Recognition of transcription termination signal by the nuclear polyadenylated RNA-binding (NAB) 3 protein

Abstract

Non-coding RNA polymerase II transcripts are processed by the poly(A)-independent termination pathway that requires the Nrd1 complex. The Nrd1 complex includes two RNA-binding proteins, the nuclear polyadenylated RNA-binding (Nab) 3 and the nuclear pre-mRNA down-regulation (Nrd) 1 that bind their specific termination elements. Here we report the solution structure of the RNA-recognition motif (RRM) of Nab3 in complex with a UCUU oligonucleotide, representing the Nab3 termination element. The structure shows that the first three nucleotides of UCUU are accommodated on the β-sheet surface of Nab3 RRM, but reveals a sequence-specific recognition only for the central cytidine and uridine. The specific contacts we identified are important for binding affinity in vitro as well as for yeast viability. Furthermore, we show that both RNA-binding motifs of Nab3 and Nrd1 alone bind their termination elements with a weak affinity. Interestingly, when Nab3 and Nrd1 form a heterodimer, the affinity to RNA is significantly increased due to the cooperative binding. These findings are in accordance with the model of their function in the poly(A) independent termination, in which binding to the combined and/or repetitive termination elements elicits efficient termination.

FIGURE 1.

Overview of the RRM of Nab3 sequence, topology, NMR spectra, solution structure, and domain structure in Nab3 and Nrd1. A, amino acid sequence of the S. cerevisiae Nab3 RRM along with its secondary structure elements and general consensus of RNP1 and RNP2 motifs. B, a schematic drawing of the domain structure of Nab3 and Nrd1. C, two-dimensional ¹H-¹⁵N HSQC spectrum of 2.5 mm uniformly ¹⁵N,¹³C-labeled Nab3 RRM in 50 mm sodium phosphate buffer (pH 8.0), 300 mm NaCl, and 10 mm β-mercaptoethanol (90% H₂O, 10% D₂O). The spectrum was acquired at 303 K on a Bruker Avance 600 MHz spectrometer. The assignments are labeled by the one-letter code of amino acids accompanied by a sequence number. The side chain resonances of asparagine and glutamine are connected by horizontal lines. D, stereo view of the 20 lowest energy structures of Nab3 RRM. The protein backbone is shown as a wire model. E, stereo view of the representative (the lowest energy) structure of Nab3 RRM shown as a ribbon diagram. The figure was generated with MOLMOL (38).

FIGURE 2.

NMR titration experiments of Nab3 RRM with UCUU RNA. A, ¹H-¹⁵N HSQC spectra of Nab3 RRM alone (in red) and in the presence of 1 eq of 5′-UCUU-3′ (in blue) at 303 K. B, close-up views of the ¹H-¹⁵N HSQC spectra, showing selected chemical shift changes during the titration. C, quantification of chemical shift perturbations of Nab3 RRM upon binding to UCUU RNA. The combined chemical shift perturbations ([ω_HNΔδ_HN]² + [ω_NΔδ_N]²)^1/2, where ω_HN = 1 and ω_N = 0.154 are weight factors of the nucleus (52), are plotted versus the amino acid residue number. Large changes occur on the β-sheet surface. The assignments of residues indicated by asterisks could not be obtained for neither the free nor bound protein, or indicates proline residues.

FIGURE 3.

Overview of the solution structure of the Nab3 RRM in complex with UCUU. A, stereo view of the 20 lowest energy structures of the Nab3 RRM-UCUU complex. The protein backbone is shown as a wire model in black. The RNA heavy atoms are shown as a wire model in red. B, stereo view of the representative (the lowest energy) structure of the Nab3 RRM-UCUU complex. The RNA is represented as a white stick model and the protein is shown as a ribbon model with residues that contact the RNA shown in yellow. Putative hydrogen bonds are shown by dotted magenta lines. C, scheme showing contacts between Nab3 RRM and the UCUU RNA. Protein residues that form putative hydrogen bonds to the RNA are shown in blue and the one having hydrophobic interactions are in yellow. A hypothetic recognition of U4 is labeled by a gray question mark. D, solvent-accessible surface representation of Nab3 RRM colored by electrostatic potential (blue, positive; red, negative) and stick representation for the RNA of the representative structure of the complex. Figures were generated with MOLMOL (38).

