![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 17, Issue 7, 3211-3220, July 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of *Pharmacology and Molecular Physiology and Biophysics, School of Medicine, Vanderbilt University, Nashville, TN 37232-8548
Submitted February 27, 2006;
Revised March 31, 2006;
Accepted April 21, 2006
Monitoring Editor: Weis Karsten
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Two mammalian enzymes in this family, ADAR1 and ADAR2, can catalyze the hydrolytic deamination of multiple sites in synthetic dsRNAs, or mediate the site-specific modification of naturally occurring viral and cellular mRNA transcripts containing extended duplex regions formed by the presence of imperfect inverted repeats (Rueter and Emeson, 1998; Bass, 2002). The most extensively studied substrates of A-to-I conversion are transcripts encoding ionotropic glutamate receptor (GluR) subunits and the 2C-subtype of the serotonin receptor (5-HT2CR), in which editing leads to nonsynonymous codon changes that generate channels with altered electrophysiologic and ion permeation properties (Seeburg and Hartner, 2003) and receptors with decreased G protein-coupling efficiency (Burns et al., 1997; Niswender et al., 1999; Berg et al., 2001). A-to-I modifications have also been described in nontranslated RNAs and noncoding regions of mRNA transcripts, suggesting that such RNA modifications affect other aspects of RNA function, including splicing, translation efficiency, nuclear retention, and transcript stability (Rueter et al., 1999; Athanasiadis et al., 2004; Blow et al., 2004; Kim et al., 2004; Levanon et al., 2004; DeCerbo and Carmichael, 2005; Prasanth et al., 2005).
ADAR2 displays a modular organization with two tandem dsRNA-binding motifs (dsRBMs) connected by a flexible linker and a conserved adenosine deaminase domain toward the carboxy terminus, for which the structures have recently been determined (Macbeth et al., 2005; Stefl et al., 2006). The dsRBM is a highly conserved 65–75 amino acid (aa) domain, present in many eukaryotic proteins with diverse cellular functions, that forms an 
1-
1-2-3-2 topology in which the two -helices are packed along a face of a three-stranded antiparallel -sheet, with most of the potential RNA-binding residues exposed on one surface (Fierro-Monti and Mathews, 2000; Tian et al., 2004). Structural analyses of dsRBMs complexed to short, synthetic RNA duplexes have suggested that dsRBM binding is independent of RNA sequence, because the majority of dsRBM–RNA interactions involve direct contact with the 2'-hydroxyl groups of the ribose sugars and direct or water-mediated contacts with nonbridging oxygen residues of the phosphodiester backbone (Ryter and Schultz, 1998; Ramos et al., 2000; Blaszczyk et al., 2004). The two dsRBMs of ADAR2, with >80% amino acid sequence similarity, adopt the same fold as all other members of the dsRBM family, although the two domains differ slightly from one another in the orientation of 1 helix relative to the other secondary structural elements (Stefl et al., 2006). Like other dsRBM-containing proteins, ADAR2 demonstrates a high-affinity for dsRNA and can edit up to 50% of the adenosine moieties in perfect dsRNA (Melcher et al., 1996; Liu et al., 1999; Lehmann and Bass, 2000; Cho et al., 2003; Dawson et al., 2004). However, ADAR2 can also demonstrate site-specific A-to-I conversion in mRNA transcripts (Emeson and Singh, 2001; Bass, 2002), providing a paradox by which such specificity can be achieved in the absence of sequence-specific dsRBM-RNA contacts. Recent studies have indicated that the dsRBMs of ADAR2 bind selectively to imperfect RNA duplexes in a manner distinct from that of an RNA-dependent protein kinase (PKR)-derived dsRBM, suggesting that individual dsRBMs possess intrinsic binding selectivity that influence substrate specificity for the parent protein (Stephens et al., 2004).
In addition to their role(s) in substrate recognition, the dsRBMs of ADAR1 and ADAR2 also have been shown to be critical for the nucleolar localization of these enzymes, representing an important mechanism by which RNA editing can be modulated by the sequestration of enzymatic activity from RNA substrates in the nucleoplasm (Desterro et al., 2003; Sansam et al., 2003). The localization of ADAR2 to the nucleolus is dependent not only on the ability of ADAR2 to bind to RNA duplexes via its dsRBMs but also on the presence of rRNA, suggesting that ADAR2 is targeted to the nucleolus by directly binding duplex regions in mature or pre-rRNA transcripts.
In this study, we demonstrate that deletion or the introduction of point mutations in each dsRBM of ADAR2 has a differential effect on both the site-selective modification of adenosine moieties in distinct ADAR2 substrates as well as in the localization of ADAR2 to the nucleolus. Although the dsRBMs of ADAR2 can functionally replace one another for subnuclear targeting, such alterations cannot support site-selective editing activity, emphasizing that the unique sequence/structure of each motif subserves specific roles in substrate recognition and subsequent ADAR2 function. More interestingly, observations that dsRBM point mutations have substrate-dependent effects upon site-specific editing in different RNA targets suggests that the precise amino acid residues involved in ADAR2–RNA interactions are unique for each RNA substrate, providing a molecular mechanism by which dsRBMs can specifically recognize multiple RNA structural determinants and enable ADAR2 to mediate the selective deamination of adenosine residues in a broad range of heterogenous RNA targets.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Mutation of lysine to alanine at positions 127 and 281 of the open reading frame was performed by PCR-mediated overlap extension (Warrens et al., 1997). Construction of eGFP-
dsRBM mutants was performed using the wild-type eGFP-ADAR2 plasmid as a template for run-around PCR (Coolidge and Patton, 1995) to delete amino acids 76–148 (dsRBM1), 230–301 (dsRBM2) and 76–301 (dsRBM1/2) from the ADAR2b open reading frame. Plasmids expressing enhanced yellow fluorescent protein (eYFP)-dsRBM fusion proteins were constructed by PCR amplification of a region of the ADAR2b open reading frame encoding dsRBM1 or dsRBM2, corresponding to amino acids 74–147 or 231–301, respectively, and subcloned into pEYFP-C1 (Clontech). Construction of plasmids encoding eGFP-ADAR2b isoforms with tandem copies of a single dsRBM was performed using PCR-mediated mutagenesis (Ausubel et al., 1998) to delete the region encoding dsRBM2 (aa 234–300) and replace it with amino acids 73–146 (dsRBM1/1) or by deleting the region encoding dsRBM1 (aa 76–148) and replacing it with amino acids 231–301 (dsRBM2/2). For the dsRBM1/1 plasmid, a single amino acid change was introduced (L301V relative to the wild-type ADAR2b sequence) from subcloning dsRBM1 into the position initially occupied by dsRBM2. Nucleolin cDNA was amplified by PCR with sense (5'-CGGAATTCCTATTCAAACTTCGTCTTCTTTCCTT-3') and antisense (5'-GAAGATCTGATCGCCACCATGGTGAAG-3') primers using pSport6-nucleolin (Open Biosystems, Huntsville, AL) as a template, and introduced EcoRI and BglII restriction sites (underlined) were used for subsequent subcloning into the pEYFP-C1 vector (Clontech). Construction of the GluR-2 (R/G site) minigene for transfection analysis was made by PCR amplification of mouse genomic DNA with sense (5'-CCGAAGCTTACGCACACTAAGGATCC-3') and antisense (5'-GGCCTCTAGATACAAACCGTTAAGAGTCTTA-3') primers to introduce HindIII and XbaI restriction sites (underlined) for subcloning into pRC/CMV2 (Invitrogen, Carlsbad, CA).
