Structural biology

Steps in the right direction

The ring-shaped helicase enzyme Rho moves along RNA using ATP as an energy source. Coordinating ATP hydrolysis with nucleic-acid binding seems to determine the direction and mechanism of helicase movement.

Helicases are molecular motors, fuelled by energy from ATP hydrolysis, that move along nucleic-acid molecules opening up double strands of DNA, RNA or DNA–RNA hybrids. These enzymes function in fundamental cellular processes such as genome replication, recombination and transcription1. The six subunits of the hexameric helicases are arranged in a ring that encircles a nucleic-acid strand. The ATP-binding sites, which catalyse ATP hydrolysis, are located at the interfaces between each subunit.

Deciphering how the six subunits coordinate their ATPase cycles to move the ring in a specific direction has been a major research topic, one that is advanced by two studies investigating the hexameric helicase Rho. In Cell, Thomsen and Berger2 report a high-resolution structure of Rho bound to RNA and ATP mimics. The structure suggests that ATP is hydrolysed in an ordered, sequential manner, and that the helicase tracks along RNA taking steps of one nucleotide per molecule of ATP hydrolysed. Meanwhile, Schwartz et al.3 investigate Rho movement using a chemical method, and report in Nature Structural and Molecular Biology that Rho makes essential interactions with the RNA every seven nucleotides, which suggests a much larger step size.

The Rho helicase is involved in terminating transcription in bacteria. When a copy of RNA is made from DNA, a short segment of the RNA at the leading edge of synthesis is paired with the DNA template to form a stable RNA–DNA hybrid within the active site of the transcribing RNA polymerase. To release the newly formed RNA, the Rho helicase binds to the RNA, moves along it in the 5′-to-3′ direction, and dissociates the RNA transcript from the RNA polymerase by disrupting the RNA–DNA complex.

A previous structure4 captured Rho as an open ring, with RNA wrapped around the primary RNA-binding sites involved in initiation of transcription termination. In this structure, the RNA was absent from the central channel that couples ATP hydrolysis5 to the conformational changes required for movement along nucleic acids. Thomsen and Berger's structure2 captures Rho in a closed-ring conformation with six nucleotides of RNA bound in the central channel of the ring, and with six ATP mimics (ADP˙BeF3) bound to the ATPase sites (Fig. 1).

Figure 1: Structure of the Rho hexameric helicase.

Thomsen and Berger's structure2 of Rho bound to RNA and ATP mimics (ADP˙BeF3) shows that Rho's six subunits interact with the RNA (purple) in the central channel through the Q and R loops (shown here as one loop per subunit). The ATP mimics are bound at the subunit interfaces. The different colours of the spheres depict the various conformational states of their ATP-binding site. Hydrolysis-competent T* subunits bind ATP and RNA tightly. The T subunits (before ATP hydrolysis), E subunit (involved in nucleotide exchange) and D subunit (bound to the product of ATP hydrolysis) all bind ATP and the central RNA weakly. The T and E subunits each interact with one nucleotide of RNA at a time, whereas the D subunit has limited interactions with RNA.

The present structure2 provides a snapshot of the ring moving along RNA, and shows that each of the six helicase subunits interacts differently with the RNA, which makes a complete turn in the central channel. Two RNA-binding elements, the Q and R loops, extend from each subunit of the Rho hexamer into the central channel, with several amino-acid residues from each of the loops associating with the backbone ribose and phosphates of the RNA. Interaction of the Q loop with the 2′-hydroxyl of the ribose of RNA explains the strict specificity of Rho for RNA. These RNA-binding elements organize into a spiral staircase to track the RNA backbone in a manner that is remarkably similar to that observed in the viral DNA-processing hexameric helicase E1 (ref. 6). As E1 and Rho share little sequence similarity, these findings indicate that hexameric helicases have converged on a common mode of nucleic-acid binding.