FIGURE 4.

Equilibrium binding of Nab3 RRM, Nrd1 RRM, and Nrd1-Nab3 heterodimer with fluorescently labeled RNA monitored by fluorescence anisotropy. A, Nab3 RRM was titrated with UCUU and GUAA (each 10 nm), and their binding isotherms are shown as red circles and triangles, respectively. B, Nab3 RRM was titrated with snR47, snR13, and SL substrates (each 10 nm), and their binding isotherms are shown as red circles, squares, and triangles, respectively. C, Nrd1 RRM was titrated with GUAA and snR13 (each 10 nm) and their binding isotherms are shown as blue inverted triangles and circles, respectively. D, Nrd1-Nab3 heterodimer was titrated with snR13 (100 pm). E, summary of the association constants (K_a) for the RRM of Nrd1 (in blue) and Nab3 (in red) in their free forms as well as for the Nrd1-Nab3 heterodimer (in black). Logarithmic scale of K_a is shown to cover a wide range of affinities. F, RNA sequences used in the affinity measurements. The buffers contained the same ion strength and pH values for all proteins. Equilibrium dissociation constant (K_d) was calculated from the best fit to the data using a single-site binding isotherm. Error is denoted as S.E. The data were normalized for visualization purposes (A–C).

FIGURE 5.

The important residues of Nab3 RRM that are required for RNA binding and cell viability. A, equilibrium binding of Nab3 RRM mutants with fluorescently labeled RNA monitored by fluorescence anisotropy. The Nab3 RRM R331A, N361A, S399A, and E397A mutants along with the wild-type of Nab3 RRM were titrated with fluorescently labeled snR47 substrate. Equilibrium association constants (K_a) are shown for individual mutants with S.E. B, residues Arg³³¹, Ser³⁹⁹, and Asn³⁶¹ are required for yeast viability. The indicated Nab3 RRM mutants were expressed episomally from pRS415 plasmids in the yeast strain with the endogenous NAB3 driven by the GAL1 promoter. Mutant strains were spotted on plates containing 2% glucose and a control galactose plate and incubated for 3 days at the indicated temperatures. Growth on glucose-containing plates leads to the repression of GAL1-driven wild-type Nab3, and thus shows the functionality of the different Nab3 mutants. Vector is a control where the GAL1::NAB3 strain contains an empty pRS415 plasmid, wt is the wild-type NAB3. C, expression of Nab3 proteins from pRS415 in glucose-containing medium. Western blot analysis was perfomed with protein extracts from the original GAL1::NAB3 strain (DLY889) grown in galactose-containing medium and extracts from DLY889 transformed with plasmids carrying wild-type and mutant NAB3 grown for 20 h in glucose-containing medium. Air2 was used as a loading control.

FIGURE 6.

Recognition of YCU by Nab3 and PTB RRMs. A, sequence alignment of PTB RRM1, -2, -3, and -4 and Nab3 RRM whose structures have been solved. The alignment was performed using ClustalW (53) and manually optimized using the three-dimensional structural information (49). For the RRMs of PTB, amino acids interacting with the RNA are shown in red boxes, residues in gray and black boxes are located in the β-sheet and residues in yellow and cyan boxes are in the α- or 3₁₀ helices, respectively (49). Residues in gray boxes form the hydrophobic core of the domains. For the RRM of Nab3, residues in red boxes are significantly perturbed upon RNA binding. B, comparison of Nab3 RRM (left: in yellow and red schematics) and PTB RRM1 (right: in cyan schematics) binding to UCU nucleotides (represented as a stick model). The protein residues that mediate the specific recognition are highlighted as a stick model. The C-terminal region of PTB RRM1 that mediates the recognition of U₃ is shown in magenta.