Cell Culture and Transfection
Human embryonic kidney (HEK) 293 cells and NIH/3T3 mouse fibroblasts (American Type Culture Collection, Manassas, VA) were maintained in alpha minimum essential medium and DMEM (Invitrogen), respectively, and supplemented with 10% (vol/vol) bovine calf serum (Hyclone Laboratories, Logan, UT). HEK293 cells were transiently transfected by calcium phosphate coprecipitation (Ausubel et al., 1998) with 8 µg of plasmid encoding wild-type or mutant eGFP-ADAR2 fusion proteins and 2 µg of either a rat ADAR2 (–1 site) or GluR-2 (R/G site) minigene reporter plasmid, as described previously (Rueter et al., 1999; Dawson et al., 2004). For analysis of subnuclear fluorescence, NIH/3T3 cells were plated on 35-mm glass-bottomed MatTek dishes (MatTek, Ashland, MA) and transiently transfected with 1 µg of plasmids encoding wild-type or mutant eGFP-ADAR2 fusion proteins and 1 µg of the eYFP-nucleolin plasmid using FuGENE6, according to the manufacturer’s instruction (Roche Diagnostics, Indianapolis, IN).
Quantitative Fluorescence Microscopy
Twenty-four hours after transfection, the subcellular localization of eGFP and eYFP fusion proteins was determined by multispectral confocal microscopy (LSM510 Meta; Carl Zeiss, Thornwood, NY); eGFP and eYFP signals were simultaneously excited at 488 nm, and the total fluorescence emission in the range of 510–628 nm was passed through a spectral grating and discriminated with 10.7-nm resolution using a multi-anode (multichannel) array detector. Multichannel signals for eGFP-only and eYFP-only samples were recorded and subsequently used as reference "signatures" to apply with a linear unmixing algorithm (Zimmermann et al., 2003) for clear discrimination of both fluorophores when mixed in the same sample. All images were acquired using a 40x/1.3 Plan Neofluar objective lens. Average nucleolar eGFP fluorescence intensity (Fn) was defined as the eGFP signal overlapping with eYFP-nucleolin and was compared with the average eGFP fluorescence intensity from a comparable area of the nucleoplasm (Fo) using the Image J image analysis software (http://rsb.info.nih.gov/ij/; National Institute of Mental Health, Bethesda, MD).
Western Blotting Analysis
Crude nuclear extracts were prepared from HEK293 cells, 60 h after transient transfection, as described previously (Schreiber et al., 1989) and diluted with dialysis buffer to maintain enzymatic activity (30 mM HEPES, pH 7.6, 300 mM NaCl, 10% glycerol, 1 mM EDTA, 0.5 mM EGTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 2 mg/ml leupeptin, and 0.1% aprotinin) (Rueter et al., 1999). Equivalent volumes for each protein sample were resolved by PAGE (7.5–12% SDS-PAGE) and transferred to a nitrocellulose membrane (Hybond-C Super; GE Healthcare, Little Chalfont, Buckinghamshire, United Kingdom). The membrane was probed with an affinity-purified ADAR2-specific antiserum raised against amino acids 6–66 of the rat ADAR2 open reading frame (Sansam et al., 2003), detected with an Alexa Fluor 680-labeled donkey anti-sheep IgG secondary antibody (0.4 ng/µl), and quantified using an Odyssey infrared imaging system (LI-COR Biotechnology, Lincoln, NE). For transient cotransfection analysis with ADAR2 and GluR-2 minigenes, relative ADAR2 protein expression was also normalized to the level of a constitutive nuclear protein, histone deacetylase 2 (HDAC2), using an affinity-purified goat polyclonal antibody (4 ng/µl; Santa Cruz Biotechnology, Santa Cruz, CA).
