Common Mechanisms of DNA translocation motors in Bacteria and Viruses Using One-way Revolution Mechanism without Rotation

Abstract

Biomotors were once classified into two categories: linear motor and rotation motor. For decades, the viral DNA-packaging motor has been popularly believed to be a five-fold rotation motor. Recently, a third type of biomotor with revolution mechanism without rotation has been discovered. By analogy, rotation resembles the Earth rotating on its axis in a complete cycle every 24 hours, while revolution resembles the Earth revolving around the Sun one circle per 365 days (see animations http://nanobio.uky.edu/movie.html). The action of revolution that enables a motor free of coiling and torque has solved many puzzles and debates that have occurred throughout the history of viral DNA packaging motor studies. It also settles the discrepancies concerning the structure, stoichiometry, and functioning of DNA translocation motors. This review uses bacteriophages Phi29, HK97, SPP1, P22, T4, T7 as well as bacterial DNA translocase FtsK and SpoIIIE as examples to elucidate the puzzles. These motors use a ATPase, some of which have been confirmed to be a hexamer, to revolve around the dsDNA sequentially. ATP binding induces conformational change and possibly an entropy alteration in ATPase to a high affinity toward dsDNA; but ATP hydrolysis triggers another entropic and conformational change in ATPase to a low affinity for DNA, by which dsDNA is pushed toward an adjacent ATPase subunit. The rotation and revolution mechanisms can be distinguished by the size of channel: the channels of rotation motors are equal to or smaller than 2 nm, whereas channels of revolution motors are larger than 3 nm. Rotation motors use parallel threads to operate with a right-handed channel, while revolution motors use a left-handed channel to drive the right-handed DNA in an anti-parallel arrangement. Coordination of several vector factors in the same direction makes viral DNA-packaging motors unusually powerful and effective. Revolution mechanism avoids DNA coiling in translocating the lengthy genomic dsDNA helix could be advantage for cell replication such as bacterial binary fission and cell mitosis without the need for topoisomerase or helicase to consume additional energy.

1. Introduction

The importance of nanomotors (Grigoriev et al.,2004) for cells or nanotechnology is akin to that of mechanical motors to daily life. Mechanical motors power cars to drive us to destinations, and nanobiomotors translocate DNA and RNA to facilitate biological functions. Extensive studies on nanobiomotors have resulted in not only many fabulous and marvelous findings, but also conjectures, puzzles, mysteries, and even zealous debates and disputes. For decades, the popular viral DNA-packaging motor (Fig. 1) has been believed to be a rotation motor with double-stranded (ds) DNA serving as a bolt and the protein channel as a nut (Hendrix,1978). The finding of the twisted connector structure, as revealed by crystal studies (Guasch et al.,2002; Simpson et al.,2001), seemingly strengthen the popular believing of a rotation mechanism in the scientific community. However, if the DNA packaging motor (Fig. 1) were considered to be a rotation machine, either the motor or the DNA would have to rotate during dsDNA advancing. But extensive studies have revealed that neither the channel nor the dsDNA rotates during transportation. When the connector and the procapsid protein were fused to each other and the rotation of the connector within the procapsid was not possible, the motors were still active in packaging, implying that connector rotation is not necessary for DNA packaging (Baumann et al.,2006; Maluf and Feiss,2006). Studies using single-molecule force spectroscopy combined with polarization spectroscopy have also indicated that the connector did not rotate (Hugel et al.,2007). Tethering of DNA ends to bead clusters demonstrated that DNA translocation by the motor was still active, and no rotation of the beads was observed either (Chang et al.,2008; Shu et al.,2007) (Fig. 2). The possibility of undetectable of bead rotation due to bond freedom between bead and dsDNA was excluded by applying bead clusters (Chang et al.,2008; De-Donatis et al.,2013).

Fig. 1.

Depiction of the structure and function of phi29 DNA packaging motor. A. Schematic of hexameric pRNA. B. AFM images of loop-extended hexameric pRNA. C. Illustrations of the phi29 DNA packaging motor in side view (left) and bottom view (right) (adapted from (Schwartz et al.,2013a) with permission of Elsevier). D and E. Illustration of the rotation and revolution motions. (adapted from (De-Donatis et al.,2013)).

Fig. 2.

Experiment of T4 and phi29 motor revealing that neither connector nor dsDNA rotation is required for active DNA packaging. A. Direct observation of DNA packaging horizontally using a dsDNA with its end linked to a cluster of magnetic beads for stretching the DNA. “a” and “b” are real-time sequential images of DNA-magnetic beads complexes (adapted from (Chang et al.,2008) with permission from the AIP publishing group). B. Schematic of the single molecule experimental geometry of phi29 DNA packaging motor to show no rotation of the connector during DNA packaging (adapted from (Hugel et al.,2007)). C. Experiment revealing that T4 motor connector does not rotate during packaging. The packaging activity is not inhibited with N-terminal of motor connector protein fused and tethered to its protease immune binding site on capsid (adapted from (Baumann et al.,2006) with permission from John Wiley and Sons).

Many mysteries have since been encountered in the course of the study of dsDNA translocation motors (Chen and Guo,1997b; Grimes and Anderson,1997; Guo et al.,1998; Hendrix,1978; Moffitt et al.,2009; Serwer,2003; Serwer,2010). Many thorough reviews on the packaging of dsDNA viruses are available (Casjens,2011; Guo and Lee,2007; Lee et al.,2009; Sun et al.,2010; Yu et al.,2010). Tetrameric, pentameric and hexameric motor models have been proposed for the motor structures and motor mechanisms (Guo et al.,1998; Moffitt et al.,2009; Ortega and Catalano,2006; Shu et al.,2007; Sun et al.,2007). Recently, a third type of biomotor has been discovered in the phi29 DNA packaging motor which exercise a revolution mechanism without rotation (Fig. 1) (for animation see http://nanobio.uky.edu/movie.html). The finding has solved many mysteries and puzzles. This review intends to address specifically the revolution mechanism and how this finding reconciles the data reported during the last 35 years, the questions of how the motor transports close circular dsDNA without breaking any covalent-bonds or changing the DNA topology, and how the motor controls a one-way DNA movement despite the intrinsic bidirectional nature of the dsDNA helix.

2. The ASCE biomotors including the AAA+ and FtsK-HerA superfamilies display a hexameric arrangement

It has been a long and fervent debate over whether the DNA packaging motor of dsDNA viruses contains six or five copies of components to drive the motor. Viral structure studies reveal that viral icosahedron capsid is composed of many pentamers and hexamers (Caspar and Klug,1962). It has been found that the DNA packaging motor of dsDNA viruses resides within the pentameric vertex (Bazinet and King,1985). Since the motor channel for dsDNA bacteriophages is a dodecamer (Agirrezabala et al.,2005; Cardarelli et al.,2010; Carrascosa et al.,1982; Doan and Dokland,2007; Kochan et al.,1984; Valpuesta et al.,1992), this 12-fold symmetry can be interpreted as a hexamer of dimers. Such special structural arrangement has, since 1978, led to the popularity of a novel rotation motor with a five-fold/six-fold mismatch gearing mechanism (Hendrix,1978). Recent publications of the hexameric pRNA crystal structure (Zhang et al.,2013), the AFM images of the hexameric pRNA ring (Shu et al.,2013a; Shu et al.,2013b; Zhang et al.,2013), and the finding of hexameric ATPase gp16 (Schwartz et al.,2013a), have resolved the debate. More importantly, finding of the revolution mechanism without rotation (Guo et al.,2013b; Schwartz et al.,2013b; Schwartz et al.,2013a; Zhao et al.,2013) helps resolve and conciliate the hexamer and pentamer discrepancy, since motors with any subunit stoichiometry can utilize the revolution mechanism. However, the use of a revolution mechanism by a hexamer motor has raised an intriguing question: Why do most ASCE (additional strand catalytic E) biomotors including the AAA+ (ATPases Associated with diverse cellular Activities) and FtsK-HerA superfamilies display a hexameric but not pentameric arrangement?

