DEAD-Box Proteins Unwind Duplexes by Local Strand Separation Quansheng Yang, Mark Del Campo, Alan M. Lambowitz, Eckhard Jankowsky  Molecular Cell  Volume 28, Issue 2, Pages 253-263 (October 2007) DOI: 10.1016/j.molcel.2007.08.016 Copyright © 2007 Elsevier Inc. Terms and Conditions

Figure 1 The Length of Single-Stranded Substrate Regions Affects Functional Binding of Ded1p, but Not the Strand-Separation Step (A) RNA substrate duplexes with progressively longer single-stranded regions (5–35 nt) and identical 16 bp duplexes (for sequences, see Experimental Procedures). (B) Dependence of the unwinding rate constants for substrates shown in (A) on Ded1p concentration. Data points are averages from multiple independent measurements. Error bars represent the standard deviation. Curves of unwinding rate constants as a function of Ded1p concentration for substrates with 35, 25, and 15 nt overhangs were fit to a sigmoidal binding isotherm (Hill equation), according to kunwind(max) = kunwind · (Ded1p)n · (K1/2(unwind)n + [Ded1p]n)−1. K1/2(unwind) is the apparent functional binding constant, and n is the Hill coefficient; values are shown in Table 1. Unwinding rate constants at enzyme saturation were kunwind(max)[35 nt] = 3.4 ± 0.1 min−1, kunwind(max)[25 nt] = 3.3 ± 0.2 min−1, and kunwind(max)[15 nt] = 3.4 ± 0.1 min−1. Lines through the data points for the substrates with a 5 nt overhang and the blunt-end duplex represent a trend. Molecular Cell 2007 28, 253-263DOI: (10.1016/j.molcel.2007.08.016) Copyright © 2007 Elsevier Inc. Terms and Conditions

Figure 2 Ded1p Does Not Require a Specific Helix Terminus for Duplex Unwinding (A) Unwinding of DNA-RNA chimeric substrates with single-stranded regions by Ded1p. Black lines indicate RNA segments; gray lines indicate DNA. 3′ and 5′ denote the respective ends of the oligonucleotides. Sequence was identical for all substrates (Experimental Procedures). Unwinding rate constants are indicated as sliding point and were measured at saturating concentrations of Ded1p (600 nM, Figure 1). Saturation was verified for each substrate. Data points are averages of multiple measurements. (B) Unwinding of blunt-end substrates with 600 nM Ded1p. (C) Unwinding of DNA-RNA chimeric substrates with additional alterations. In substrate vii, the polarity of the single-stranded overhang is reversed. Arrows indicate 3′-5′ polarity; the overhang is covalently attached to the duplex region via a 5′-5′ linkage. Substrate viii is a multiple-piece substrate in which the RNA terminus of the chimeric strand is blocked by an adjacent DNA strand. Ends are indicated. Reactions were performed as described above, and saturation of each substrate with Ded1p was verified. (D) Representative PAGE of the unwinding reaction with the multipiece substrate viii. Cartoons at the left indicate the mobility of individual species; asterisks indicate the radiolabels. The zero time point represents the reaction prior to ATP addition. Samples were removed from 30 s to 30 min. 95°C represents the control for denatured strands. Molecular Cell 2007 28, 253-263DOI: (10.1016/j.molcel.2007.08.016) Copyright © 2007 Elsevier Inc. Terms and Conditions