Quantitative Analysis of RNA Editing
Analysis of A-to-I conversion, using nuclear extracts from HEK293 cells transfected with wild-type and mutant eGFP-ADAR2 fusion proteins, was performed with in vitro-transcribed RNA substrates using the MEGAscript T7 transcription kit (Ambion, Austin, TX) or T3 RNA polymerase (Promega, Madison, WI), as described previously (Burns et al., 1997; Dawson et al., 2004); the concentration of transcribed RNA was determined by UV absorbance spectrometry at 260 nm. On hundred femtomoles of each RNA substrate was incubated with the nuclear extracts containing equivalent amounts of each eGFP-ADAR2 protein at 30°C, in a total volume of 50 µl, for varying times, and the reactions were terminated by freezing on dry ice. For preliminary studies (Supplemental Figure S1), HEK293 nuclear extracts containing the wild-type eGFP-ADAR2 fusion protein were variably diluted to determine the linear range of enzymatic activity for each RNA substrate. Subsequent analysis of editing activity for wild-type and mutant eGFP-ADAR2 fusion proteins was performed using the empirically determined conditions, as follows: dsRNA (D79N): extract dilution, 15; incubation time, 60 min; GluR-2 (Q/R site): extract dilution, 30; incubation time, 20 min; GluR-2 (R/G site): extract dilution, 270; incubation time, 30 min; 5-HT2CR (D-site): extract dilution, 30; incubation time, 8 min; and ADAR2 (–1 site): extract dilution, 30; incubation time, 5 min. Quantification of editing efficiency for [-32P]ATP-labeled D79N dsRNA was performed by thin layer chromatography (TLC) (Rueter et al., 1995), and analysis of editing for the remaining in vitro reaction products was performed using a modified primer-extension analysis, as described previously (Burns et al., 1997; Dawson et al., 2004; Feng et al., 2006).
Total RNA was isolated from HEK293 cells, 60 h after transient transfection, using TRI Reagent (Molecular Research Center, Cincinnati, OH). First-strand cDNA was synthesized using avian myeloblastosis virus reverse transcriptase (Promega) with minigene-specific primers for ADAR2 (5'-GGATCCCCCGGGCTGCAG-3') and GluR-2 (5'-GGCCGAATTCTACAAACCGTTAAGAGTCTTA-3') using 5 µg of total RNA. To quantify the relative expression of ADAR2 mRNA splice variants resulting from editing at the –1 site, the ADAR2 cDNA was amplified by PCR with a 6-carboxyfluorescein (6-FAM)–labeled sense primer and a nonlabeled antisense primer (Feng et al., 2006). Amplicons corresponding to alternatively spliced ADAR2 variants were resolved by 2.5% agarose gel electrophoresis and quantified by PhosphorImager analysis (GE Healthcare). Quantification of R/G site editing was performed using a modified primer-extension analysis, as described previously (Dawson et al., 2004).
Statistical Analysis
Student’s t test was performed using GraphPad Prism (GraphPad Software, San Diego, CA). Values are reported as mean ± SEM; p < 0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Sansam et al., 2003). To further examine the role of individual dsRBMs in maintaining the steady-state nucleolar localization of ADAR2 in living cells, we used multicolor fluorescence microscopy using a series of eGFP-ADAR2 mutants in transfected NIH/3T3 mouse fibroblasts (Figure 1), along with an eYFP-nucleolin fusion protein that was included to define the nucleolar compartment. The relative nucleolar localization of wild-type and mutant eGFP-ADAR2 proteins was quantified by measuring the average fluorescence intensity of eGFP-ADAR2 that overlapped with the eYFP-nucleolin signal in nucleoli (Fn) compared with the average fluorescence intensity of eGFP-ADAR2 from a corresponding area of the nucleoplasm (Fo). Because eGFP and eYFP have strongly overlapping emission spectra, we applied a computational spectral separation technique to resolve the signals from each fluorophore (Zimmermann et al., 2003). Transfection of eGFP alone resulted in a diffuse pattern of fluorescence in the cytoplasm and nucleus, with no preferential concentration in nucleoli, whereas the pattern of eYFP-nucleolin fluorescence was highly restricted to the nucleolar compartment (Figure 2A). The fluorescence pattern for cells cotransfected with both eGFP and eYFP-nucleolin was identical to that observed when each fluorophore was transfected independently, with little observable bleed-through between the emission channels for these simultaneously excited fluorescent proteins (Figure 2A).
|
|
dsRBM1/2 (76–301)] produced the expected pattern of diffuse nuclear fluorescence (Figure 2A), suggesting that dsRNA binding was required to maintain the steady-state localization of ADAR2 in nucleoli (Desterro et al., 2003; Sansam et al., 2003). Expression of a mutant eGFP-ADAR2 fusion protein (K127A, K281A), containing substitutions for highly conserved amino acids in the loop between the 3 and 2 regions for all dsRBMs (Tian et al., 2004), resulted in a diffuse pattern of nuclear fluorescence nearly identical to the pattern observed when the region containing both dsRBMs was deleted from the fusion protein (Figure 2A). Because analogous mutations in PKR and Staufen have been shown to ablate dsRNA-binding activity (McMillan et al., 1995; Ramos et al., 2000), these results further demonstrate that the localization of ADAR2 depends upon its ability to bind dsRNA. The expression of eGFP-ADAR2 mutants containing independent deletions of dsRBM1 [dsRBM1 (76–148)] or dsRBM2 [dsRBM2 (230–301)] also resulted in reduced nucleolar fluorescence for both fusion proteins (Figures 1 and 2, A and B), indicating that both dsRBMs are essential for the normal steady-state localization of ADAR2 in nucleoli. The relative nucleolar fluorescence for the dsRBM1 mutant was significantly less than for the fusion protein lacking dsRBM2 (p < 0.02), suggesting a greater role for dsRBM1 in the subnuclear compartmentalization of ADAR2. Independent substitutions of K127 and K281 decreased relative nucleolar fluorescence to the same extent (K127A and K281A); however, the magnitude of this effect was less than that observed for the dsRBM deletions (Figure 2), suggesting that additional contacts between the dsRBMs and nucleolar binding site(s) are required for normal nucleolar localization.
Because deletion or even subtle point mutations of ADAR2 dsRBMs could result in structural alterations that cause ADAR2 mislocalization, we also examined the subcellular localization of eYFP when expressed as a fusion protein with either dsRBM1 (aa 74–147) or dsRBM2 (aa 231–301) alone (Figure 2C); the precise borders for each motif were based upon amino acid sequence homology to other dsRBM-containing proteins and the recently resolved NMR structure for these domains in ADAR2 (Fierro-Monti and Mathews, 2000; Stefl et al., 2006). Both eYFP-dsRBM1 and eYFP-dsRBM2 were localized to the cytoplasm and nucleus, presumably because of the absence of a nuclear localization signal and the passive diffusion of in these small fusion proteins (
40 kDa) through the nuclear pore (Suntharalingam and Wente, 2003); however, eYFP-dsRBM1 was localized to nucleoli significantly better than eYFP-dsRBM2 (p < 0.001; Figure 2C), further confirming a nonequivalent role for these highly conserved domains in the maintenance of ADAR2 nucleolar localization.