The fully liganded Rho structure2 has six ATP mimics bound at the subunit interfaces (Fig. 1). Each of the sites binds ATP either tightly (T* state in Fig. 1) or weakly (T, E or D states). The subunits that bind ATP tightly also bind RNA tightly, whereas the subunits that bind ATP weakly bind RNA weakly. Rho helicase moves in the 5′-to-3′ direction along RNA, whereas E1 moves in the 3′-to-5′ direction along its DNA substrate. Knowledge of the structures of two hexameric helicases that move in opposite directions on nucleic acid represents a unique opportunity to examine the structural basis of directionality.

One might assume that helicases moving on nucleic acid in opposite directions would bind nucleic acid in opposite orientations. Surprisingly, this is not the case. Rho and E1 bind nucleic acid with the same relative polarity, and the chirality of their amino-acid spiral staircase is also the same. But the order of ATPase sites with weakly and tightly bound ATPs around the ring in E1 is the opposite in Rho. This suggests that the cycles of ATP hydrolysis that drive movement of the helicase along the nucleic acid might occur in opposite directions around the ring in the two enzymes. Moreover, in contrast to Rho, the subunits of E1 that make tight interactions with DNA are those that have hydrolysed ATP (rather than those that bind it tightly). This raises an intriguing alternative possibility to explain the opposing directions of translocation — the on and off switches for nucleic-acid binding and release might be reversed in the two helicases.

Although the sequence of ATP binding and hydrolysis cannot be established from a single structural snapshot, the arrangement of Thomsen and Berger's Rho structure2 supports the ordered sequential model of ATP hydrolysis and translocation (the one-nucleotide-per-ATP model)6. Conceptually, this model of helicase function is similar to the binding change or rotational catalysis mechanism of the ATP synthase7, which has sequence and structural similarity to the Rho protein. The one-nucleotide-per-ATP model2,6 (Fig. 2a) proposes that each Rho subunit hydrolyses ATP sequentially around the ring. After ATP hydrolysis at the subunit that is tightly bound to RNA, one nucleotide of RNA is released at the 5′ end from the subunit at the top of the amino-acid staircase. ATP binding and product release in subunits at the bottom of the staircase trigger binding, in an orchestrated manner, of one nucleotide from the 3′ end of RNA. In this way, a single cycle of ATP hydrolysis leads to movement with a step of one nucleotide in the 5′-to-3′ direction.

Figure 2: Two models of hexameric helicase nucleic-acid translocation.

As indicated by the crystal structure2, five nucleotides of RNA (spheres) interact with five subunits of the helicase (hands A–E). The red closed hand is the subunit that has the tightest grip on the RNA, and is activated for ATP hydrolysis. The subunits represented by orange hands are in various tight and weak binding states. The green, open hand is the D subunit, which has limited interactions with RNA. a, The one-nucleotide-per-ATP stepping model proposes that, with every ATP-hydrolysis cycle, the subunit with the tightest grip on RNA (red hand) releases one nucleotide of RNA from the 5′ end, and a subunit grabs one nucleotide of RNA from the 3′ end (green hand). Thus, Rho moves along RNA in the 5′-to-3′ direction with steps of one nucleotide per ATP hydrolysed. b, In the zipper-lock model, which involves movement along more than one nucleotide per ATP hydrolysed, the tightly bound subunit functions as the zipper-lock. With every ATP-hydrolysis cycle, the zipper is unlocked, which causes a global change in the RNA-binding staircase (orange hands become blue), releasing a large segment of RNA from the 5′ end. A new zipper-lock is formed at an adjacent subunit, which restores the RNA-binding staircase, allowing it to spontaneously bind a new segment of RNA from the 3′ end. Thus, by fluctuating between tight and weak states, the hexamer ring translocates more than one nucleotide per ATP hydrolysed. Pi, inorganic phosphate.