One fundamental process of all living organisms is to transport dsDNA. The ASCE superfamily including the AAA+ and FtsK-HerA superfamilies, is a clade of nanomotors that facilitates a wide range of functions (Ammelburg et al.,2006; Guo and Lee,2007; Snider et al.,2008; Snider and Houry,2008), many of which are involved in dsDNA riding, tracking, packaging, and translocation. These functions are critical to DNA repair, replication, recombination, chromosome segregation, DNA/RNA transportation, membrane sorting, cellular reorganization, cell division, bacterial binary fission and many other processes (Ammelburg et al.,2006; Martin et al.,2005). Despite their functional diversity, a common characteristic of this family is their ability to convert energy obtained from binding of ATP or hydrolysis of the γ-phosphate bond of ATP into a mechanical force (Chemla et al.,2005; Guo et al.,1987c; Hwang et al.,1996; Sabanayagam et al.,2007; Schwartz et al.,2012). This usually involves the conformational changes of an ATPase enzyme. The change in the conformation generates a gain or a loss of conformational entropy and the affinity for its substrate, thus a mechanical movement (De-Donatis et al.,2013). This movement is in turn used to either make or break contacts between macromolecules, resulting in local or global protein unfolding, complex assembly or disassembly, or translocation of DNA, RNA, proteins or other macromolecules (Guenther et al.,1997; McNally et al.,2010). The feature of these nanomotors is the hexameric arrangement that members of this family often display (Aker et al.,2007; Chen et al.,2002; Mueller-Cajar et al.,2011; Wang et al.,2011; Willows et al.,2004). The hexamer arrangement facilitates bottom-up assembly in nanomachine manufacturing and produce stable structure engagements and robust machines that can be artfully functionalized in human cells to remedy functional defects.

Different models have been proposed for the motor of dsDNA viruses (Aathavan et al.,2009; Guo and Lee,2007; Moffitt et al.,2009; Serwer,2010; Yu et al.,2010; Zhang et al.,2012). Until recently, it has been believed that they contain a five-fold/six-fold mismatch structure (Hendrix,1978). In 1986, the phi29 DNA packaging motor was constructed in vitro (Guo et al.,1986), and has been found to have three co-axial rings: pRNA, connector, and gp16 ATPase ring (Guo et al.,1987a; Guo et al., Fujisawa et al.,1991; Morita et al 1993; Ibarra et al.,2001; Lee and Guo,2006) (Fig. 1). In 1998, the pRNA ring was determined to exist as a hexameric ring (Guo et al.,1998; Zhang et al.,1998) (featured by Cell (Hendrix,1998)). In 2000, it was verified by Cryo-electron microscopy (Cryo-EM) to be hexameric in shape (Ibarra B et al.,2000). But studies by others have put forward a pentameric model (Chistol et al.,2012; Morais et al.,2008; Yu et al.,2010). However, biochemical analysis (Guo et al.,1998; Hendrix,1998; Zhang et al.,1998), single molecule photobleaching study (Shu et al.,2007), gold labeling imaging by electron microscopy (EM) (Moll and Guo,2007; Xiao et al.,2008), and RNA crystal structure studies (Zhang et al.,2013) have all revealed hexameric assembly of pRNA. One interesting theory has been proposed that the motor initially assembles as a hexamer but one of the subunits departed before DNA packaging starts, thus generating a pentamer (Morais et al.,2001; Morais et al.,2008; Simpson et al.,2000). However, single molecule photobleaching analysis of DNA-packaging intermediates showed that the active motor still contained six copies of pRNA during DNA translocation (Shu et al.,2007) (Fig. 3), and pRNA dimers were the building blocks for hexameric ring, which is assembled through the pathway of 2 → 4 → 6 pRNAs.

Fig. 3.

Single molecule photo-bleaching and confirmation of the presence of six copies of phi29 motor pRNA vial dual-view imaging of procapsids containing three copies of Cy3-pRNA and three copies of Cy5-pRNA. A. pRNA dimer design constructed with Cy3- and Cy5-pRNA. B. Typical fluorescence image of procapsids with dual-labeled pRNA dimers. C. Comparison of empirical photobleaching steps with theoretical prediction of Cy3-pRNA in procapsids bound with dual-labeled dimers. D. Photobleaching steps of procapsids reconstituted with the dimer. (adapted from (Shu et al.,2007) with permission from John Wiley and Sons).

The formation of gp16 into an active hexameric complex in the phi29 DNA packaging has been further demonstrated by using a Walker B mutant gp16 that could bind but not hydrolyze ATP, as the activity of the assembly containing a different number of mutant monomers followed a binomial distribution model (Chen et al.,1997; Schwartz et al.,2013a; Trottier and Guo,1997). Empirical results have been obtained in conjunction with stoichiometric data to show that the motor complex is hexameric (Schwartz et al.,2013a). Moreover, qualitative DNA binding assays, capillary electrophoresis assays (CE), and electrophoretic mobility shift assays (EMSA) have demonstrated that the final oligomeric state of ATPase is hexamer (Schwartz et al.,2013a) (Fig. 4).

Fig. 4.

Stoichiometric assays showing the formation of the hexamer of phi29 motor ATPase gp16. A. Native gel revealed six distinct bands characteristic of the six oligomeric states of the ATPase; the hexamer increases as the concentration of protein increased. B. Slab gel showing the binding of ATPase to dsDNA in a 6:1 ratio; imaged in GFP (upper) and Cy3 channels (lower) for ATPase and dsDNA, respectively. C. Quantification by varying the molar ratio of [ATPase]:[DNA]. The concentration of bound DNA plateaus at a molar ratio of 6:1 (adapted from (Schwartz et al.,2013a) with permission from the Elsevier)

There has been some discrepancies over the motor stoichiometries and the mechanisms. A compression mechanism found in the T4 DNA packaging motor (Dixit et al.,2012; Ray et al.,2010) agrees with the revolution mechanism and the “Push through a One-Way Valve” model as previously described (Guo and Lee,2007; Jing et al.,2010; Zhang et al.,2012; Zhao et al.,2013). There seems to be another discrepancy over the pentamer model of the phi29 motor, as other studies have demonstrated four steps, instead of five, of motor action, as revealed by laser trap experiments (Fig. 5A). To reconcile these discrepancies, the group proposed a model in which one of the subunits in the pentamer ring is inactive, and the other four subunits are active during the DNA packaging process (Moffitt et al.,2009; Yu et al.,2010). This four steps have been recently shown to be a result from the four relaying lysine layers embedded inside the phi29 connector channel inner wall (Fig. 5B, 5C) (Zhao et al.,2013).

Fig. 5.

Experiment and the proposed mechanism revealing four steps of pauses for each circle during the packaging of phi29 dsDNA. A. Design (left) and results (right) of four sample packaging traces. (adapted from (Chistol et al.,2012) with permission of the Elsevier). B. The presence of four lysine residues of motor channel protein leads to the formation of four positively charged rings in different motors (adapted from (Fang et al.,2012) with permission from Elsevier). C. Diagram showing DNA revolution inside phi29 connector channel with four steps of pause due to the interaction of four positively charged lysine rings with the negatively charged dsDNA phosphate backbone. DNA revolution across the 12 connector channel subunits is shown. The uneven electrical static interaction resulting from the mis-gearing between 10.5-base pairs per DNA helical turn and 12 subunit per round of the channel is the cause leading to the observation of uneven steps (adapted with permission from (Zhao et al.,2013). Copyright (2013) American Chemical Society). (PDB IDs: Phi29-gp10, 1H5W. (Guasch et al.,2002); SPP1-gp6, 2JES. (Lebedev et al.,2007); P22-gp1, 3LJ5. (Olia et al.,2011)).