Substrate-dependent dsRBM Contribution to Site-selective Editing Activity
Given the differential effects that mutation or deletion of dsRBMs have on the nucleolar localization of ADAR2, it is unclear to what extent these domains also control the site-selective editing of ADAR2 substrates. To further examine the roles of individual dsRBMs, we used a transiently transfected tissue culture model system in which a series of eGFP-ADAR2 mutants were assessed for their ability to catalyze site-selective A-to-I conversion in two well characterized ADAR2 targets, the –1 and R/G sites of ADAR2 and GluR-2 transcripts (Supplemental Figure S2), respectively (Lomeli et al., 1994; Rueter et al., 1999; Dawson et al., 2004). HEK293 cells were chosen for this analysis because previous studies demonstrated a low level of endogenous editing activity in this cell line (Maas et al., 1996; Burns et al., 1997; Rueter et al., 1999; Schaub and Keller, 2002).
Editing of the –1 site within ADAR2 pre-mRNAs generates a proximal 3'-splice site within intron 4 to direct the inclusion of an additional 47 nucleotides in the ADAR2 open reading frame (Rueter et al., 1999; Dawson et al., 2004; Feng et al., 2006). As an indirect index of –1 site editing, we quantified the relative abundance of minigene-derived ADAR2 splice variants, containing (+47) or lacking (–47) this alternatively spliced cassette, using a reverse transcription (RT)-PCR–based strategy with a 6-FAM–labeled sense PCR primer (Feng et al., 2006). Cotransfection of eGFP and the minigene resulted in sole expression of the –47 RNA isoform, indicative of the absence of –1 site editing, whereas coexpression of wild-type eGFP-ADAR2 generated the +47 isoform almost exclusively (Figure 3A). Deletion of dsRBM1 (dsRBM1) or substitution of the conserved lysine in this domain (K127A) had little effect on the extent of editing-dependent alternative splicing for the minigene, yet deletion of dsRBM2 (dsRBM2) or deletion of both dsRBMs (dsRBM1/2) completely ablated use of the proximal 3'-splice site. These results indicate that dsRBM1 is not required, whereas dsRBM2 is critical, for site-specific editing at the –1 site of ADAR2 pre-mRNA transcripts. Further substitution of a conserved lysine in dsRBM2 (K281A) or conserved lysines in both dsRBMs (K127A, K281A) reduced the extent of alternative splicing by 44–61%, indicating K281 contributes only partially to dsRBM2-dependent editing at the –1 site (Figure 3A). Similar analyses, using a minigene derived from the GluR-2 transcript, demonstrated that although deletion of dsRBM1 (dsRBM1) caused a 19% decrease in R/G site editing (p < 0.01), deletion of dsRBM2 resulted in a 57% reduction of ADAR2 activity (p < 0.001) (Figure 3B). These results were distinct from those observed for the –1 site, where dsRBM1 was completely dispensable and dsRBM2 was essential for A-to-I conversion (Figure 3A), yet such a disparity could not be explained by differences in the relative expression of wild-type or mutant fusion proteins in transfected cells (Supplemental Figure S3). Overall, these results suggest that substrate-dependent dsRBM recognition of RNA targets is a critical component of site-specific adenosine deamination.
|
; Raitskin et al., 2001; Desterro et al., 2003; Sansam et al., 2003) and that translocation of ADAR2 to the nucleoplasm results in increased editing activity (Sansam et al., 2003), making comparisons of editing activity for mutations that simultaneously affect both subnuclear accumulation and site-specific editing difficult to interpret.
To circumvent these problems, we used an in vitro editing system using crude nuclear extracts from HEK293 cells transiently transfected with different eGFP-ADAR2 mutants. The relative protein level for wild-type and mutant eGFP-ADAR2 proteins in HEK293 nuclear extracts was determined by quantitative Western blotting analysis using an affinity-purified antiserum directed against amino acids 6–66 of wild-type ADAR2 (Figure 4) (Sansam et al., 2003; Dawson et al., 2004), and all proteins were diluted to achieve the same final concentration in the in vitro editing reaction. A band corresponding to the expected molecular weight for each eGFP-ADAR2 protein was observed, along with minor band representing a stable degradation product (Figure 4), and no signal was seen for mock-transfected HEK293 cells (our unpublished data). In addition to the RNA substrates previously used for transfection studies (Figure 3), the in vitro analyses also took advantage of a variety of ADAR2 substrates that are distinct at the level of both nucleotide sequence and predicted RNA secondary structure (Emeson and Singh, 2001; Dawson et al., 2004), including RNA targets containing the D-site of 5-HT2CR transcripts (Burns et al., 1997), the Q/R site of GluR-2 RNAs (Sommer et al., 1991; Higuchi et al., 1993), and a perfect, synthetic dsRNA (D79N) derived from a portion of the 2A adrenergic receptor (Burns et al., 1997; Lakhlani et al., 1997) (Supplemental Figure S2). For each RNA editing substrate (100 fmol), the amount of the eGFP-ADAR2 protein and the duration of the reaction were empirically determined to assure that all of the editing reactions were performed in the linear range of the assay (Supplemental Figure S1). The extent of total inosine production for the perfect dsRNA was determined by TLC (Rueter et al., 1995), and site-specific editing of the remaining ADAR2 substrates was quantified using modified primer-extension analyses (Sansam et al., 2003; Dawson et al., 2004).
