Probing the mechanisms of DEAD-box proteins as general RNA chaperones: the C-terminal domain of CYT-19 mediates general recognition of RNA

Abstract

The DEAD-box protein CYT-19 functions in the folding of several group I introns in vivo and a diverse set of group I and group II RNAs in vitro. Recent work using the Tetrahymena group I ribozyme demonstrated that CYT-19 possesses a second RNA-binding site, distinct from the unwinding active site, which enhances unwinding activity by binding nonspecifically to the adjacent RNA structure. Here, we probe the region of CYT-19 responsible for that binding by constructing a C-terminal truncation variant that lacks 49 amino acids and terminates at a domain boundary, as defined by limited proteolysis. This truncated protein unwinds a six-base-pair duplex, formed between the oligonucleotide substrate of the Tetrahymena ribozyme and an oligonucleotide corresponding to the internal guide sequence of the ribozyme, with near-wild-type efficiency. However, the truncated protein is activated much less than the wild-type protein when the duplex is covalently linked to the ribozyme or single-stranded or double-stranded extensions. Thus, the active site for RNA unwinding remains functional in the truncated CYT-19, but the site that binds the adjacent RNA structure has been compromised. Equilibrium binding experiments confirmed that the truncated protein binds RNA less tightly than the wild-type protein. RNA binding by the compromised site is important for chaperone activity, because the truncated protein is less active in facilitating the folding of a group I intron that requires CYT-19 in vivo. The deleted region contains arginine-rich sequences, as found in other RNA-binding proteins, and may function by tethering CYT-19 to structured RNAs, so that it can efficiently disrupt exposed, non-native structural elements, allowing them to refold. Many other DExD/H-box proteins also contain arginine-rich ancillary domains, and some of these domains may function similarly as nonspecific RNA-binding elements that enhance general RNA chaperone activity.

Fig. 1

Papain digestion of CYT-19 and Δ578–626. A, Complete time course of digestion. Lane 1, undigested CYT-19; lanes 2–7, time course of CYT-19 digestion by papain. Lane 8, undigested Δ578–626; lanes 9–14, time course of Δ578–626 digestion by papain under the same conditions. Digestion times for each protein are: lanes 2 and 9, two min; lanes 3 and 10, six min; lanes 4 and 11, 15 min; lanes 5 and 12, 45 min; lanes 6 and 13, 60 min; lanes 7 and 14, 120 min. B, Early times of papain digestion. Lanes 1 and 8 show the undigested CYT-19 and Δ578–626 proteins, respectively. Other lanes show digestion times of: 30 s (lanes 2 and 9), 1 min 30 s (lanes 3 and 10), 2 min 30 s (lanes 4 and 11), 6 min (lanes 5 and 12), 8 min 30 s (lanes 6 and 13), and 20 min (lanes 7 and 14).

Fig. 2

Acceleration of substrate dissociation from the Tetrahymena ribozyme. A, Secondary structure of the ribozyme. The P1 duplex, formed between the substrate and the ribozyme, is highlighted in black, with the substrate cleavage site indicated by a dashed arrow. B, Progress curves of substrate dissociation in the absence of CYT-19 (○, k_obs = 0.033 min⁻¹), in the presence of 100 nM (□, k_obs = 0.32 min⁻¹) or 200 nM (Δ, k_obs = 0.49 min⁻¹) Δ578–626 and 2 mM ATP-Mg²⁺, 200 nM Δ578–626 without ATP (◇, k_obs = 0.028 min⁻¹) or 20 nM (σ, k_obs = 1.3 min⁻¹) or 100 nM (ν, k_obs = 4.7 min⁻¹) wild-type CYT-19 and 2 mM ATP-Mg²⁺. C, Dependence of substrate dissociation rate on the concentration of Δ578–626 (○) and wild-type CYT-19 (□). Results from equivalent reactions without CYT-19 are also shown ( ). Four independent determinations gave a k_cat/K_M value of 3.6 (± 1.8) × 10⁶ M⁻¹ min⁻¹ for the Δ578–626 protein, and three independent determinations gave a k_cat/K_M value of 4.1 (± 1.4) × 10⁷ M⁻¹ min⁻¹ for the wild-type protein.

