolecule across the outer membrane (Fig. 1B).

Certainly, more work is needed to clarify the complex role of pili in natural transformation. Our model can be challenged by laser tweezers experiments in the near future. If pili drive translocation across the outer membrane of Gram-negative bacteria, the rate of DNA uptake should be variable but peak in the range of velocities found for pilus retraction. Force will be definitely lost as a result of insufficient binding between DNA and the pilus fibre. In case DNA is pulled across the outer membrane by binding of ComE to the macromolecule, velocities would be expected to be much slower, while forces will be in the range of binding forces of ComE to DNA.

Sequence-specific selection of DNA during transformation

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References

Another important issue concerning DNA uptake is the fact that some Gram-negative bacteria (e.g. Neisseria) preferentially take up DNA from related bacteria, as they recognize a genus-specific sequence (DNA uptake signal, DUS). The identity of the DUS receptor is enigmatic. It is clearly absent from the pili extruding the cell body, and the secretin PilQ as well as the periplasmic DNA binding protein ComE do not discriminate DNA by sequence. In Neisseria, a pseudopilin, ComP, was shown to enhance DUS-dependent uptake in a dose-dependent fashion, but itself did not bind DUS-containing DNA in vitro (Aas et al., 2002a). It was concluded that ComP eventually stabilizes the DUS receptor. PilV was identified as an antagonist of ComP, as its absence led to increased ComP levels. Thus, PilV and ComP are thought to modulate DUS-dependent DNA uptake efficiency in Neisseria (Aas et al., 2002b).

In order to recognize a distinct DNA sequence, at least core parts accessible to DNA have to be highly conserved in the DUS receptor. Thus, the permanent presence of a conserved DUS receptor at the surface of the bacterium disagrees with surface structures underlying frequent variation. Because DUS is clearly involved in neisserial DNA uptake, i.e. crossing the outer membrane, the receptor may interact with PilQ at the inner side of the complex. Unspecific binding of DNA to PilQ might result in exposure of the DUS receptor at the surface. Subsequent selection for DUS at the receptor might be followed by relocalization of the DUS receptor and transfer of a small loop of DNA into the periplasm. This loop is putatively sufficient for further binding to ComE and final pulling of the macromolecule across the outer membrane (Fig. 1B). It is imaginable that the competence pilus is involved in active shuttling of the receptor upon DNA binding. The model explains why DUS has to be situated on the DNA molecule that is transported and why DUS-containing oligos do not promote non-DUS DNA uptake. Moreover, the observation that non-DUS DNA does not compete with DUS-containing DNA for uptake, but even led to enhanced transformation efficiency in some neisserial strains (Duffin and Seifert, 2010), might be explained by enhanced surface exposure of the DUS receptor upon unspecific binding of DNA to PilQ.

Type IV secretion system-mediated DNA translocation across the outer membrane

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References

In H. pylori, the VirB homologues system ComB was directly shown to mediate transport of dsDNA across the outer membrane (Fig. 1C) (Stingl et al., 2010). The crystal structure of outer membrane components of the VirB system revealed an extraordinary huge complex of 17.2 nm in diameter, bearing a central cavity of 3.2 nm (Chandran et al., 2009). Because one of the essential players, VirB2, proposed to coat the inner surface of this cavity, was missing in the structure, the channel might be somewhat narrower in vivo.