|
; Burns et al., 1997; Rueter et al., 1999; Schaub and Keller, 2002). Deletion of the region containing both dsRBMs and the intervening linker (dsRBM1/2) reduced editing in all RNA targets to near background levels, as did deletion of dsRBM2 alone (dsRBM2), demonstrating that dsRBM2 is essential for the editing of perfect dsRNAs as well as naturally occurring ADAR2 substrates containing imperfect inverted repeats. Interestingly, deletion of dsRBM1 (dsRBM1) decreased editing in a substrate-dependent manner, because the extent of editing for GluR-2 (Q/R site) and 5-HT2CR (D-site) transcripts was <10% of that observed for wild-type eGFP-ADAR2 (Figure 5, B and C), whereas editing of the remaining ADAR2 substrates was variably reduced by 30–70% (Figure 5). These results not only show that the dsRBMs of ADAR2 are required for editing activity but also confirm previous observations in transfected cells (Figure 3) regarding the substrate-dependent contributions of each motif to site-specific editing.
|
; Ramos et al., 2000). These results suggest that although dsRBM1 is critical for the editing of specific ADAR2 substrates, the precise amino acid residues required for A-to-I conversion may be distinct from those necessary for dsRNA binding. By contrast, the K281A mutation caused a modest decrease (13–25%) in the extent of editing for the perfect dsRNA (D79N) and the GluR-2 (R/G) transcript, whereas editing efficiency for the GluR-2 (Q/R), 5-HT2CR (D), and ADAR2 (–1) sites was reduced by greater than 85% (Figure 5C). The observation that dsRBM2 was critical for efficient editing of all targets examined, yet only a subset of these substrates was affected by the K281A mutation, demonstrates the relative importance of specific amino acid residues in dsRBM2-dependent editing activity.
Function-dependent Replacement and Transposition of ADAR2 dsRBMs
The contribution of dsRBM1 to the site-selective editing of ADAR2 seems to be substrate dependent, whereas dsRBM2 is critical for enzymatic activity with all substrates examined (Figure 5), despite the high degree of conservation at both the levels of amino acid sequence and protein structure (Tian et al., 2004; Stefl et al., 2006). This functional inequality between dsRBM1 and dsRBM2 could result from subtle sequence or structural differences affecting RNA–protein interactions or the relative position of each motif in the ADAR2 protein. To determine whether the unique properties of each dsRBM regarding site-specific editing and nucleolar localization are a function of their relative position, three additional eGFP-ADAR2 mutants (Figure 6A) containing tandem copies of either dsRBM1 (dsRBM1/1) or dsRBM2 (dsRBM2/2), or a mutant in which the positions of the dsRBMs were interchanged (dsRBM2/1), were used for in vitro editing analysis with the same five ADAR2 substrates presented in Figure 5. To assess the level of mutant eGFP-ADAR2 expression in crude nuclear extracts from HEK293 cells, quantitative Western blotting analysis was performed (as in Figure 4) to adjust the enzyme input to equivalent levels for the in vitro reaction (Figure 6B). If the properties of each dsRBM in ADAR2 are dependent simply upon its relative position, replacement of either dsRBM with the alternate motif should maintain wild-type editing activity, yet such replacement mutations significantly reduced site-specific A-to-I conversion (Figure 6C). Replacement of dsRBM2 with dsRBM1 (dsRBM1/1) resulted in a near complete loss of editing for all substrates examined, providing a pattern of activity that was very similar to the dsRBM2 deletion mutant (dsRBM2). Similarly, replacement of dsRBM1 with dsRBM2 (dsRBM2/2) resulted in a pattern of editing that was nearly identical to protein lacking dsRBM1 (dsRBM1), suggesting that replacement of either RNA-binding motif was no better than deleting the domain. Consistent with these findings, an eGFP-ADAR2 mutant in which the positions of the dsRBMs were interchanged (dsRBM 2/1) also had negligible editing activity with any of the RNA substrates tested (our unpublished data). These results demonstrate that the differential properties of dsRBM1 and dsRBM2 to site-selective editing are related to their unique sequences/structures rather than their relative positions within ADAR2.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
), the interferon-mediated antiviral response (Williams, 2001), and the deamination of specific adenosine moieties as a consequence of RNA editing (Emeson and Singh, 2001; Bass, 2002). A majority of the proteins involved in the recognition of duplex RNA contain multiple copies of a highly conserved 65- to 75-aa dsRNA-binding motif, providing a molecular mechanism to facilitate the interaction of these proteins with their dsRNA targets (Fierro-Monti and Mathews, 2000; Doyle and Jantsch, 2002; Carlson et al., 2003; Tian et al., 2004). ADAR2 is a dsRNA-specific adenosine deaminase that also contains two tandem copies of this highly conserved motif (Tian et al., 2004). To determine whether these highly similar dsRBMs play equivalent roles in the nucleolar localization of ADAR2 and their ability to promote the deamination of selective adenosine residues, we have examined the subcellular localization of a series of wild-type and mutant eGFP-ADAR2 fusion proteins and their ability to catalyze site-specific editing events on multiple ADAR2 substrates. Results from these analyses have demonstrated a functional inequality between the two highly conserved dsRBMs, where dsRBM1 tends to play a greater role in localizing ADAR2 to the nucleolus and dsRBM2 is critical for the editing of ADAR2 targets. The relative nucleolar localization of ADAR2 depends primarily upon the number of functional dsRBMs present, whereas site-specific A-to-I conversion is also strongly affected by the nature and the organization of these motifs, indicating that each dsRBM possesses an intrinsic ability to recognize specific determinants in duplex RNAs. More importantly, the contributions of dsRBM1 and K281 to editing activity are substrate dependent, consistent with previous findings that dsRBMs can specifically recognize distinct structural determinants on naturally occurring ADAR2 targets (Stephens et al., 2004; Stefl et al., 2006).