3. A motor with a revolution rather than rotation mechanism in dsDNA translocation has been discovered in bacteriophage phi29

Viral dsDNA packaging motors consist of a protein channel and two packaging components (Fig. 1). A discovery made 25 years ago has shown that the larger component serves as part of the ATPase complex, and the smaller one is responsible for dsDNA binding and cleavage (Guo et al.,1987c; Guo et al.,1998). This notion has now been well-grounded. The motor connector contains a center channel encircled by 12 copies of the protein gp10 that serves as a pathway for dsDNA translocation (Guasch et al.,2002; Jimenez et al.,1986). Besides the well-characterized connector channel core, the motor of bacterial and a hexameric packaging RNA ring (Guo et al.,1987a; Guo et al.,1998; Shu et al.,2007; Zhang et al.,2013) (Fig. 1). During packaging, ATP binds to one ATPase subunit and stimulates ATPase to adapt a conformation, with a conformational entropy alteration (De-Donatis et al.,2013), resulting in a higher affinity to dsDNA. Upon ATP hydrolysis, the ATPase virus phi29 involves an ATPase gp16 (Grimes and Anderson,1990; Guo et al.,1987b; Guo et al.,1987c; Ibarra et al.,2001; Lee and Guo,2006; Lee et al.,2008) assumes a new conformation, and another conformational entropy resume, resulting in a lower affinity to dsDNA, thus pushing dsDNA away from the subunit and transferring it to an adjacent subunit by a power stroke (Fig. 6) (Schwartz et al.,2012; Schwartz et al.,2013b; Schwartz et al.,2013a). EMSAs have shown that the motor ATPase gp16's affinity towards dsDNA increases in the presence of ATPγs (Fig. 6) but remains low in the presence of ADP, AMP, or in the absence of nucleotide (Schwartz et al.,2012). Cooperativity and sequential action among the ATPase subunits also promote the revolution of dsDNA along the connector channel (Chen and Guo,1997a; Moffitt et al.,2009), as determined by Hill constant and binomial behavior of the activity in the inhibition assays due to incorporation of mutant subunits (Fig. 7). The DNA revolves unidirectionally along the hexameric channel wall, but neither the dsDNA nor the hexameric ring rotates (Schwartz et al.,2013b; Zhao et al.,2013) (Fig. 8). One ATP is hydrolyzed in each transitional step and six ATPs are consumed in one cycle to translocate the dsDNA one helical turn of 360° (Schwartz et al.,2012; Schwartz et al.,2013b). The binding of ATPase on the channel wall along the same phosphate backbone chain, but at a location 60° different from the last contact, drives dsDNA forward 1.75 bp each step (10.5 bp/turn ÷ 6 ATP = 1.75 bp/ATP) and revolves it without any rotation of the ATPase or dsDNA (Schwartz et al.,2013b; Zhao et al.,2013) (Fig. 9).

Fig. 6.

Gel showing the binding of dsDNA with the ATPase under different conditions, revealing the different conformations and different dsDNA affinities of the ATPase upon ATP binding and hydrolysis (adapted from (Schwartz et al.,2013a) with the permission of Elsevier).

Fig. 7.

Inhibition assays revealing stoichiometry of motor components and cooperativity of motor subunits. A. Binomial distribution assay reveals that ATPase gp16 possesses a 6-fold symmetry (highlighted) in the DNA packaging motor (adapted from (Schwartz et al.,2013a) with the permission of Elsevier). B. Binomial distribution assay demonstrating one inactive Walker B ATPase mutant subunit within the hexamer blocked motor activity. C. The inhibition ability of inactive Walker B mutants in the absence (C, E) and presence (D, F) of dsDNA reveals cooperativity of motor subunits, supporting the sequential action and revolution mechanism (adapted from (Schwartz et al.,2013b) with the permission of Elsevier).

Fig. 8.

Schematic showing the mechanism of sequential revolution in translocating dsDNA. A. The binding of ATP to one ATPase subunit stimulates the ATPase gp16 to adopt a conformation with a higher affinity for dsDNA. ATP hydrolysis forces gp16 to assume a new conformation with a lower affinity for dsDNA, thus pushing dsDNA away from the subunit and transferring it to an adjacent subunit. Rotation of neither the hexameric ring nor the dsDNA is required since the dsDNA revolves around the diameter of the ATPase. In each transitional step, one ATP is hydrolyzed, and in one cycle, six ATPs are required to translocate dsDNA one helical turn of 360° (10.5 bp). An animation is available at http://nanobio.uky.edu/movie.htm. (adapted from (Schwartz et al.,2013b) with permission of the Elsevier). B. Diagram of CryoEM results showing offset of dsDNA in the channel of bacteriophage T7 DNA packaging motor. The dsDNA did not appear in the center of the channel, instead, the dsDNA tilted toward the wall of the motor channel (adapted from (Guo et al.,2013a) with the permission of the National Academy of Sciences). C. The revolution of dsDNA along the 12 subunits of the connector channel (adapted from (Schwartz et al.,2013b) with permission of the Elsevier).

Fig. 9.

Model of sequential mechanism of sequence action of phi29 DNA packaging motor. Binding of ATP to the conformationally disordered ATPase subunit stimulates an entropic and conformational change of the ATPase, thus fastening the ATPase at a less random configuration. This lower entropy conformation enables the ATPase subunit to bind dsDNA and prime ATP hydrolysis. ATP hydrolysis triggers the second entropic and conformational change, which renders the ATPase into a low affinity for dsDNA thus pushing the DNA to the next subunit that has already bound ATP. These sequential actions promote the movement and revolution of the dsDNA around the hexameric ATPase ring.

An anti-parallel arrangement (an opposite handedness) exists between the dsDNA helix and the channel subunits of both the connector dodecamer, as revealed by crystallography (Guasch et al.,2002; Simpson et al.,2001). The connector helix has 12 subunits that display a tilt of 30°. Since one turn of the helix of dsDNA advances 360°, the dsDNA shifts 30° every time it passes one of the 12 subunits (360° ÷ 12 = 30°), and the angle is compensated for by the 30° tilting of the subunit of channel wall (Fig. 1C), resulting in the transportation of the helices without rotation, coiling, or a torsion force. The contact between the connector and the dsDNA chain is transferred from one point on the phosphate backbone to another along one strand in a 5′ to 3′ direction (Moffitt et al.,2009; Schwartz et al.,2013b; Zhao et al.,2013).

An effective mechanism of coordination between gp16 and dsDNA operates by using the ATP hydrolysis cycle as means of regulation. Motor ATPase tightly clinches dsDNA after binding to ATP and subsequently pushes the dsDNA away after ATP hydrolysis. Only one molecule of ATP is sufficient to generate the high affinity state for DNA in the ring of the motor ATPase. The mixed oligomers of the wild type and the mutants display a negative cooperativity and a communication between the subunits of gp16 oligomer (Schwartz et al.,2012; 2013b; 2013a). In the presence of dsDNA, a rearrangement occurs within the subunits of gp16 that enables them to communicate between each other and to “sense” the nucleotide state of the reciprocal subunit. This observation supports that only the subunit that binds to the substrate at any given time is permitted to hydrolyse ATP, thus performing translocation while the other subunits are in the “stalled” or “loaded” state. This suggests an extremely high level of coordination on the function of the protein, likely the most efficient process to couple energy production with DNA translocation via ATP hydrolysis among the biomotors studied so far (Fig. 9).

More recently, Kumar and Grubmuller reported their studies on the physical and structural properly of the connecter of phi29 DNA packaging motor (Kumar and Grubmuller,2014; see also the Commentary by Guo, 2014). Their new evidence based on independent biophysical studies supports the one-way revolution mechanism of biomotors. Using dynamic simulations, the elasticity and stiffness of connector channel was probed down to the atomic level. The author used their biophysical data to rule out untwist-twist DNA packaging model (Simpson et al.,2000) proposed previously. Their equilibrium and non-equilibrium force probe via MD simulations revealed that the connector as a whole is softer than the procapsid, yet the central region of the connector is one of the most rigid proteins known. Their calculation revealed that the previously proposed 12° untwisting requires ten times more energy than is available from the single hydrolysis of an ATP molecule (Kumar and Grubmuller,2014). Moreover, the umbrella sampling simulations suggest 100 kJ/mol is required for only a 4° untwisting of the entire connector. Kumar and Grubmuller's data also demonstrates why the connector is able to withstand a large pressure variation of up to ~60 atm. With data together made them conclude that the overall heterogeneity of the connector's mechanical properties aligns well with the newly resolved one-way revolution model reported (Schwartz et al.,2012; 2013b; 2013a). The authors pointed out that the connector acts as a one-way valve preventing leakage of DNA from the procapsid during packaging by the inner flexible connector loop to serve as a check-valve (Fang et al.,2012; Zhang et al.,2012; Zhao et al.,2013). The value and novelty of these conclusions is, not only in the solid data, but the authors’ rational interpretation of the mechanism following data acquisition. They did not simply follow the literature; instead, they used their own logical and insightful view to piece together their findings independently (Kumar and Grubmuller,2014).