Using laser tweezers, we demonstrated for the first time substrate translocation through a type IV secretion system in real time (Stingl et al., 2010). Compared with what was known for B. subtilis, it became evident that a completely different process was monitored. DNA uptake in H. pylori occurred at velocities comparable with those measured for pilus retraction, however, at significantly lower forces. It was initially surprising that the DNA uptake process was reverted at high forces, meaning that previously imported DNA could be extracted from the bacterium. Most strikingly, the transport system was unaffected by rapid extraction of the macromolecule, because subsequent import upon release of external force was observed repeatedly. Uptake of covalently labelled fluorescent DNA, which was shown to cross the outer membrane but failed to cross the inner membrane, had similar kinetics as unmodified DNA. In contrast, more frequent pauses of the transport process and more frequent rupture of the bound molecule was observed, indicative of steric hindrance caused by the fluorophor at the pore complex. These results are consistent with a relatively large pore size and a loose contact of the outer membrane complex with the DNA substrate. How is DNA pulled across the outer membrane in H. pylori? A comE homologue is absent from its genome. However, the pilin VirB2 was shown to directly interact with DNA in Agrobacterium (Cascales and Christie, 2004) and it was shown to be dislocated from the cytoplasmic membrane by the ATPase VirB4 (Kerr and Christie, 2010). The VirB4 homologue in H. pylori ComB4 constitutes the only type IV secretion system ATPase involved in natural transformation (Karnholz et al., 2006). F-pilus retraction of a type IV secretion system accompanied by rotation about the long axis of the filament was demonstrated recently (Clarke et al., 2008). It remains to be shown, if the VirB2 homologue in H. pylori (ComB2) also interacts with DNA. Our laser tweezers data hint to a pilus retraction-like mechanism of DNA import and, given a direct interaction of ComB2 with DNA, translocation might be directly dependent on disassembly and/or rotation of ComB2 pilin subunits by the type IV secretion apparatus. Consistent with such a mechanism, the velocities were quite variable and highly sensitive towards alteration of external force. Loss of force during transduction of energy from the pilus disassembly machinery towards substrate translocation is expected concerning binding and unbinding events during translocation, in particular upon elevation of external pulling force.

The second step – DNA transport across the cytoplasmic membrane

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References

The second step is the transport of DNA across the cytoplasmic membrane, mediated by a highly conserved cytoplasmic membrane channel (ComEC) present in all known competent bacteria. In contrast to the outer membrane or the cell wall, the cytoplasmic membrane enables ion gradients to build up and to energize the cell. Hence, in contrast to the situation at the outer membrane, DNA cannot be transported in loose contact with the pore complex. ssDNA probably enters the cytoplasm of competent bacteria. For this purpose, a free DNA end appears to be necessary as a starting point (Provvedi et al., 2001). While the second strand is degraded (as shown for B. subtilis and Haemophilus influenzae (Barany et al., 1983; Provvedi et al., 2001) the first ssDNA strand enters the cytoplasm. From experiments with Streptococcus pneumoniae and H. influenzae, it was concluded that transport across the cytoplasmic membrane proceeds with 3′-5′ direction, leading to a free 3′ end for initiating recombination (Barany et al., 1983; Méjean and Claverys, 1993). For B. subtilis similar experiments failed in showing directional translocation of ssDNA into the cytoplasm (Vagner et al., 1990).

A fluorophore covalently attached to DNA affects the size and the structure of the macromolecule. Covalently labelled fluorescent DNA was able to cross the outer membrane, but was excluded from the cytoplasm (Stingl et al., 2010). The latter was evident for both H. pylori and B. subtilis. Thus, crossing the cytoplasmic membrane is highly specific for the DNA substrate. The fact that all transformable bacteria use the same membrane channel (ComEC, named ComA in Neisseria) might be an indication for the complexity of the process, which does not leave much variation (Fig. 1). The highly conserved membrane channel ComEC is best studied in B. subtilis (Draskovic and Dubnau, 2005). It probably functions as an oligomer and multiply traverses the membrane. The competence domain, which is conserved in competent bacteria, was proposed to form the water-filled channel, through which DNA traverses into the cytoplasm. Species-specific variations in other parts of the protein might reflect different adaptations to diverse interaction partners of ComEC during function.

