Article pubs.acs.org/biochemistry
Coordinated Action of Two Double-Stranded RNA Binding Motifs and an RGG Motif Enables Nuclear Factor 90 To Flexibly Target Different RNA Substrates Tobias Schmidt,† Paul Knick,† Hauke Lilie,‡ Susann Friedrich,† Ralph Peter Golbik,† and Sven-Erik Behrens*,† †
Institute of Biochemistry and Biotechnology (NFI), Section of Microbial Biotechnology, and ‡Section of Protein Biochemistry, Martin Luther University Halle-Wittenberg, Kurt-Mothes-Strasse 3, D-06120 Halle/Saale, Germany S Supporting Information *
ABSTRACT: The mechanisms of how RNA binding proteins (RBP) bind to and distinguish different RNA molecules are yet uncertain. Here, we performed a comprehensive analysis of the RNA binding properties of multidomain RBP nuclear factor 90 (NF90) by investigating specifically the functional activities of two doublestranded RNA binding motifs (dsRBM) and an RGG motif in the protein’s unstructured C-terminus. By comparison of the RNA binding affinities of several NF90 variants and their modes of binding to a set of defined RNA molecules, the activities of the motifs turned out to be very different. While dsRBM1 contributes little to RNA binding, dsRBM2 is essential for effective binding of double-stranded RNA. The protein’s immediate C-terminus, including the RGG motif, is indispensable for interactions of the protein with single-stranded RNA, and the RGG motif decisively contributes to NF90’s overall RNA binding properties. Conformational studies, which compared wild-type NF90 with a variant that contains a pseudophosphorylated residue in the RGG motif, suggest that the NF90 C-terminus is involved in conformational changes in the protein after RNA binding, with the RGG motif acting as a central regulatory element. In summary, our data propose a concerted action of all RNA binding motifs within the frame of the full-length protein, which may be controlled by regulation of the activity of the RGG motif, e.g., by phosphorylation. This multidomain interplay enables the RBP NF90 to discriminate RNA features by dynamic and adaptable interactions. serve as protein−protein interaction sites.10,11 The solution structures of isolated tandem dsRBMs of the RBPs ADAR2 (adenosine deaminase acting on RNA) and PKR (protein kinase RNA-activated) suggested that the two motifs may act differently but jointly during RNA binding.12,13 Little about how various RNA binding motifs such as dsRBMs and RGG motifs cooperate in the context of an entire protein is yet known.3 Nuclear factor 90 (NF90), also known as NFAR-1, DRBP76, or TCP80, is a multifunctional protein14 that participates in several stages of cellular gene expression, including transcription,15 nuclear export of RNA,16 translation,17,18 and miRNA biogenesis.19 Best documented is the role of NF90/ NFAR-1 acting as an RBP that controls gene expression on the posttranscriptional level. Thus, NF90 was shown to bind to the untranslated regions (UTRs) of various mRNAs and to modulate their turnover.20−23 In most cultivated cells, NF90 is located predominantly in the nucleus.24 However, many of NF90’s activities implicate a cytoplasmic localization involving shuttling, which is triggered by phosphorylation.25,26
RNA binding proteins (RBPs) are important regulators of mammalian gene expression that may affect the maturation, subcellular distribution, and translation and decay of mRNAs.1,2 These activities, which are often linked with the nuclear/ cytoplasmic shuttling of the proteins, imply that RBPs interact with numerous RNA molecules. The knowledge of the molecular features that determine how an RBP binds to and specifically discriminates between different types of RNA molecules is just beginning to emerge.3 RBPs commonly contain several RNA binding motifs, such as dsRBM and/or RGG motifs, the composition and positioning of which are assumed to decisively impact the protein’s RNA binding properties.3,4 Double-stranded RNA binding motifs (dsRBMs), which consist of ∼70 amino acids (aa), are evolutionarily highly conserved domains that interact with high affinity with α-helical dsRNA. Among several other conserved residues, dsRBMs contain a central phenylalanine residue, which is a crucial determinant of the fold and the RNA binding properties of the motif4−7 (Figure S1). RGG motifs, also termed RGG/RG box or RGG box, commonly consist of several copies of arginine and glycine repeats.8 While RGG motifs are known to bind nonspecifically to nearly all types of nucleic acids,8,9 their modes of interactions with RNAs are uncertain. With RBPs that contain several dsRBM copies, these were shown to display different RNA binding activities6 and to © 2016 American Chemical Society
Received: October 1, 2015 Revised: January 19, 2016 Published: January 22, 2016 948
DOI: 10.1021/acs.biochem.5b01072 Biochemistry 2016, 55, 948−959
Article
Biochemistry
Figure 1. Features of NF90, the NF90 isoforms, and the applied NF90 variants. Quality of the purified proteins. (A) Organization of NF90 isoforms NF90a and NF90b. The double-stranded RNA binding motifs (dsRBM), the RGG motifs, the dimerization zinc finger motifs, and isoform-specific elements are indicated. Arrows mark the amino acid positions that were mutated in the variants used in this study. The bottom panel shows a prediction of NF90b’s secondary structure elements using the PredictProtein-Server40 (filled boxes, α-helix; empty boxes, β-sheet; see the text). (B) Coomassie-stained SDS gel of purified and refolded NF90bwt (WT), NF90bΔC63 (ΔC63), and NF90bΔC46 (ΔC46). (C) UV absorption spectrum (see the text) and (D) sedimentation equilibrium analysis (analytical ultracentrifugation) of purified and refolded NF90bwt in assay buffer. The molecular mass was determined with an apparent Svedberg constant of 3.95, corresponding to 60 kDa. This revealed the monomeric state of the protein (theoretical mass of 76 kDa). (E) Far-UV circular dichroism (CD) spectrum of purified NF90bwt. The acquired data were normalized to mean residue weight (MRW) ellipticities.
constellation, NF90b’s dsRBMs function very differently, while the RGG motif was indicated to have an important regulatory role. The experimental data suggest that the dynamic, joint activity of the three RNA binding motifs enables NF90b to recognize structural features of target RNAs.