4. The dsDNA packaging motor of many, if not all, dsDNA bacteriophage uses the same revolution mechanism without rotation

All the dsDNA viruses known so far use a similar mechanism to translocate their genomic DNA into preformed protein shells, termed procapsids, during replication (For reviews, see Guo and Lee,2007; Rao and Feiss,2008; Serwer,2010; Zhang et al.,2012). It has been reported that the motors of bacteriophages phi29, HK97, SPP1, P22 and T7 share a common revolution mechanism without rotation (De-Donatis et al.,2013). The special structure of these motors composes several factors that coordinate to drive the dsDNA with a one-way traffic mechanism. The 30° left-handed twist of the channel wall produces an anti-parallel arrangement with the right-handed helix of the dsDNA, resulting in a one-orientation trend. The same anti-parallel arrangement with the 30° left-handed twist has also been found in motor channels of other dsDNA viruses including phi29 (Guasch et al.,1998), HK97 (Cardarelli et al.,2010), SPP1 (Lhuillier et al.,2009), P22 (Cingolani et al.,2002) and T7 (Agirrezabala et al.,2005) (see Section 8). Of these motor channels, the primary amino acid sequences were not conserved, while the 2D and 3D structures of the twisted swivel region are perfectly conserved and aligned (Fig. 11). Distinction of revolution from rotation by channel size and chirality are applicable to bacteriophages SPP1, HK97, P22, T4, T7 and phi29 (see Section 7 and 8). Unidirectional flow loops within the channel of SPP1 and phi29 promote one-directional processing for the one-way trafficking of dsDNA (Fig. 12). The electropositive lysine layers present in all phage channels interact with a single strand of the electronegative dsDNA phosphate backbone (Fig. 5B, 5C). The electrostatic interaction between lysine layers and phosphage background will result in a relaying contact that facilitates one-way motion and generation of transitional pausing during dsDNA translocation. Coordination of these vector forces in the same direction makes viral DNA-packaging motors unusually effective.

Fig. 11.

Quaternary structures showing the presence of the left-handed 30° tilting of the connector channel of different bacteriophages. A and B. The external view and the cross section view of the motor, showing the anti-parallel configuration between the left-handed connector subunits and the right-handed dsDNA helices. The 30° tilt of the helix (highlighted) relative to the vertical axis of the channel can be seen in a cross-section internal view of the connector channel and the view of its single subunit as shown in B (adapted from (Schwartz et al.,2013b) with permission from Elsevier, adapted from (Agirrezabala et al.,2005) with permission from Elsevier and adapted from (De-Donatis et al.,2013)). (PDB IDs: Phi29-gp10, 1H5W. (Guasch et al.,2002); HK97 family-portal protein, 3KDR; SPP1-gp6, 2JES. (Lebedev et al.,2007); P22-gp1, 3LJ5. (Olia et al.,2011), T7-gp8 EM ID: EMD-1231. (Agirrezabala et al.,2005)).

Fig. 12.

The role of the flexible inner channel loop in DNA one-way traffic. A. Flexible loops within the phi29 (left) and SPP1 (right) connector channels function to interact with DNA, facilitating DNA to move forward but blocking reversal of DNA during DNA packaging. B. Demonstration of one-way traffic of dsDNA through wild type connectors using a ramping potential and by switching polarity (right). C. SsDNA is translocated via two-way traffic with a loop deleted connector (adapted from (Zhao et al.,2013) with permission from American Chemical Society).

The revolution mechanism is in consent with the finding in many viruses revealing that their dsDNA is spooled inside the viral capsid which occurs naturally and automatically, free from rotation tangles (Jiang et al.,2006; Lander et al.,2006; Molineux and Panja,2013; Petrov and Harvey,2008) (Fig. 10). A random arrangement occurs at first, with or without a free end, followed by a more ordered orientation of DNA as it continues to enter the procapsid and coil tighter. In phi29, the dsDNA has been found to show as a ring of density encircling around the gate region of the connector and to traverse the connector-tail tube (Fig. 10A) (Duda and Conway,2008a; Tang et al.,2008). The alternative explanation of the observation of such toroidal structure revealed by cryo-EM might be the consequence of the averaging effect of DNA revolution, since cryoEM is a collective image of many isolated dsDNA densities, the accumulation of the density of individual revolving intermediates with DNA tilted to the channel wall will produce a toroidal image. The force generation mechanism of viral DNA packaging motors has been investigated by using the members of the ASCE family, which have Walker A and Walker B motifs. These motifs convert energy from the binding or hydrolysis of ATP into a mechanical force, a process that usually involves a conformational change of ATPase (see next section). The spooling phenomenon caused by revolution is also evident in some FtsK DNA pumps (Fig. 13,) (Demarre et al.,2013; Grainge,2013) .

Fig. 10.

Spooling of DNA within capsids of phages to support the revolution mechanism. The DNA Spooling inside the capsids are shown using example of A. phi29 bacteriophage (adapted from (Duda and Conway,2008b) with permission from Elsevier and (Tang et al.,2008) with permission from Elsevier), B. λ bacteriophage (adapted from (Petrov and Harvey,2011) with permission from Elsevier), C. ε 15 bacteriophage (adapted from (Jiang et al.,2006) with permission from Nature Publishing Group), D. T7 bacteriophage (adapted from (Cerritelli et al.,1997) with permission from Elsevier), and E. P22 bacteriophage (adapted and redrew from (Lander et al.,2006) and adapted from (Zhang et al.,2000) with permission from Elsevier).

Fig. 13.

Bacterial hexameric DNA translocase FtsK may use the revolution mechanism. A. Diagram showing the contact of one strand of the dsDNA to the inner channel wall of the hexameric ATPase. The adjacent contact between DNA and ATPase will move around the inner surface of the channel without any rotation of the ATPase or DNA. B. Each of DNA contact points is expected to be separated by 60° along the inner surface of the ATPase hexameric channel. C. Sequential action of dsDNA translocation. DNA is shown as a line. T represents for ATP-bound and D for ADP-bound. (adapted from (Crozat and Grainge,2010) with permission from John Wiley and Sons).

5. Cellular dsDNA translocases also use the revolution mechanism without rotation

The cellular components that show the high-level similarity to the hexameric revolution motor come from two families of bacterial motor proteins: The FtsK family, an ASCE DNA motor protein group that transports E. coli DNA and separates intertwined chromosomes during cell division (Iyer et al.,2004; Snider et al.,2008; Snider and Houry,2008); and the SpoIIIE family (Demarre et al.,2013), which is responsible for transportation of DNA from a mother cell into the pre-spore during the cell division of Bacillus subtilis (Bath et al.,2000; Grainge,2008). The FtsK and SpoIIIE DNA transportation systems rely on the assembly of a hexameric machine. Although the precise action mode of revolution for FtsK and SpoIIIE motor has not been pointed out, authors of these review strongly believe that available evidence explicitly demonstrated that both motors use a revolution mechanism to translocate dsDNA without rotation (Fig. 8 and 13). The translocation of dsDNA with 1.6-1.75 base per hexamer subunit of FtsK (Crozat and Grainge,2010; Massey et al.,2006) perfectly agrees with the quantification in phi29 and T3 DNA packaging motor that each ATPase subunit uses one ATP to package 1.75 nucleotide (Guo et al.,1987c; Schwartz et al.,2013b; Schwartz et al.,2013a; Zhao et al.,2013; ; Morita et al 1993). The model in which dsDNA touches the internal surface of the hexameric ring (Crozat and Grainge,2010) is in agreement with the revolution mechanism that only one strand, but not both, of the dsDNA touches the internal wall of the motor channel and revolves without rotation (Schwartz et al.,2013b; Zhao et al.,2013). The proposed model of 60° per step for FtsK hexamer (Crozat and Grainge,2010) is in agreement with the finding of 30° per step for the 12-subunit phi29 dodecameric connector channel (Fig. 11) and 60° per steps for the hexameric phi29 ATPase gp16 (Guo et al.,2013b; Schwartz et al.,2013b; Schwartz et al.,2013a; Zhao et al.,2013).