Figure 3 Unwinding of Substrates without RNA Termini by Ded1p and Mss116p (A) Unwinding of substrates with a 10 nt, internal RNA segment by Ded1p. Substrate ix contains a 25 nt overhang 5′ to the duplex region; substrate x is blunt ended. Sequences were as in Figures 1 and 2. Cartoons illustrate substrate design (black, RNA; gray, DNA). The numbers indicate the length of the respective DNA and RNA segments; the asterisk depicts the radiolabel. PAGE panels show representative unwinding reactions with 1000 nM Ded1p. The mobility of duplex and single-stranded species is indicated by the cartoons on the left. The zero time point represents the reaction prior to ATP addition. Samples were removed from 5 s to 30 min. Unwinding rate constants (kunwind) were calculated as in Figure 1. Plots show the dependence of unwinding rate constants on the Ded1p concentration for substrates ix and x. Data points are averages of multiple measurements; error bars indicate the standard deviation. The curve for the substrate with the overhang was fit to a sigmoidal binding isotherm (Figure 1B), yielding kunwind(max) = 8.8 ± 0.6 min−1, K1/2(unwind) = 608 ± 34 nM, n = 4.0 ± 0.7 (For comparison between sigmoidal and hyperbolic curve fits, see Figure S4.) The lines through the data points for the blunt-end substrate x represent a trend. (B) Unwinding of substrates with a 5 nt, internal RNA segment by Ded1p. Sequences and reactions were as in (A. Lines in the plots showing the dependence of unwinding rate constants on the Ded1p concentration represent a linear trend. (C) Unwinding of substrates with a 10 nt, internal RNA segment by Mss116p. Sequences, reactions, and data analysis were as in (A, except for the reaction temperature, which was 24°C (Experimental Procedures). Representative unwinding reactions shown in the PAGE panels were performed with 400 nM Mss116p. For substrate ix with single-stranded overhang, kunwind(max) = 3.3 ± 0.2 min−1, K1/2(unwind) = 36 ± 4 nM, n = 3.5 ± 1.1. For the blunt-end substrate kunwind(max) = 4.2 ± 0.1 min−1, K1/2(unwind) = 192 ± 5 nM, n = 2.0 ± 0.1. (D) Unwinding of substrates with a 5 nt, internal RNA segment by Mss116p. Sequences, reactions, and data analysis were as described above in (C). For substrate x with a single-stranded overhang, kunwind(max) = 0.44 ± 0.01 min−1, K1/2(unwind) = 165 ± 3, nM, n = 3.8 ± 0.2; for the substrate without a single-stranded overhang, kunwind(max) = 0.43 ± 0.1 min−1, K1/2(unwind) = 146 ± 5 nM, n = 3.6 ± 0.1. Molecular Cell 2007 28, 253-263DOI: (10.1016/j.molcel.2007.08.016) Copyright © 2007 Elsevier Inc. Terms and Conditions

Figure 4 Dependence of Unwinding Rate Constants of Ded1p and Mss116p on the Length of the RNA Segment in the Duplex Unwinding rate constants for Ded1p with the blunt-end substrates were determined as in Figure 3A, with 1000 nM Ded1p (left panel). Sequences for all substrates were identical; the numbers indicate the length of the RNA (black) and the DNA (gray) segments. Unwinding rate constants for Mss116p were determined as in Figure 3C, with 400 nM enzyme (right panel). Note the logarithmic scale of the y axis. Molecular Cell 2007 28, 253-263DOI: (10.1016/j.molcel.2007.08.016) Copyright © 2007 Elsevier Inc. Terms and Conditions

Figure 5 Duplex Unwinding from Terminal and Internal Duplex Regions by Ded1p and Mss116p (A) Unwinding of substrates with 5 nt RNA segments at different duplex positions by Ded1p. Sequences for all substrates were identical. Numbers show the lengths of the RNA (black) and the DNA (gray) segments. Termini are indicated. Unwinding rate constants were determined as in Figure 3A, with 1000 nM Ded1p. (B) Unwinding of substrates with 5 nt RNA segments at various duplex positions by Mss116p. Unwinding rate constants were determined as in Figure 3C, with 400 nM Mss116p, which was saturating for all substrates. Molecular Cell 2007 28, 253-263DOI: (10.1016/j.molcel.2007.08.016) Copyright © 2007 Elsevier Inc. Terms and Conditions

Figure 6 Three-Step Mechanism for Duplex Unwinding by DEAD-Box Proteins Step 1: the DEAD-box protein binds to the duplex either at a terminal or internal region. This loading is facilitated by single-stranded regions if the RNA segment is not shorter than a critical length between 10 and 5 nucleotides and the single-stranded region also exceeds a critical length between 5 and 15 nt. Step 2: the duplex is opened locally in an ATP-dependent reaction. This opening can occur anywhere in the duplex, provided RNA residues are accessible. The number of base pairs opened is hypothetical. Step 3: dissociation of remaining base pairs. For longer and/or more stable duplexes, this step will greatly affect the overall efficiency of the strand separation. The overall rate of duplex unwinding will thus depend on loading efficiency (step 1, accounting for effects of single-stranded regions and effects of enzyme concentrations), the rate by which the duplex is locally opened (step 2, accounting for effects of ATP concentration), and the lifetime of this locally opened state, as well as on the stability of remaining base pairs (step 3, accounting for the effect of duplex length and/or stability). Molecular Cell 2007 28, 253-263DOI: (10.1016/j.molcel.2007.08.016) Copyright © 2007 Elsevier Inc. Terms and Conditions