What are the characteristics of DNA import into the cytoplasm? Using laser tweezers, single DNA uptake events were followed in B. subtilis (Maier et al., 2004). Like for outer membrane transport in H. pylori, the uptake events were highly processive, meaning that several µm of DNA tether were taken up. In contrast, B. subtilis DNA uptake was tenfold slower and occurred also against high pulling forces without reduction in velocity. Because the forces of DNA uptake in B. subtilis were in the range of those measured for pilus retraction, it was proposed that pseudopilus retraction itself is implicated in driving DNA import in B. subtilis (Maier et al., 2004). However, development of high forces is not restricted to pilus retraction. FtsZ and SpoIIIE, two completely different motor proteins involved in chromosome segregation and partitioning, as well as viral DNA packaging motors exert similar high forces (Smith et al., 2001; Pease et al., 2005; Ptacin et al., 2008). Genetic studies established that ComFA plays a role in DNA entry into the cytoplasm (Londono-Vallejo and Dubnau, 1994). ComFA is a DEAD-box helicase, which is important for DNA uptake but not binding. Inactivation of the Walker A motif present in ComFA abolished DNA uptake. Moreover, the amount of ComFA was decreased in a comEC mutant, indicating stabilization of ComFA by ComEC upon protein–protein interactions (Kramer et al., 2007). Hence, ComFA is a good candidate translocating DNA into the cytoplasm in B. subtilis. A putative role of ssDNA binding proteins in the cytoplasm in preventing back diffusion of DNA was also discussed (Chen and Dubnau, 2004). But unlike for ComFA, direct evidence for a role in uptake rather than protection of the imported molecule against degradation is still missing for ssDNA binding proteins. comFA is absent from the genome of most other bacteria. However, the close homologue PriA is ubiquitously found and involved in restart of chromosomal replication upon stall of the replication fork as well as homologous recombination and DNA repair. PriA was shown to be involved in neisserial DNA transformation, although it has to be clarified whether it directly plays a role in DNA import or rather in downstream processes (Kline and Seifert, 2005).

What is the energy source for DNA uptake?

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References

Very early studies indicated a role of the proton motive force (pmf), in particular the proton gradient, as driving force of DNA uptake into the cytoplasm of B. subtilis (van Nieuwenhoven et al., 1982). This led to the proposition of a H+/DNA symport mechanism. In contrast, ComEA-dependent binding of DNA to B. subtilis was unaffected by the presence of the pmf. Single molecule uptake experiments showed that the DNA motor stopped upon addition of a protonophor (dissipating the complete proton motive force), while the ATP level was maintained throughout the experiment (Maier et al., 2004). In the latter experiment, it remains unclear whether the DNA uptake complex was compromised by lack of an energized membrane. Localization of polar residues within membrane proteins is sensitive towards membrane polarization. However, the fact that the ATP level was maintained for some time in B. subtilis in the absence of a pmf, strengthens the conclusions of van Nieuwenhoven that a proton gradient is essential for DNA uptake. Still, the data do not exclude an additional role of ATP in energization of the process.

What serves as energy source for transfer of DNA to the cytoplasmic membrane? Obviously, the pmf is not involved in this process in B. subtilis, as binding was unaffected by dissipation of the pmf (van Nieuwenhoven et al., 1982). As mentioned above, the role of ATP is unclear. In H. pylori, short preincubation of the cells with an inhibitor of the F0F1-ATP-synthase completely abolished DNA uptake across the outer membrane as visualized using fluorescently labelled DNA (Stingl et al., 2010). Because H. pylori is unable to gain ATP by fermentation, the addition of an ATP-synthase inhibitor leads to immediate dissipation of the ATP level (unpublished results). Thus, the data are consistent with our model of an active outer membrane DNA translocation process, like e.g. a retraction mechanism by disassembly of ComB2 subunits via the ComB4 ATPase (Fig. 1C). However, we cannot exclude that loss of ATP also affected constant turnover of the ComB system. Single molecule experiments will be valuable in order to clarify whether ATP really serves as energy source for DNA uptake in H. pylori.

Are the two steps of DNA uptake spatially and temporally uncoupled?

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References

Data on B. subtilis indicated that DNA uptake takes place at the cell poles (Hahn et al., 2005). Moreover, by fusion of GFP variants to proteins, important for DNA transfer across the cell wall or across the cytoplasmic membrane, colocalization of these proteins was detected (Kramer et al., 2007). Strong interactions were confirmed by protein destabilization upon knock-out of single components. For example, the helicase and putative motor for DNA translocation across the cytoplasmic membrane, ComFA, stabilized the traffic ATPase ComGA, involved in assembly of the pseudopilus. Thus, in B. subtilis, the protein complexes involved in either of the two steps appear to be intimately associated. A recent study corroborated the idea that cell wall and cytoplasmic membrane constituents of the B. subtilis DNA translocation machinery form a huge complex at the cell poles that is relatively static (Kaufenstein et al., 2011).