5.1. FtsK

FtsK is responsible for dsDNA bidirectional translocation and conjugation between bacterial cells (Burton and Dubnau,2010; Pease et al.,2005). It is a multi-domain protein consisting of a 600-amino acid linker, a C-terminal ATPase domain FtsK(C) containing α, β and γ domains, and a N-terminal membrane-spanning domain FtsK(N) (Aussel et al.,2002; Barre et al.,2000; Yu et al.,1998). The ability of FtsK to move on DNA molecules in vitro using ATP as the energy source suggests that it is a DNA motor protein (Pease et al.,2005). The crystal structure of FtsK(C) indicates a ring-like hexamer (Lowe et al.,2008; Massey et al.,2006). Electron microscopy (EM) of FtsK(C) has revealed a DNA-dependent hexamer formation and DNA passage through the hexameric ring. Based on these data, a rotary inchworm mechanism has been proposed for FtsK to translocate dsDNA (Crozat and Grainge,2010; Massey et al.,2006) with ATPase subunits acting in a sequential manner (Fig. 13). Hexameric FtsK(C) translocates DNA through its central channel where protein-DNA contacts involve one or two hexamers. During each cycle of ATP binding and hydrolysis within each FtsK subunit, one motif binds tightly to the helix while the other progresses forward along the dsDNA. Such process that causes translational movement of the dsDNA is repeated by the subsequent handing off of the helix to the next adjacent subunit (Massey et al.,2006). A monomer undergoes a catalytic cycle in which DNA translocates 1.6-1.75 bp with little rotation. Then DNA binds to the next subunit after the catalysis of a second subunit (Massey et al.,2006). The hexameric ring holds dsDNA, with one functional subunit contacting the DNA at a time. The monomer of the functional subunit experiences an ATP catalytic cycle and translocates DNA through the channel by the hinged movement of the α domain and the β domain. This action carries the DNA backbone to the next functional subunit inside the same ring by a sequential hand-off mechanism. It performs the same exercise of DNA-binding, catalytic cycling and translocation. Such a cycle of DNA translocation is facilitated by the interaction between helical structure of DNA and the functional subunit of the hexameric ring without rotation of the protein ring against the DNA (Massey et al.,2006). Furthermore, the cycle of DNA translocation may follow a sequential escort mechanism in which, at one catalytic step, multiple α and/or β domains interact with DNA, drag the strand, and release it until they change hands with the adjacent domains so that DNA is moved using the flexibility of the subunits and their loops (Crozat et al.,2010).

5.2. SpoIIIE

Like FtsK, SpoIIIE shares a conserved C-terminal domain that harbors three subdomains, α, β, and γ (Fiche et al.,2013; Kaimer and Graumann,2011). The α and β subdomains translocate DNA through ATP binding and DNA-dependent hydrolysis (Barre,2007), and the γ subdomain recognizes specific DNA sequences to guide DNA translocation (Lowe et al.,2008; Sivanathan et al.,2006). These subdomains assemble into a hexameric ring that accommodates dsDNA in its central channel (Massey et al.,2006). The SpoIIIE DNA translocase actively moves DNA into the spore during Bacillus subtilis sporulation (Bath et al.,2000; Sharp and Pogliano,2002). The DNA translocation seems to be conducted by two hexamers, one within each membrane. One hexamer appears to pump DNA at a time, and then the other takes over. This view of DNA translocation receives support from the visualization of two chromosomal sites with both arms of the chromosome having moved simultaneously into the forespore. Further characterization of the SpoIIIE translocation pore assembly at the sporulation septum suggests a stepwise process (Fleming et al.,2010). An unstable channel of DNA translocation is first formed by the transmembrane domain of SpoIIIE localized to the leading edge of the closing septum. This channel is stabilized by binding of DNA to the translocase domains, followed by DNA translocation through the separating membranes. The channel is destabilized with release of the DNA from the translocase domain. The destabilization generates a larger opening for the final loop of the circular chromosome to migrate into the forespore.

Recently, a conceptually alternative model for SpoIIIE/FtsK motors was proposed (Cattoni et al.,2013). In contradiction with the previous model in which FtsK hexamer assembly is KOPS (FtsK Orienting/Polarizing Sequence) dependent (Graham et al.,2010; Lowe et al.,2008), it was shown that SRS (Sequence Retrieval System) sequences do not have a role in protein assembly by Fluorescence Correlation Spectroscopy and Atomic Force Microscopy. This new model involves three sequential steps: (a) binding of pre-formed hexamers to non-specific DNA, which might involve conformational transitions of the protein hexamer from an open- (inactive) to a closed-ring structure (active), (b) specific localization of SpoIIIE to SRS by target search exploration, and (c) sequence-specific activation of the motor mediated by binding and oligomerization of SpoIIIE-γ on SRS.

6. Rotational dsDNA translocases with only one ssDNA strand within the channel

Some hexameric dsDNA translocation motors such as RuvA/RuvB involved in DNA homologous recombination (Han et al.,2006; Rafferty et al.,1996) are sophisticated, and some other dsDNA riding motors such as those involved in DNA repair only display two protein subunits. The mechanisms of these motors are so complicated that they go beyond the scope of this review and will not be addressed here. However, many of other dsDNA tracking ATPases such as DNA helicases are assembled into hexamers to rotate along nucleic acids with one strand of DNA or RNA passing through the channel. These rotation nanomotors include, but are not limited to, helicases (Bailey et al.,2007; Castella et al.,2006; Enemark and Joshua-Tor,2006), DNA polymerase (Kato et al.,2001), and RecA family proteins (Cox,2003; Xing and Bell,2004b), the latter of which plays a key role in replication, repair, and genetic recombination of DNA. Most rotation dsDNA translocases share the mechanisms that are distinct from the revolution dsDNA translocases.

6.1 Helicase DnaB

It has long been believed that the mechanism of translocation of certain DNA or RNA operates by a rotation movement. A recent study from the crystal structure of the bacterial replicative DNA helicase DnaB has revealed that DnaB operates in the mechanism similar to the phi29 DNA packaging motor concerning hexameric stoichiometry and one-way sequential action (Itsathitphaisarn et al.,2012), but distinct from in the following respects: 1) the DNA inside the hexameric channel is a ssDNA instead of a dsDNA; 2) the motion is rotation instead of revolution; 3) the channel is right-handed instead of left-handed; and 4) the channel is smaller than 3nm that is found in revolutionary motor. DnaB is a hexameric motor that can unwind dsDNA in front of the replication fork to provide ssDNA templates for the DNA polymerase III holoenzyme (Kaplan and Steitz,1999; LeBowitz and McMacken,1986). Based on the crystal structure of the DnaB hexamer in a complex with GDP-AIF4 and ssDNA, a hand-over-hand translocation mechanism has been proposed by the Steitz group (Fig. 14). In this mechanism, the sequential hydrolysis of NTP is coupled to the 5′-3′ translocation of the subunits with a step size of two nucleotides, while some other helicases use the 3′-5′ direction (Thomsen and Berger,2009). The sequential hand-by-hand migration of the individual subunits along the helical axis of an ssDNA template leads to DNA translocation (Fig. 14). Although the translocation substrate is ssDNA, it revolves as an A form structure; that is, the ssDNA is present in the complex as a helix similar to a dsDNA helix, albeit smaller in diameter, during translocation.

Fig. 14.

Sequential rotation mechanism of the hexameric DNA helicase DnaB. One strand of the dsDNA displaces within the hexameric channel and the other strand is outside the ATPase channel(adapted from (Itsathitphaisarn et al.,2012) with permission from Elsevier).