ADAR2 has been shown to bind to perfect dsRNA regions and catalyze nonspecific editing (Melcher et al., 1996; Liu et al., 1999; Lehmann and Bass, 2000; Cho et al., 2003; Dawson et al., 2004), bind to specific editing sites (Ohman et al., 2000; Stephens et al., 2004; Stefl et al., 2006), or bind to other duplex regions with no productive A-to-I conversion (Klaue et al., 2003). All of these binding events contribute to macroscopic measurements of ADAR2 affinity for an RNA substrate, but they do not necessarily contribute to specific editing (Klaue et al., 2003), thereby uncoupling binding affinity from site-selective editing activity. Consistent with these observations, previous studies have demonstrated that a mutation in dsRBM1 (K127A) decreased binding affinity to a GluR-2 (R/G site) substrate (Macbeth et al., 2004), yet this mutation had no effect on editing of the R/G site (Figure 5). Given the disparity between site-specific editing and dsRBM binding, we focused on dsRBM–RNA interactions related to site-selective editing, rather than ADAR2-substrate affinity, using both deletion and substitution analyses to compare the role(s) that each dsRBM plays in the editing of multiple ADAR2 substrates.
Previous observations that dsRBMs bind to perfect dsRNA, in a sequence-independent manner (Schreiber et al., 1989; Ryter and Schultz, 1998; Ramos et al., 2000; Blaszczyk et al., 2004), have suggested that the dsRBMs of ADAR2 may recognize all of their RNA targets in a similar manner. Consistent with this idea, ADAR2 has been shown to nonspecifically deaminate adenosine moieties in a wide range of synthetic RNA duplexes (Melcher et al., 1996; Liu et al., 1999; Lehmann and Bass, 2000; Cho et al., 2003; Dawson et al., 2004). However, the substrate-dependent contribution of dsRBM1 to editing activity (Figure 5C) not only indicates that this motif makes specific contacts with imperfect dsRNA targets but also suggests that this dsRBM can interact with different structural determinants in each naturally occurring substrate. Further support for this model of ADAR2–RNA interaction has been provided by recent biochemical and NMR analyses demonstrating that both dsRBM1 and dsRBM2 can bind to unique structural determinants in different RNA targets encoding either the R/G or Q/R sites of GluR-2 transcripts (Stephens et al., 2004; Stefl et al., 2006).
How can the same dsRBMs in ADAR2 specifically recognize distinct sequence/structural determinants in multiple RNAs? Deletion analyses have demonstrated that dsRBM2 is critical for the editing of all RNA substrates examined, yet substitution (K281A) of a highly conserved lysine residue that resides on the RNA-binding surface of dsRBM2 only affects editing activity on a subset of these transcripts (Figure 5). Using a substrate containing the Q/R site, the K281A mutation reduced editing to background levels, suggesting that K281 is a key residue mediating dsRBM2–RNA interactions. By contrast, the same mutation had little effect on editing of the R/G site (Figure 5), indicating that dsRBM2 may use distinct amino acid residues to contact different RNA targets. The lack of universal importance for K281 has been further suggested by NMR analysis where no chemical shift change in the amide proton for this residue was observed upon binding of a GluR-2 transcript containing the R/G site (Stefl et al., 2006). If the ability of dsRBMs to use unique complements of amino acids to make specific RNA contacts is ubiquitous, this represents a powerful mechanism by which to achieve site-selective recognition for a broad range of potential targets. Although we cannot eliminate other possible interpretations, such as K281A-mediated structural alterations that affect binding in a substrate-specific manner, this possibility is unlikely because the corresponding lysines in other dsRBMs have been shown to make direct contact with dsRNA (Ryter and Schultz, 1998; Ramos et al., 2000) and that analogous mutations in PKR and Staufen ablated dsRNA binding (McMillan et al., 1995; Ramos et al., 2000).
The distinct effects observed for the deletion of dsRBM1 and dsRBM2 on editing activity (Figure 5) could result from the precise sequence/structure of each domain or their relative location within the ADAR2 protein. Replacement of one dsRBM with another (dsRBM1/1 and dsRBM2/2) provided patterns of editing activity that were similar to mutants lacking the replaced dsRBM, indicating that functional differences in A-to-I conversion come from subtle sequence/structure difference between these motifs. The lack of interchangeability between these domains further suggests that the unique binding preferences of each dsRBM contribute to site-selective editing activity.
The steady-state accumulation of ADAR2 in nucleoli has been suggested to represent a regulatory mechanism by which to modulate editing activity at its site of action in the nucleoplasm (Desterro et al., 2003; Sansam et al., 2003). The nucleolar localization of ADAR2 is dependent upon its dsRBMs and the presence of rRNA, suggesting that ADAR2 is targeted to the nucleolus by directly binding extended duplex regions in mature or pre-rRNA transcripts (Sansam et al., 2003). Several other dsRBM-containing proteins including ADAR1, PKR, Staufen, RNA helicase A, RNA helicase II/Gu, and NF-
B repressing factor have also been shown to accumulate in nucleoli (Tian and Mathews, 2001; Valdez et al., 2002; Desterro et al., 2003; Macchi et al., 2004; Niedick et al., 2004; Zhang et al., 2004), suggesting that dsRBM-mediated binding may serve as a general mechanism for nucleolar localization. Observations that deletion or mutation of either dsRBM can significantly affect the extent of nucleolar targeting (Figure 2) suggest that either both motifs are required for binding to specific dsRNAs or that each motif binds to duplex RNAs that represent only a subset of the total nucleolar ADAR2-binding sites. Unlike the significant reductions in editing activity observed by replacement or interchange of the dsRBMs, such modifications had little effect on ADAR2 subnuclear localization (Figure 6), suggesting that the nucleolar localization of ADAR2 results from a general dsRNA-binding activity for one or multiple nucleolar binding sites rather than the specific dsRBM–RNA interactions required for site-selective adenosine deamination.
The lack of obvious sequence requirements for the binding of dsRBM-containing proteins led to the initial conclusion that these motifs only confer general dsRNA-binding affinity with little sequence preference, yet most of the data regarding this hypothesis focused upon model RNA substrates that contained a perfect RNA duplex (Ryter and Schultz, 1998; Ramos et al., 2000; Blaszczyk et al., 2004). By contrast, more recent studies using naturally occurring dsRNAs, formed by intramolecular base pairing between imperfect, inverted repeats, have indicated that dsRBMs have unique functional properties based upon intrinsic binding preferences and affinities (Spanggord et al., 2002; Stephens et al., 2004; Stefl et al., 2006). Although the common features of the dsRBMs provide them with a general ability to interact with extended RNA duplexes, the subtle differences in amino acid sequence and structure between these domains also allow individual motifs to play diverse, substrate-specific roles that ultimately define the function(s) of their parent proteins.