It remains to be shown whether the pseudopilus is dispensable for DNA translocation into the cytoplasm of B. subtilis, e.g. by characterization of DNA uptake into protoplasts lacking ComGC. In contrast, a comEC mutant, defective in DNA translocation into the cytoplasm, still binds DNA in a ComEA-dependent manner. Because ComEA is anchored to the cytoplasmic membrane (Inamine and Dubnau, 1995), it has to be concluded that the first step of DNA uptake, i.e. the transfer of DNA across the cell wall, functions independently from the second. ComEA proteins localize exceptionally dispersed (Hahn et al., 2005) and constitute a relatively mobile protein fraction (Kaufenstein et al., 2011). Interestingly, cytoplasmic RecA proteins, important for homologous recombination of imported DNA, were detected only at one cell pole, preferentially harbouring the larger DNA translocation complex (Kaufenstein et al., 2011). Hence, a tentative model would include: (i) translocation of DNA across the cell wall by several translocation complexes, (ii) buffering DNA at the cell membrane by recruitment of ComEA proteins and (iii) subsequent activation of only one ComEC complex for translocation of DNA across the cytoplasmic membrane. If this model holds true, it will be of special interest to decipher the criterion for the decision, at which cell pole uptake of DNA into the cytoplasm occurs.

Also in H. pylori, imported DNA preferentially localized at the cell poles, indicating similar spatial distribution of the DNA uptake machineries as in B. subtilis (Stingl et al., 2010). In addition, these results showed that multiple DNA uptake machineries were concomitantly active in one cell (at least concerning outer membrane DNA transport). Via fluorescent visualization we observed that the temporal persistence of DNA in the periplasm was variable in wild-type H. pylori. In the comEC mutant, blocked for inner membrane transport, DNA was stably detected in the periplasm. These data showed that outer and inner membrane DNA transports are mechanistically and temporally uncoupled in H. pylori. As for B. subtilis, the first step is obviously independent from the second. In contrast, any putative contribution of the ComB system to ComEC-dependent DNA translocation into the cytoplasm remains elusive. For other Gram-negative bacteria, like Neisseria and Campylobacter, the situation appears to be similar. Knock-out of the comEC homologue probably led to accumulation of DNA in the periplasm in Neisseria from which unprocessed linear plasmid was reisolated (Facius et al., 1996). A similar mutant in Campylobacter was capable of DNA uptake (probably into the periplasm) but devoid of natural transformation, which remains to be confirmed (Jeon et al., 2008). Hence, this indicates a general principle that at least the first step in DNA uptake is functionally independent of the second. Because the second step is by nature dependent on DNA delivery by the first, it will be a challenge to unambiguously show that the cytoplasmic DNA translocation system functions independently.

Perspectives

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References

The investigation of the mechanism of bacterial DNA uptake profits from the integration of data from diverse disciplines like molecular biology, biophysics and biochemistry. Notably, genetic knock-out, conditional as well as site-directed mutants together with information about protein–protein and protein–DNA interactions can give valuable information. This is in particular true, when the data are combined with results from biophysical approaches, like fluorescent and single molecule analysis. Principle obstacles for detailed studies of DNA uptake are: (i) the limited number of active sites for uptake per cell at a distinct time point and (ii) the lack of an active DNA translocation system that is reconstituted by purified components. Interdisciplinary work will open up new vistas in the future but will strongly depend on open-mindedness of researchers keeping the fascination for each others scientific background.

References

  1. Top of page
  2. Summary
  3. The first step – transfer of DNA from the surface to the cytoplasmic membrane
  4. Sequence-specific selection of DNA during transformation
  5. Type IV secretion system-mediated DNA translocation across the outer membrane
  6. The second step – DNA transport across the cytoplasmic membrane
  7. What is the energy source for DNA uptake?
  8. Are the two steps of DNA uptake spatially and temporally uncoupled?
  9. Perspectives
  10. Acknowledgements
  11. References