6.2 TrwB

TrwB structurally resembles ring helicases and F₁-ATPase because of its spherical nature and membrane (Gomis-Ruth et al.,2001; Gomis-Ruth and Coll,2001; Matilla et al.,2010; Tato et al.,2007). It is a homo-hexamer of the TraG-like family responsible for bacterial conjugation (Cabezon and de la,2006). Its single hexamer is 110 Å in diameter and 90 Å in height with a channel ~20 Å wide. A ring of asparagine residues plugs the channel at the other end resulting in narrow end ~8 Å in diameter (Gomis-Ruth et al.,2001). Crystallographic structural analysis has revealed that the cytoplasmic domain of TrwB is hexameric (Gomis-Ruth et al.,2001). Its structural similarities to other well-documented DNA motors suggest that TrwB may translocate ssDNA through its central channel with energy derived from ATP hydrolysis. The central channel in TrwB could be occupied by a single DNA strand, which might be transferred through the internal cavity in a 5′-3′ direction (Cabezon and de la,2006). This notion receives support from biochemical data showing that TrwB is a DNA-dependent ATPase (Tato et al.,2005) and a structure-specific DNA-binding protein (Matilla et al.,2010), likely to pump DNA by a “binding change mechanism” assigned to F₁-ATPase (Cabezon and de la,2006). The ATPase activity is stimulated by interacting with TrwA (Tato et al.,2007), a tetrameric oriT-specific DNA-binding protein (Moncalian and de la,2004). Maximal ATPase rates can be reached in the presence of TrwA and dsDNA (Tato et al.,2007). These findings lead to a model of DNA translocation that occurs toward the membrane through the TrwA-TrwB interaction.

6.3 RecA

The RecA family of ATP-dependent recombinases plays a crucial role in genetic recombination and dsDNA break repair in Archaea, Bacteria, and Eukaryota. Extensive biochemistry, genetics, electron microscopic X-ray crystallography studies have been carried out for the RecA family proteins (Dunderdale and West,1994; Kowalczykowski and Eggleston,1994). RecA can interacts with ssDNA to form right-handed filament with ssDNA as a complex, which is the only active form of the protein, with approximately six monomers of RecA per helical turn (Di et al.,1982). The interaction can be enhanced by the binding of ATP (Menetski and Kowalczykowski,1985). EM studies of the active filament has revealed that ATP binding triggers a re-orientation between the RecA ATPase domains, leading to the relative rotation of the protein on DNA substrate during the translocation of DNA by the energy of ATP hydrolysis.

7. Revolution and rotation mechanisms can be determined by motor channel size (Fig. 15)

Fig. 15.

Comparison of the size of channels between biomotors using rotation mechanism (left panel) and biomotors using revolution mechanism (right panel). The dsDNA biomotors with a channel size of twice the width of dsDNA argues against the a bolt and nut treading mechanism, supporting the conclusion of revolution rather than rotation (adapted from (Agirrezabala et al.,2005) with permission from Elsevier, adapted from (Sun et al.,2008) with permission from Elsevier and adapted from (De-Donatis et al.,2013)). (PDB IDs: RepA, 1G8Y. (Niedenzu et al.,2001); TrwB: 1E9R. (Gomis-Ruth et al.,2001); ssoMCM, 2VL6. (Liu et al.,2008); Rho, 3ICE. (Thomsen and Berger,2009); E1, 2GXA. (Enemark and Joshua-Tor,2006); T7-gp4D, 1E0J. (Singleton et al.,2000); FtsK, 2IUU. (Massey et al.,2006); Phi29-gp10, 1H5W. (Guasch et al.,2002); HK97 family-portal protein, 3KDR; SPP1-gp6, 2JES. (Lebedev et al.,2007); P22-gp1, 3LJ5. (Olia et al.,2011). T7-gp8 EM ID: EMD-1231. (Agirrezabala et al.,2005)).

If dsDNA advances through the center of the motor channel, the channel needs to have a size similar to the diameter of dsDNA. Thus, the diameter of the channel should be close to 2 nm for dsDNA or 1-2nm for ssDNA so that DNA can touch the channel wall while passing through the channel. For revolution motors, however, dsDNA advances by touching the channel wall instead of proceeding through the center of the channel. Thus, the channel should be much larger than the diameter of the DNA so that there is sufficient room for revolution. By inspecting the crystal structures of different motors, the diameter of the channel of the revolution motor is larger than 3 nm and the rotation motor is smaller than 2 nm (De-Donatis et al.,2013). The revolution of dsDNA along the hexametric ATPase ring is evidenced by the following parameters: the width of dsDNA helix is 2 nm, but the diameter of the narrowest region of the all connector channels of phi29 (Guo et al.,1987c), SPP1, T4, T7, HK97, and FtsK is 3-5 nm, as revealed by X-ray crystallography (Fig. 15) (Agirrezabala et al.,2005; Lebedev et al.,2007; Massey et al.,2006; Sun et al.,2008; Valpuesta and Carrascosa,1994). Here the 5nm refers to the narrowest region of the channel and the other regions of the channel are much larger than 3-5nm in diameter. Such a size ratio makes it impossible to exercise the bolt-and-nut-tracing rotation mechanism, since for any instantaneous contact, the twisting of the DNA helix is required for rotation, however, with such a large channel, only one, but not two or more, ATPase subunit can touch the DNA (De-Donatis et al.,2013).

During revolution of dsDNA through the channel, dsDNA advances by touching the channel wall instead of proceeding through the center of the channel (Guo et al.,2013a). This is consistent with recent finding by Cryo-EM-imaging, showing that the T7 dsDNA core tilts from its central axis. A clear pattern indicates that the DNA core stack does tilt. A counterclockwise motion of the dsDNA was observed from the N-terminus when the connector was placed at the top (Guo et al.,2013a) (Fig. 8B). This is in agreement with direction of clockwise revolution of dsDNA viewed from the wider end (C-terminus) to the narrower end (N-terminus) of the phi29 connector (Jing et al.,2010; Schwartz et al.,2013b; Zhang et al.,2012; Zhao et al.,2013) (Fig. 8).

8. Revolution motors can be distinguished from rotation motors by detecting the chirality (Fig. 16)

Fig. 16.

The use of channel chirality to distinguish revolution motors from rotation motors. A. In revolution motors, the right-handed DNA revolves within a left-handed channel (Guasch et al.,2002; Olia et al.,2011; Zhao et al.,2013). B. In rotation motors, the right-handed DNA rotates through a right-handed channel via the parallel thread, with RecA (Xing and Bell,2004a) and DnaB (Itsathitphaisarn et al.,2012) shown as examples (adapted from (De-Donatis et al.,2013)). (PDB IDs: RecA, 1XMS. (Xing and Bell,2004b); P22-gp1, 3LJ5. (Olia et al.,2011))

As a common sense, a bolt-and-nut rotation mechanism requires that both treads of the bolt-and-nut display at the same orientation. However, the crystal structure of phi29 connector reveals an anti-parallel arrangement between motor connector subunit and the DNA helices. All 12 subunits of the connector portal protein of phi29, tilting at a 30° left-handed angle relative to the vertical axis of the channel. The left-handed channel facilitates advancement and revolution of the right-handed B-DNA in a single direction (Schwartz et al.,2013b; Zhao et al.,2013). This structural arrangement greatly facilitates the controlled motion, supporting the conclusion that dsDNA revolves, instead of rotates, through the connector channel touching each of the 12 connector subunits in 12 discrete steps of 30° transitions for each helical pitch, without producing a coiling or torsion force (Schwartz et al.,2013b). Moreover, the 30° angle of each connector subunit coincides nicely with the crystal structure of the spiral cellular clamp loader and the grooves of dsDNA (Mayanagi et al.,2009; McNally et al.,2010). The conserved regions serve a common function in DNA packaging. They assemble into a propeller-like structure composed of 12 subunits with a central channel that acts as a pathway for the translocation of DNA. The 30° anti-parallel arrangement that plays a critical role in dsDNA packaging is conserved in evolution of the viral dsDNA motor proteins known so far.