Fig. 3

Unwinding of the isolated P1 duplex by Δ578–626 (○) and wild-type CYT-19 (□). Equivalent reactions in the absence of CYT-19 are also shown (Δ). Five independent determinations gave an average k_cat/K_M value of 4.1 (± 1.8) × 10⁵ M⁻¹ min⁻¹ for the Δ578–626 protein and four independent determinations gave a value of 6.9 (± 1.7) × 10⁵ M⁻¹ min⁻¹ for the wild-type protein (25 °C, pH 7.0, 5 mM Mg²⁺). The inset shows the sequence of the isolated P1 duplex.

Fig. 4

CYT-19-mediated unwinding of the construct containing P1 and P2. A, Secondary structure of the P1-P2 substrate. For consistency, the orientations of the P1 and P2 domains are the same as in Figs. 2 and 3. B, The dependence of the unwinding rate on the concentrations of Δ578–626 (○) and wild-type CYT-19 (□). Equivalent reactions in the absence of CYT-19 are also shown (Δ). Two independent determinations gave a k_cat/K_M value of 9.3 (± 1.0) × 10⁵ M⁻¹ min⁻¹ for the Δ578–626 protein and three independent determinations gave a value of 2.7 (± 0.6) × 10⁷ M⁻¹ min⁻¹ for the wild-type protein. All data from these determinations are shown.

Fig. 5

Equilibrium binding of Δ578–626 to the Tetrahymena ribozyme, as monitored by nitrocellulose filter binding. The data shown for the Δ578–626 protein (○) gave a K_d value of 200 nM, and the data for the wild-type CYT-19 protein, from an experiment performed side-by-side, (∇) gave a K_d value of 30 nM. The experiments shown did not include added nucleotides; analogous experiments in the presence of 2 mM ADP-Mg²⁺ or 2 mM AMP-PNP-Mg²⁺ gave indistinguishable results for both the wild-type and Δ578–626 proteins.

Fig. 6

Acceleration by CYT-19 of misfolded Tetrahymena ribozyme re-folding to the native state. The progress of re-folding was followed by the onset of substrate cleavage activity by the ribozyme in the absence of CYT-19 (λ) or in the presence of various concentrations of Δ578–626 (○) or wild-type CYT-19 (performed side-by side) (τ). An equivalent experiment for the wild-type CYT-19 protein has been published previously, and these data are included for comparison (∇). The dependences of re-folding rate constant on protein concentration gave k_cat/K_M values of 8.6 (± 0.4) × 10⁴ M⁻¹ min⁻¹ and 2.9 (± 0.4) × 10⁴ M⁻¹ for the wild-type and Δ578–626 proteins, respectively.

Fig. 7

Δ578–626 is defective in facilitating splicing of the Neurospora mt LSU intron. A, Disappearance of precursor RNA (20 nM) was followed at 25°C in the presence of 100 nM CYT-18 dimer in the absence of CYT-19 (○) or with 100 nM CYT-19 (∇), 100 nM Δ578–626 (◇), or 500 nM Δ578–626 (□). The RNA was also incubated without any proteins (×) as a negative control. RNA splicing was performed as described in Materials and Methods. B, Dependence of the splicing rate constant on the concentrations of Δ578–626 (○, λ) and wild-type CYT-19 (□, ν). Equivalent reactions in the presence of CYT-18 and absence of CYT-19 are also shown (Δ, σ). The data are from the experiment shown in panel A (filled symbols) and an analogous experiment using independent preparations of proteins (open symbols). Linear fits to the data shown gave k_cat/K_M values of 2.9 (± 0.8) × 10⁵ M⁻¹ min⁻¹ and 1.9 (± 0.2) × 10⁴ M⁻¹ for the wild-type and Δ578–626 proteins, respectively.

Fig. 8

Model for the role of the RNA binding site within the C-terminal domain (RBD) in enhancing duplex unwinding activity of CYT-19. A, CYT-19 binds with low specificity to structured RNA (blue cylinders, which depict the secondary structure elements of the Tetrahymena ribozyme) via its C-terminal domain and efficiently unwinds helices that are not packed tightly against the body of the structured RNA (the P1 duplex is shown as red and green strands). B, Enhanced unwinding activity by CYT-19 is also observed upon addition of a simple duplex to P1 (P2 is shown as a single blue cylinder), suggesting that the primary recognition by the C-terminal domain is for double-stranded RNA located adjacent to the helix to be unwound. See Discussion for further details.