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
The online version of this article contains supplemental material at MBC Online (http://www.molbiolcell.org). 
Address correspondence to: Ronald Emeson ( ron.emeson{at}vanderbilt.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ausubel, F., Brent, R., Kingston, R., Moore, D., Seidman, J., Smith, J., Struhl, K. (1998). In: Current Protocols in Molecular Biology, New York: John Wiley & Sons.
Bass, B. L. (2002). RNA editing by adenosine deaminases that act on RNA. Annu. Rev. Biochem 71, 817–846.[CrossRef][Medline]
Bass, B. L., Nishikura, K., Keller, W., Seeburg, P. H., Emeson, R. B., O’Connell, M. A., Samuel, C. E., Herbert, A. (1997). A standardized nomenclature for adenosine deaminases that act on RNA. RNA 3, 947–949.[Medline]
Berg, K. A., Cropper, J. D., Niswender, C. M., Sanders-Bush, E., Emeson, R. B., Clarke, W. P. (2001). RNA-editing of the 5-HT(2C) receptor alters agonist-receptor-effector coupling specificity. Br. J. Pharmacol 134, 386–392.[CrossRef][Medline]
Blaszczyk, J., Gan, J., Tropea, J. E., Court, D. L., Waugh, D. S., Ji, X. (2004). Noncatalytic assembly of ribonuclease III with double-stranded RNA. Structure 12, 457–466.[Medline]
Blow, M., Futreal, P. A., Wooster, R., Stratton, M. R. (2004). A survey of RNA editing in human brain. Genome. Res 14, 2379–2387.
Burns, C. M., Chu, H., Rueter, S. M., Hutchinson, L. K., Canton, H., Sanders-Bush, E., Emeson, R. B. (1997). Regulation of serotonin-2C receptor G-protein coupling by RNA editing. Nature 387, 303–308.[CrossRef][Medline]
Carlson, C. B., Stephens, O. M., Beal, P. A. (2003). Recognition of double-stranded RNA by proteins and small molecules. Biopolymers 70, 86–102.[CrossRef][Medline]
Cho, D. S., Yang, W., Lee, J. T., Shiekhattar, R., Murray, J. M., Nishikura, K. (2003). Requirement of dimerization for RNA editing activity of adenosine deaminases acting on RNA. J. Biol. Chem 278, 17093–17102.
Coolidge, C. J. and Patton, J. G. (1995). Run-around PCR: a novel way to create duplications using polymerase chain reaction. Biotechniques 18, 762–764.[Medline]
Dawson, T. R., Sansam, C. L., Emeson, R. B. (2004). Structure and sequence determinants required for the RNA editing of ADAR2 substrates. J. Biol. Chem 279, 4941–4951.
DeCerbo, J. and Carmichael, G. G. (2005). Retention and repression: fates of hyperedited RNAs in the nucleus. Curr. Opin. Cell Biol 17, 302–308.[CrossRef][Medline]
Desterro, J. M., Keegan, L. P., Lafarga, M., Berciano, M. T., O’Connell, M., Carmo-Fonseca, M. (2003). Dynamic association of RNA-editing enzymes with the nucleolus. J. Cell Sci 116, 1805–1818.
Doyle, M. and Jantsch, M. F. (2002). New and old roles of the double-stranded RNA-binding domain. J. Struct. Biol 140, 147–153.[CrossRef][Medline]
Emeson, R. and Singh, M. (2001). Adenosine-to-inosine RNA editing: substrates and consequences. In: RNA Editing, ed. B. Bass. Oxford, United Kingdom: Oxford University Press, 109–138.
Feng, Y., Sansam, C. L., Singh, M., Emeson, R. B. (2006). Altered RNA editing in mice lacking ADAR2 autoregulation. Mol. Cell Biol 26, 480–488.
Fierro-Monti, I. and Mathews, M. B. (2000). Proteins binding to duplexed RNA: one motif, multiple functions. Trends Biochem. Sci 25, 241–246.[CrossRef][Medline]
Higuchi, M., Single, F. N., Kohler, M., Sommer, B., Sprengel, R., Seeburg, P. H. (1993). RNA editing of AMPA receptor subunit GluR-B: a base-paired intron-exon structure determines position and efficiency. Cell 75, 1361–1370.[CrossRef][Medline]
Kim, D. D., Kim, T. T., Walsh, T., Kobayashi, Y., Matise, T. C., Buyske, S., Gabriel, A. (2004). Widespread RNA editing of embedded alu elements in the human transcriptome. Genome Res 14, 1719–1725.
Klaue, Y., Kallman, A. M., Bonin, M., Nellen, W., Ohman, M. (2003). Biochemical analysis and scanning force microscopy reveal productive and nonproductive ADAR2 binding to RNA substrates. RNA 9, 839–846.
Lakhlani, P. P., MacMillan, L. B., Guo, T. Z., McCool, B. A., Lovinger, D. M., Maze, M., Limbird, L. E. (1997). Substitution of a mutant alpha2a-adrenergic receptor via "hit and run" gene targeting reveals the role of this subtype in sedative, analgesic, and anesthetic-sparing responses in vivo. Proc. Natl. Acad. Sci. USA 94, 9950–9955.
Lehmann, K. A. and Bass, B. L. (2000). Double-stranded RNA adenosine deaminases ADAR1 and ADAR2 have overlapping specificities. Biochemistry 39, 12875–12884.[CrossRef][Medline]
Levanon, E. Y., et al. (2004). Systematic identification of abundant A-to-I editing sites in the human transcriptome. Nat. Biotechnol 22, 1001–1005.[CrossRef][Medline]
Liu, Y., Emeson, R. B., Samuel, C. E. (1999). Serotonin-2C receptor pre-mRNA editing in rat brain and in vitro by splice site variants of the interferon-inducible double-stranded RNA-specific adenosine deaminase ADAR1. J. Biol. Chem 274, 18351–18358.