The similarity in secondary structure pattern and quaternary structure of the portal proteins among phi29 and the other dsDNA tailed bacteriophages SPP1, T7, T4, and HK97 family phages (Agirrezabala et al.,2005; Lebedev et al.,2007; Massey et al.,2006; Sun et al.,2008; Valpuesta and Carrascosa,1994), particularly the 30° anti-parallel arrangement between the connector subunit and the DNA helices, provides strong evidence that the role the portal protein plays in phi29 DNA packaging is shared by all other dsDNA tailed bacteriophages (De-Donatis et al.,2013). However, for rotation motors such as DnaB helicase (Itsathitphaisarn et al.,2012), the channel displays right-handed twisting in contrast to the left-handed chirality found in the revolutionary viral DNA packaging motors (Schwartz et al.,2013b; Zhao et al.,2013).

9. Special aspects of motor actions

9.1. The viral DNA translocases transport dsDNA in one direction

The direction of translocation is controlled by five features in phi29 packing motor: 1) ATPase undergoes a series of entropy transitions and conformational changes during ATP and dsDNA binding. Hydrolysis of ATP results in a second change in conformation, possibly with an entropy alteration of the ATPase, one with a low affinity for dsDNA, causing it to push dsDNA to advance and revolve within the channel. 2) The 30° angle of each subunit of the left-handed motor connector channel runs anti-parallel to the dsDNA helix, coinciding with the 12 subunits of the connector channel (360° ÷ 12 = 30°), as revealed by crystallography. 3) The unidirectional flowing property of the internal channel loops serves as a ratchet valve to prevent dsDNA reversal. 4) The 5′-3′ single-directional revolution of dsDNA advances with one strand contacting the phi29 motor connector channel wall. 5) Four lysine layers interact with a single strand of the dsDNA phosphate backbone (De-Donatis et al.,2013; Fang et al.,2012; Jing et al.,2010; Zhao et al.,2013).

The one-way flow loops within the channel facilitate DNA moving towards the capsid but not reversing (Fig. 12). The dsDNA runs through the connector, potentially making sequential contact at each subunit. A mutant connector was constructed with the deletion of the internal loop containing residues 229-246. The mutant procapsids with the loop-deleted connector failed to produce any virion. The packaged dsDNA reversed its direction and slided out of the mutant procapsid after being packaged (Fang et al.,2012; Geng et al.,2011; Grimes et al.,2011; Isidro et al.,2004; Serwer,2010). These results suggest that the channel loops act as a ratchet during DNA translocation to prevent the DNA from leaving, in line with the “push-through one-way valve” model (Fang et al.,2012; Jing et al.,2010; Zhang et al.,2012; Zhao et al.,2013). The connector is a one-way valve that only allows dsDNA to move into the procapsid unidirectionally, and the two-way traffic property has never been detected for wild type connectors under experimental conditions (Jing et al.,2010). However, when all of the internal loops were deleted, translocation of dsDNA through the connector was blocked, as demonstrated by single pore conductance assay (Fang et al.,2012), and the one-way traffic channel became two-way traffic for ssRNA or ssDNA (Geng et al.,2013). The channel loops might act as a clamp; when internal loops of the connector were deleted, two-way traffic of DNA was observed, as demonstrated by both scanning potential and polarity switch (Geng et al.,2013; Zhao et al.,2013). Analysis of the crystal structure and Cryo-EM density map of SPP1 channel loops reveals that these loops are located in a close proximity to dsDNA via non-ionic interactions. The channel loops of phi29 and SPP1 might play similar roles in directional traffic of dsDNA to enter the capsid through the connector channel (Orlova et al.,2003).

The 5′-3′ single-directional movement along one strand advances DNA towards the capsid without reversal. The packaging direction is from 5′ to 3′, and dsDNA is processed by contacting the connector with one strand of DNA in the 5′ to 3′ direction, as modification of phi29 genome DNA in the 5′ to 3′ direction strand was found to stop dsDNA packaging (Aathavan et al.,2009) (Fig. 17). One complete circle of revolution is equal to one helical turn of 10.5 base pairs of dsDNA. The dsDNA revolves along the motor using a single strand in the 5′ to 3′ direction (Oram et al.,2008). It has been reported that a 3′ single-strand overhang could be packaged under conditions extending from the 100 bp duplex. Extensions up to 12 bases at the end of the DNA did not inhibit the initiation of translocation. However, extensions to 20 or more bases significantly blocked the DNA packaging of the T4 motor (Fig. 17). The 20-base gap has been consistently found to be vulnerable, whether it was at the 3′ end or in the middle of the DNA strand (Oram et al.,2008).

Fig. 17.

Effect of DNA chemistry and structure on its packaging. A. Demonstration of blockage of dsDNA packaging by single-stranded gaps. When a single-stranded gap is present, only the left-end fragment of phi29 genomic DNA is packaged (adapted from (Moll and Guo,2005) with permission from Elsevier). B. T4 DNA packaging assays reveals that single stranded extensions with less than 12 bases at the DNA end do not inhibit translocation, whereas the ones with more bases do have a significant effect on the packaging (adapted from (Oram et al.,2008) with permission from Elsevier). C. Chemical modification of the negatively charged phosphate backbone on DNA packaging. Modification on the 3′→5′ strand does not block dsDNA packaging, but alternation on the other direction seriously affects DNA packaging. The results support the finding of the revolution mechanism showing that only one strand of the dsDNA interacts with the motor channel during revolution (adapted from (Aathavan et al.,2009) with permission from Nature Publishing Group).

9.2. Electrostatic interaction of lysine layers and phosphate backbone and mismatch between 10.5 bases per DNA patch and the 12 subunits connector channel wall result in uneven steps in DNA translocation

The electrostatic force arises from the relay interaction of the electropositive lysine layers with the electronegative DNA phosphate backbone. The negatively charged phi29 connector interior channel surface is decorated with 48 positively charged lysine residues, as revealed by the connector crystal analysis (Guasch et al.,2002) (Fig. 5). The residues that exist as four 12-lysine rings are derived from the 12 protein subunits that enclose the channel. DNA revolves through 12 subunits of the connector per cycle, with only one strand touching the channel wall. During a complete 360° revolution, the negatively charged phosphate backbone makes contact with the same positively charged layer of the lysine ring (Schwartz et al.,2013b; Schwartz et al.,2013a; Zhao et al.,2013). This results in uneven speed with four pauses during the DNA translocation process (Chistol et al.,2012; Schwartz et al.,2013b; Zhao et al.,2013).

The crystal structure shows that the length of the connector channel is 7 nm (Badasso et al.,2000; Guasch et al.,2002; Jimenez et al.,1986) and four lysine layers fall vertically within a 3.7 nm range (Guasch et al.,2002), spaced approximately ~0.9 nm apart. Since B-type dsDNA has a pitch of 0.34 nm/bp, ~2.6 bp exist for each rise along its axis that interact with the lysine ring (0.9 nm/0.34 nm·bp⁻¹ = ~2.6 bp). For every 10.5 bp, one strand of DNA revolves 360° through the channel. This results in a 0.875 mismatch between the base with the negatively charged phosphate group and the channel subunits with positively charged lysine ring (10.5/12 = 0.875). The dsDNA phosphate backbone interacts with the positively charged lysine in the next subunit, and the distance variation due to this mismatch is compensated for by the introduction of the next lysine layer. Thus, the contact point between the phosphate and the lysine shifts to a spot on the next lysine ring where the phosphate can make its next contact. This transition leads to a slight pause in DNA advancement. When dsDNA translocates through three subunits, the leading phosphate transitions into the next lysine layer in order to compensate for the imperfect match that occurs between the phosphate and each lysine residue during DNA advancement through the connector (Fig. 5). The parameters calculated in this section is an average, and are only used to help interpret the data and to rectify the model reported in two recent papers published in Nature and Cell (Moffitt et al.,2009; Chistol et al.,2012). While in these two papers, solid and informative data were produced by single molecule studies via optical traps to reveal four discrete steps in packaging 10 base pairs of phi29 DNA, the incorrect interpretation has led to a four-step rotation model that is contradictory and can not be agreeable with motor mechanisms reported by all other groups. The calculation in this section by no means represents a precise mechanical action of the lysine layer as long as the distance is concern, since the distances of four lysine layers in phi29 and in other phage display a signification variation. Indeed, it has been found that the mutation of only one layer of the four lysine rings does not significantly affect motor action (Fang et al.,2012), indicating that the interaction of lysine and phosphate is only an auxiliary force but not the main force involved in motor action. This also indicates that the uneven speed of the four steps of transition and pausing caused by the four lysine layers is not an essential function of the motor. The lysine layers and the 10.5 bases per helical turn are not a perfect match and the distance between layers is not constant. The lysine layers are also reported in the inner walls of the portal proteins of SPP1 and P22, which may potentially interact with their genomic DNA during packaging. Such pauses in speed during DNA translocation have been reported in both phi29 (Chistol et al.,2012; Moffitt et al.,2009) and T4 (Kottadiel et al.,2012).