Several isoforms of NF90 are expressed in human cells by alternative splicing of pre-mRNAs of the ILF3 gene locus (Figure 1A).27 All forms share a bipartite nuclear localization signal (NLS) and an N-terminal dimerization zinc finger domain (DZF) that appears to function as an interaction interface of other DZF-containing proteins.28 All isoforms also contain two dsRBMs and an RGG motif that are suspected to be essentially involved in RNA binding. Isoforms NF90a and NF90b differ only by an insert of four amino acids (NVKQ) that is located between the dsRBMs. Here, we took NF90b as an example of a multidomain RNA binding protein to investigate the mechanisms that define its binding specificities and that enable the protein to discriminate between different RNA molecules. Moreover, we examined the roles of the two dsRBMs and the RGG motif of NF90 during the protein’s RNA binding process. For this, we applied a combination of biophysical methods to the wild-type (wt) NF90b and to NF90b variants with mutations in the binding motifs and the protein’s C-terminus. Comparative studies of the binding of the purified protein variants to a set of model RNAs revealed a cooperative and coordinated action of the three motifs in the course of the RNA binding process. Within this
■
MATERIALS AND METHODS Plasmid Constructs. The NF90b gene (Homo sapiens) was cloned into a pET-21(a)+ vector that was used for expression in Escherichia coli. The described NF90b gene variants were generated by primer-mediated mutagenesis and polymerase chain reaction (Table S1). The truncated gene variants of NF90bΔC63 and NF90bΔC46 were generated by introduction of an additional amber stop codon at the desired positions. Expression and Purification of NF90bwt and NF90b Variants. The genes of NF90b or its respective variants were expressed in E. coli BL21-CodonPlus (DE3)-RP. Biomass production was conducted using fermentation. Gene expression was induced by adding 1 mM isopropyl 1-thio-D-galactopyranoside. NF90b was found in inclusion bodies (IB) when expressed in E. coli. Accordingly, the wild-type (NF90wt) protein as well as all protein variants were purified from the IB fraction, whose 949
DOI: 10.1021/acs.biochem.5b01072 Biochemistry 2016, 55, 948−959
Article
Biochemistry Table 1. Synthetic 5′-FAM-EX-Labeled RNA Molecules That Were Applied for Binding Studies
a
Extinction coefficients were provided by IBA.
and of the RNA conformation, the data were converted to mean residue ellipticity ΘMRW. Intrinsic Fluorescence. Fluorescence spectra were recorded on a Fluoromax-4 spectrofluorometer (Jobin Yvon) using 3 μM protein in assay buffer supplemented with sodium chloride at the indicated concentrations. The excitation monochromator was set to 280 or 295 nm, and the excitation and emission slit widths were set to 1 and 5 nm, respectively; the scanning speed was 100 nm/min with a response time of 1 s. At least five spectra were measured, averaged, and normalized to the fluorescence intensity at 343 nm. Substrates Used in the Nucleic Acid Binding Studies. Fluorescently labeled oligonucleotides were purchased from IBA GmbH (Göttingen, Germany). The concentrations were determined by absorbance at 260 nm using the respective extinction coefficients. We prepared double-stranded nucleic acids by mixing the corresponding single strands in an equimolar ratio, heating them to 95 °C, cooling them, and finally purifying them by nondenaturing polyacrylamide gel electrophoresis before storing them at −80 °C. The RNA molecules that were used in this study are summarized in Table 1. Measurement of RNA Binding Constants and Binding Mode Analysis (LEM). The protein of interest was added to 5′-FAM-EX-labeled RNA (25−300 nM) in assay buffer supplemented with the indicated concentrations of NaCl. Fluorescence changes were monitored on a Fluoromax-4 spectrofluorometer (Jobin Yvon) at 20 °C. After equilibrium had been attained, the signal amplitudes of the 5′-FAM-EXlabeled RNAs were measured (excitation at 491 nm, emission at 515 nm) and corrected for the volume change. Fluorescence intensities relative to the starting fluorescence were plotted against the protein concentration. Fitting the binding isotherms according to eq 134−36 with KaleidaGraph (Synergy Software) yielded the KD values of the interaction of the protein and the labeled RNA (see the Supporting Information for the mathematical derivation).
isolation followed a modified protocol described by Rudolph et al.29 Briefly, cells were harvested, centrifuged, resuspended in 0.1 M Tris-HCl (pH 7.0) and 1 mM EDTA, and lysed using a French press homogenizer. Following DNaseI treatment and incubation with Triton X-100, the inclusion bodies were collected by centrifugation, washed, and resolubilized using 8 M urea as a denaturant. After clarification by ultracentrifugation, the denatured protein solution was fractionated by cation exchange chromatography (GE Healthcare), and bound protein was eluted by applying an ionic strength gradient. For refolding, pooled fractions of denatured protein were continuously dropped and rapidly diluted in refolding buffer [50 mM sodium phosphate, 100 mM NaCl, 5% (v/v) glycerol, 500 mM arginine, and 10 mM DTT (pH 7.2)] with a final dilution factor of at least 10-fold and a final protein concentration of