Lomeli, H., Mosbacher, J., Melcher, T., Hoger, T., Geiger, J. R., Kuner, T., Monyer, H., Higuchi, M., Bach, A., Seeburg, P. H. (1994). Control of kinetic properties of AMPA receptor channels by nuclear RNA editing. Science 266, 1709–1713.
Maas, S., Melcher, T., Herb, A., Seeburg, P. H., Keller, W., Krause, S., Higuchi, M., O’Connell, M. A. (1996). Structural requirements for RNA editing in glutamate receptor pre-mRNAs by recombinant double-stranded RNA adenosine deaminase. J. Biol. Chem 271, 12221–12226.
Macbeth, M. R., Lingam, A. T., Bass, B. L. (2004). Evidence for auto-inhibition by the N terminus of hADAR2 and activation by dsRNA binding. RNA 10, 1563–1571.
Macbeth, M. R., Schubert, H. L., Vandemark, A. P., Lingam, A. T., Hill, C. P., Bass, B. L. (2005). Inositol hexakisphosphate is bound in the ADAR2 core and required for RNA editing. Science 309, 1534–1539.
Macchi, P., Brownawell, A. M., Grunewald, B., DesGroseillers, L., Macara, I. G., Kiebler, M. A. (2004). The brain-specific double-stranded RNA-binding protein Staufen 2, nucleolar accumulation and isoform-specific exportin-5-dependent export. J. Biol. Chem 279, 31440–31444.
McMillan, N. A., Carpick, B. W., Hollis, B., Toone, W. M., Zamanian-Daryoush, M., Williams, B. R. (1995). Mutational analysis of the double-stranded RNA (dsRNA) binding domain of the dsRNA-activated protein kinase, PKR. J. Biol. Chem 270, 2601–2606.
Melcher, T., Maas, S., Herb, A., Sprengel, R., Seeburg, P. H., Higuchi, M. (1996). A mammalian RNA editing enzyme. Nature 379, 460–464.[CrossRef][Medline]
Niedick, I., Froese, N., Oumard, A., Mueller, P. P., Nourbakhsh, M., Hauser, H., Koster, M. (2004). Nucleolar localization and mobility analysis of the NF-kappaB repressing factor NRF. J. Cell Sci 117, 3447–3458.
Niswender, C. M., Copeland, S. C., Herrick-Davis, K., Emeson, R. B., Sanders-Bush, E. (1999). RNA editing of the human serotonin 5-hydroxytryptamine 2C receptor silences constitutive activity. J. Biol. Chem 274, 9472–9478.
Ohman, M., Kallman, A. M., Bass, B. L. (2000). In vitro analysis of the binding of ADAR2 to the pre-mRNA encoding the GluR-B R/G site. RNA 6, 687–697.[Abstract]
Prasanth, K. V., Prasanth, S. G., Xuan, Z., Hearn, S., Freier, S. M., Bennett, C. F., Zhang, M. Q., Spector, D. L. (2005). Regulating gene expression through RNA nuclear retention. Cell 123, 249–263.[CrossRef][Medline]
Raitskin, O., Cho, D. S., Sperling, J., Nishikura, K., Sperling, R. (2001). RNA editing activity is associated with splicing factors in lnRNP particles: the nuclear pre-mRNA processing machinery. Proc. Natl. Acad. Sci. USA 98, 6571–6576.
Ramos, A., Grunert, S., Adams, J., Micklem, D. R., Proctor, M. R., Freund, S., Bycroft, M., St Johnston, D., Varani, G. (2000). RNA recognition by a Staufen double-stranded RNA-binding domain. EMBO J 19, 997–1009.[CrossRef][Medline]
Rueter, S. M., Burns, C. M., Coode, S. A., Mookherjee, P., Emeson, R. B. (1995). Glutamate receptor RNA editing in vitro by enzymatic conversion of adenosine to inosine. Science 267, 1491–1494.
Rueter, S. M., Dawson, T. R., Emeson, R. B. (1999). Regulation of alternative splicing by RNA editing. Nature 399, 75–80.[CrossRef][Medline]
Rueter, S. and Emeson, R. (1998). Adenosine-to-Inosine conversion in mRNA. In: Modification and Editing of RNA, ed. H. Grosjean and R. Benne. Washington, DC: ASM Press, 343–361.
Ryter, J. M. and Schultz, S. C. (1998). Molecular basis of double-stranded RNA-protein interactions: structure of a dsRNA-binding domain complexed with dsRNA. EMBO J 17, 7505–7513.[CrossRef][Medline]
Sansam, C. L., Wells, K. S., Emeson, R. B. (2003). Modulation of RNA editing by functional nucleolar sequestration of ADAR2. Proc. Natl. Acad. Sci. USA 100, 14018–14023.
Schaub, M. and Keller, W. (2002). RNA editing by adenosine deaminases generates RNA and protein diversity. Biochimie 84, 791–803.[Medline]
Schreiber, E., Matthias, P., Muller, M. M., Schaffner, W. (1989). Rapid detection of octamer binding proteins with ‘mini-extracts’, prepared from a small number of cells. Nucleic Acids Res 17, 6419.
Seeburg, P. H. and Hartner, J. (2003). Regulation of ion channel/neurotransmitter receptor function by RNA editing. Curr. Opin. Neurobiol 13, 279–283.[CrossRef][Medline]
Sommer, B., Kohler, M., Sprengel, R., Seeburg, P. H. (1991). RNA editing in brain controls a determinant of ion flow in glutamate-gated channels. Cell 67, 11–19.[CrossRef][Medline]
Sontheimer, E. J. and Carthew, R. W. (2005). Silence from within: endogenous siRNAs and miRNAs. Cell 122, 9–12.<
8 cells; mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001 compared with wild type). (C) Schematic diagram indicating the domain structures of eYFP-dsRBM1 and eYFP-dsRBM2 fusion proteins is presented; the coordinates of the dsRBM domain are indicated relative to the normal ADAR2 start codon. A representative micrograph of the subcellular localization of each fusion protein in NIH/3T3 cells is shown with corresponding Fn/Fo values (n 