9.3. The motor can transport concatemeric or closed circular dsDNA in prokaryotic or eukaryotic cells without breaking any covalent-bonds or changing topology of dsDNA

During cell segregation or binary fission, dsDNA translocases can transport circular dsDNA without breaking any covalent-bonds or changing topology of dsDNA (Bath et al.,2000; Demarre et al.,2013; Massey et al.,2006). In most dsDNA viruses, the dsDNA substrate for packaging is the concatemer dsDNA (Bravo and Alonso,1990; Everett,1981; Jacob et al.,1979; Son et al.,1988; White and Richardson,1987). How do the motors transport concatemers or closed circular dsDNA?

The ATPase monomer (not the hexamer) can bind to dsDNA (Schwartz et al.,2012; Schwartz et al.,2013b). After ATP binding, the ATPase subunit decreases in entropy and changes its conformation, leading to a high affinity for dsDNA thus assembling as a hexamer on DNA without requirement of ATP hydrolysis (De-Donatis et al.,2013). Therefore, the motor assembles into a hexamer on the DNA itself, rather than the insertion or treading of the dsDNA molecule into a preformed hexameric ring. A free 5′ or 3′ dsDNA end is not required, and the motor can theoretically transport closed circular dsDNA without breaking covalent bonds or changing DNA topology.

It was commonly believed that the binding of ATPase gp16 to the viral procapsid to form a complex is the first step in motor function during DNA packaging (Fujisawa et al.,1991; Guo et al.,1987b). However, it has been found that the first step in phi29 DNA packaging is instead the binding of multiple gp16 in a queue along the dsDNA (Fig. 18A) (Schwartz et al.,2012; Schwartz et al.,2013b). A string of multiple Cy3-gp16 complexes have been observed on dsDNA chains that has been lifted by, and stretched between, two polylysine-coated silica beads in the presence of nonhydrolysable ATPγS (Fig. 18A). This suggests that the queuing of ATPase gp16 along the DNA is the initial step in phi29 DNA packaging. Upon a conformational change stimulated by ATP binding, single subunit gathers around the dsDNA chain first and then form a complex surrounding the DNA. It has been noted that both ends of the dsDNA are tethered to the beads, an observation which proves that a free 5′ or 3′ dsDNA end is not required for the ATPase to bind dsDNA, and that assembly of the hexameric gp16 ring occurs only upon binding to DNA (Fig. 18). EM images also reveal multiple gp16 complexes binding to phi29 genome (Schwartz et al.,2013b).

Fig. 18.

Elucidation of how the motor transports closed circular dsDNA without breaking any covalent-bonds or changing the topology of the DNA. A. Direct observation of ATPase complexes queuing and moving along dsDNA. Cy3 conjugated gp16 were incubated with (a, b, e) and without (d) dsDNA tethered between two polylysine beads. The zoomed in images are shown in (c) and (f). (g, h) The motion of the Cy3-gp16 spot was analyzed and a kymograph was produced to characterize the ATPase walking. (adapted from (Schwartz et al.,2013b) with permission from Elsevier). B. A model showing that gp16 hexamer acts as an open spiral filament similar to RecA filament with a different chirality rather than as a closed ring. (adapted from (De-Donatis et al.,2013)). C. Illustration of the assembly of the motor subunit on circular dsDNA. Subunits gather around the dsDNA chain first and then form a hexamer complex surrounding the DNA.

9.4. Sequential action of the six subunits of the ATPase

Sequential action of phi29 DNA packaging motor was originally reported by Chen and Guo (Chen and Guo,1997b) and subsequently confirmed by Bustamante and coworkers (Moffitt et al.,2009). Hill constant determination and binomial distribution of inhibition assay have led to the conclusion that ATPase subunits work sequentially and cooperatively (De-Donatis et al.,2013; Schwartz et al.,2013b). This action enables the motor to work continuously without interruption, despite some observable pauses. Cooperativity among motor subunits allows the motor to continue to move, as deduced from the observation of the binding of dsDNA to only one gp16 subunit at a time. ATPase activity has been analyzed by studying the effects of the introduced mutant subunits on the oligomerization of gp16 (Chen et al.,1997; Trottier and Guo,1997). Increasing the amounts of Walker B mutants added to the overall gp16 oligomer failed to provide any significant effect on the rate of hydrolysis when ATPase activity was measured in the absence of dsDNA. However, when saturating amounts of dsDNA were added to the reaction, a strong negative cooperative effect was produced with a profile that mostly overlapped with a predicted profile in which one single inactive subunit was able to inactivate the whole oligomer. A predicted case was calculated from the binomial distribution inhibition assay (Chen et al.,1997; Trottier and Guo,1997); these results suggest that in the presence of dsDNA, a rearrangement occurs within the subunits of gp16 that enables them to communicate with each other and to “sense” the nucleotide state of the reciprocal subunit.

Binding of ATP to the conformationally disordered ATPase subunit stimulates an conformational change, with possible entropy alteration (De-Donatis et al.,2013), of the ATPase, thus fastening the ATPase at a less random configuration (Fig. 9). This lower conformation entropy enables the ATPase subunit to bind dsDNA and prime ATP hydrolysis. Based on the fact that the size of the motor channel is much larger than the width of the dsDNA, in any instantaneous time, only one ATPase subunit can touch the DNA, and only the subunit that is bound to the dsDNA at any given time is permitted to hydrolyze ATP. ATP hydrolysis triggers the second entropic and conformational change, which renders the ATPase into a low affinity for dsDNA thus pushing the DNA to the next subunit that has already bound to ATP with a high affinity for dsDNA. These sequential actions will promote the movement and revolution of the dsDNA around the internal hexameric ATPase ring.

Translocation occurs while the other subunits are in a type of “stalled” or “loaded” state. This represents an extremely high level of coordination in the function of the proteins with their DNA substrate, perhaps the most efficient process of coupling energy production via ATP hydrolysis and DNA translocation of all viral motors known so far. An effective mechanism of coordination is apparent between gp16 and dsDNA through the hydrolysis cycle as means for regulation. The cooperativity and sequential actions among hexameric ATPase subunits (Chen and Guo,1997b; Moffitt et al.,2009) promote revolution of dsDNA along the channel. The contact between the connector and the dsDNA chain is transferred from one point on the phosphate backbone to another, in line with previous reports (Moffitt et al.,2009). Formation of both gp16/procapsid complex and gp16 queue on DNA have been observed (De-Donatis et al.,2013; Guo et al.,1987b; Koti et al.,2008; Schwartz et al.,2013b). The finding has led to the proposal (De-Donatis et al.,2013) that gp16 hexamer is not an closed ring but an open spiral ring forming a filament similar to the RecA protein with different chirality (Fig. 18B) (Di et al.,1982).

10. Potential applications of the revolution motor

The discoveries discussed in this review offer a series of possible answers to the puzzles that may lead to inventions in the motion world. The riding system along one string of dsDNA provides a platform for cargo transportation at the nanoscale, and a tool for studying force generation mechanisms in a moving world. The revolution mechanism itself offers a prototype or a hint for the design of new motors involving forward motion via a tract, such as that used by roller coasters, trolley cars or launcher for flying objects such as rocket to depart from a helical tract without rotation of the object. The transition of dsDNA along 12 channel subunits offers a series of recognition sites on the dsDNA backbone and provides additional spatial variables to discriminate nucleotides based on distance parameters for nucleotide sensing. Nature has evolved a clever machine that translocates DNA double helix to avoid difficulties associated with rotation, e.g. DNA supercoiling. A video of dsDNA translocation can be found at nanobio.uky.edu/file/motion.avi.