DEAD-box proteins as RNA helicases and chaperones

Abstract

DEAD-box proteins are ubiquitous in RNA-mediated processes and function by coupling cycles of ATP binding and hydrolysis to changes in affinity for single-stranded RNA. Many DEAD-box proteins use this basic mechanism as the foundation for a version of RNA helicase activity, efficiently separating the strands of short RNA duplexes in a process that involves little or no translocation. This activity, coupled with mechanisms to direct different DEAD-box proteins to their physiological substrates, allows them to promote RNA folding steps and rearrangements and to accelerate remodeling of RNA–protein complexes. This review will describe the properties of DEAD-box proteins as RNA helicases and the current understanding of how the energy from ATPase activity is used to drive the separation of RNA duplex strands. It will then describe how the basic biochemical properties allow some DEAD-box proteins to function as chaperones by promoting RNA folding reactions, with a focus on the self-splicing group I and group II intron RNAs.

Keywords: ATPase; CYT-19; DExD/H-box protein; Group I intron; Group II intron; Mss116; RNA folding.

Figure 1

Roles of single-stranded and double-stranded extensions in duplex unwinding. A, Strand separation by a processive, 3′-5′ helicase begins when the helicase loads on a single-stranded extension with a dangling 3′-end. The helicase then translocates in the 3′ to 5′ direction into the adjacent double-stranded region. B, Single-stranded overhangs enhance strand separation by some DEAD-box proteins, but are not required to have a defined polarity with respect to the adjacent duplex. C, Double-stranded extensions can also enhance unwinding for some DEAD-box proteins, and these extensions may be RNA or DNA. D, A single-stranded RNA or DNA segment can activate RNA unwinding by DEAD-box proteins even when physically separated from the neighboring duplex, as demonstrated when biotinylated RNA constructs were linked to streptavidin (22).

Figure 2

Targeting DEAD-box proteins to RNA substrates through specific and non-specific interactions. A, Recognition of a specific RNA motif via an ancillary domain. Hairpin 92 from the 23S rRNA, shown by nucleotide letters, is recognized with high specificity by the bacterial proteins DbpA and YxiN. B, Electrostatically-driven RNA binding through an ancillary domain or unstructured extension. C, Helicase core loading on RNA substrates via dimerization, with a tethering interaction being formed by the helicase core from one monomer and the unwinding being carried out by the other monomer (22). D, Targeting through specific interactions between the DEAD-box core and a protein component of the targeted RNP complex.

Figure 3

DEAD-box protein structure and interactions with ssRNA and ATP. A, Schematic depiction of domains and motifs, as present in Mss116p. Cylinders indicate domains that are present in the crystal structure. NTE, N-terminal extension (does not include the mitochondrial targeting sequence); CTE, non-conserved C-terminal extension; BT, basic tail. B–D, Structural views of RNA and ATP binding by DEAD-box proteins, shown for Mss116p (59). B, Domain structure. Bound U₁₀ RNA and AMP-PNP are magenta and red, respectively. Note the location of the ligand-binding pockets at the interface of the two RecA-like domains. C, Interactions and conserved motifs. Conserved sequence motifs are highlighted in the same colors as in panel A. D, Stereoview of the RNA binding site. The structure has been rotated by approximately 140° relative to panels B and C. The six RNA nucleotides that are bound in a consistent manner in all available DEAD-box protein:RNA structures are indicated by labels on the first and the last nucleotides (U3 and U8). The conserved helix that necessitates RNA bending is labeled (α8) and the approximate center of the bend is indicated with an arrow. Amino acids that contact the RNA are shown as stick models.

Figure 4

Coupling of ATPase cycles to duplex unwinding activity. Pathways for strand separation with or without ATP hydrolysis are indicated by arrows and the numbers 1–3 (see text for further details). Pathway 4 depicts a futile cycle of ATP hydrolysis without complete strand separation. An additional tethering interaction is shown in the model throughout the strand separation process. Modified from ref. (13) with permission.

Figure 5

Representative structures of group I and group II intron RNAs. A, Secondary and tertiary structures of the Tetrahymena thermophila group I intron. The tertiary structure is a model that is supported by domain structures and biochemical data (113). B, Secondary and tertiary structure of the group II intron from Oceanobacillus iheyensis (114, 115).

Figure 6

Sensitivity of CYT-19 to RNA tertiary structure. Tertiary contacts between the 6-bp P1 duplex (green) and the core of Tetrahymena ribozyme (black cylinders) inhibit unwinding of the duplex by CYT-19. However, reversible “undocking” of P1 allows CYT-19 to promote strand separation of the duplex (21).