The one-nucleotide stepping model predicts that each nucleotide of RNA goes through uniform interactions with the helicase subunits during translocation. The biochemical studies described by Schwartz et al.3 challenge the predictions of this model. The authors used a chemical-interference method to probe functionally essential RNA interactions that occur as Rho tracks along and unwinds the RNA–DNA duplex. Surprisingly, they found that Rho interacts with the 2′-hydroxyl group of the RNA only about every seven nucleotides. This seems inconsistent with the one-nucleotide-per-ATP stepping mechanism2,6. To explain the large step size, Schwartz and colleagues3 instead propose a mechanism similar to that used by the ring-shaped, ATP-powered motor of the bacterial virus phage Φ29, which packages DNA into its empty phage capsid8. The authors3 suggest that Rho fluctuates between open and closed ring structures and that the two subunits at the gap in the open structure act as a 'latch' and 'lever' to bind and release RNA periodically with large 7-nucleotide movements.

To reconcile these structural2 and biochemical6 findings, a modified version of the latch–lever model, dubbed the zipper-lock model (Fig. 2b), is proposed here. As with the one-nucleotide stepping model, the zipper-lock model assumes that each subunit transitions through a round of ATP binding, hydrolysis and product release and that this occurs in a sequential manner around the ring.

In contrast to the one-nucleotide stepping model, which assumes that ATP hydrolysis causes a local conformational change that releases one nucleotide of RNA, the zipper-lock model posits that ATP hydrolysis causes a global conformational change in the ring that releases more than one nucleotide of RNA. The tightly binding subunit acts like a zipper-lock (latch) to hold the RNA in place. The zipper is unlocked after ATP hydrolysis at the locking subunit, which disrupts the RNA-binding staircase, unzipping the remaining protein–RNA interactions, while an adjacent subunit forms a new zipper-lock with a distant RNA nucleotide (Fig. 2b). The new zipper-lock reorganizes the amino-acid staircase to spontaneously bind a new, equally large segment of RNA from the 3′ end. The 7-nucleotide periodicity in the chemical-interference pattern3 can be explained by clockwise ATP hydrolysis and periodic zipper-lock formation with roughly every seventh nucleotide of RNA. Variations of this model can explain helicase steps ranging from one nucleotide to seven nucleotides. Greater than one-nucleotide movement per ATP molecule hydrolysed is also consistent with biochemical studies9 of the hexameric T7 gp4 helicase.

The structures of the fully liganded hexameric helicases2,6 represent a turning point for understanding the workings of ring-shaped helicases. And the two studies of Rho2,3 offer challenges for mechanistic deliberations and investigations to fully understand how ring-shaped helicases move on nucleic acids. Future work will involve obtaining additional snapshots of the hexameric helicase in various ATP-liganded states to sort out the directionality of ATP hydrolysis around the ring; applying new methods to elucidate the dynamics of the nucleic-acid translocation process; and measuring the chemical and physical step sizes of the hexameric helicases by biochemical and single-molecule kinetics.


  1. 1

    Patel, S. S. & Picha, K. M. Annu. Rev. Biochem. 69, 651–697 (2000).

  2. 2

    Thomsen, N. D. & Berger, J. M. Cell 139, 523–534 (2009).

  3. 3

    Schwartz, A. et al. Nature Struct. Mol. Biol. doi:10.1038/nsmb.1711 (2009).

  4. 4

    Skordalakes, E. & Berger, J. M. Cell 114, 135–146 (2003).

  5. 5

    Wei, R. R. & Richardson, J. P. J. Biol. Chem. 276, 28380–28387 (2001).

  6. 6

    Enemark, E. J. & Joshua-Tor, L. Nature 442, 270–275 (2006).

  7. 7

    Boyer, P. D. Annu. Rev. Biochem. 66, 717–749 (1997).

  8. 8

    Moffitt, J. R. et al. Nature 457, 446–450 (2009).

  9. 9

    Donmez, I. & Patel, S. S. EMBO J. 27, 1718–1726 (2008).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Patel, S. Steps in the right direction. Nature 462, 581–583 (2009) doi:10.1038/462581a

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.