Skip to main content
NCBI home page
Nucleic Acids Research logoLink to Nucleic Acids Research
. 2003 Aug 15;31(16):4874–4881. doi: 10.1093/nar/gkg681

ADAR2 A→I editing: site selectivity and editing efficiency are separate events

Annika M Källman 1, Margareta Sahlin 1, Marie Öhman 1,*
PMCID: PMC169957  PMID: 12907730

Abstract

ADAR enzymes, adenosine deaminases that act on RNA, form a family of RNA editing enzymes that convert adenosine to inosine within RNA that is completely or largely double-stranded. Site-selective A→I editing has been detected at specific sites within a few structured pre-mRNAs of metazoans. We have analyzed the editing selectivity of ADAR enzymes and have chosen to study the naturally edited R/G site in the pre-mRNA of the glutamate receptor subunit B (GluR-B). A comparison of editing by ADAR1 and ADAR2 revealed differences in the specificity of editing. Our results show that ADAR2 selectively edits the R/G site, while ADAR1 edits more promiscuously at several other adenosines in the double-stranded stem. To further understand the mechanism of selective ADAR2 editing we have investigated the importance of internal loops in the RNA substrate. We have found that the immediate structure surrounding the editing site is important. A purine opposite to the editing site has a negative effect on both selectivity and efficiency of editing. More distant internal loops in the substrate were found to have minor effects on site selectivity, while efficiency of editing was found to be influenced. Finally, changes in the RNA structure that affected editing did not alter the binding abilities of ADAR2. Overall these findings suggest that binding and catalysis are independent events.

INTRODUCTION

Adenosine deaminases that act on RNA (ADARs) are members of a family of enzymes that catalyze hydrolytic deamination of adenosines to inosines within a variety of largely or completely double-stranded RNAs (dsRNAs). Two functional ADAR enzymes that share a common modular domain organization have been characterized in mammals, ADAR1 and ADAR2. Edited substrates have been found among cellular pre-mRNAs, viral RNAs and non-coding RNAs (reviewed in 1). Inosine is recognized as guanosine by the translational machinery, thus editing creates a functional G for A replacement with the potential to change the amino acid sequence in protein coding regions. There are several known ADAR substrates where site-selective editing gives rise to functionally important isoforms of proteins from a single encoded pre-mRNA. In mammals, pre-mRNA of receptor proteins involved in neurotransmission, including serotonin receptors (2) and glutamate receptors (3–5), are edited. The glutamate receptor subunit B (GluR-B) pre-mRNA has been studied extensively and is edited at several locations (reviewed in 6). One of the A→I changes in GluR-B pre-mRNA occurs at the R/G site, where the genome encoded AGA is translated as GGA after editing, causing a change from arginine to glycine. Receptors assembled with GluR-B subunits edited at the R/G site recover faster from desensitization (5). Another editing site in GluR-B is the Q/R site, where editing causes an amino acid change from glutamine to arginine. Editing and alternative splicing generate a glutamate receptor subunit diversity required for normal brain development and function (7).

The ADAR enzymes have been shown to have slightly different but overlapping specificities. For example, both ADAR1 and ADAR2 can efficiently edit GluR-B pre-mRNA at the R/G site in vitro while the Q/R site is only edited by ADAR2 (8,9). ADAR2 has been shown to be essential for normal development in mouse (10). The Q/R site is endogenously edited to nearly 100% in the mammalian brain with a 90% reduction in editing in homozygous ADAR2–/– mice (10). There are also indications that ADAR1 is important for development, as homozygous ADAR1–/– mice reveal an embryonic lethal phenotype with defects in the hematopoetic system. Both ADAR1 and ADAR2 are also expressed in tissues other than brain; it is therefore possible that additional unknown substrates are present in other tissues (8,11,12).

ADAR1 and ADAR2 differ in the number of dsRNA-binding motifs (dsRBM) present in the enzyme. There are three copies of dsRBMs in ADAR1, while ADAR2 contains two. The second dsRBM in ADAR1 has been shown not to be essential for enzyme activity, and the two dsRBM in ADAR2 are best aligned with the first and the third motifs in ADAR1 (8,13). Other dsRNA-binding proteins such as Drosophila Staufen and the RNA-dependent protein kinase are members of the same family of functionally diverse proteins containing one or more dsRBMs (14).

So far, a relatively small number of A→I editing sites have been found and it is still unclear how ADARs achieve specific recognition of these sites. The dsRNA structure is essential for ADAR editing, and all known substrates require an editing complementary sequence (reviewed in 1). Like most dsRNA-binding proteins, the ADAR enzymes generally bind dsRNA without sequence specificity. However, in a previous report we have shown that ADAR2 selectively binds to the R/G editing site in vitro (15). All selectively edited sites have been found in duplex regions interrupted by mismatches or loops (reviewed in 6,16). Promiscuous deamination, where as much as 30% of the adenosines are edited, has been observed for long stretches of completely base paired non-coding RNA (17,18). Mismatches and loops have been suggested to be factors that allow the ADAR enzyme to achieve site-selective deamination (19).

To better understand the mechanism of editing we have used an RNA substrate containing the R/G editing site of the GluR-B pre-mRNA to study the requirements for selective editing (Fig. 1). Several studies have previously shown that both ADAR1 and ADAR2 can efficiently edit the R/G site in vitro (8,20). However, in these reports the selectivity of the editing enzymes has not been taken into account. In the present study we show that there is a distinct difference in the selectivity of editing between ADAR1 and ADAR2. We show that ADAR2 can perform R/G site-selective editing in vitro, while ADAR1 editing is promiscuous. Further, by mutational analysis we have examined the importance of the RNA structure for selective ADAR2 editing.

Figure 1.

Figure 1

Open in a new tab

Schematic RNA secondary structure of the GluR-B stem–loop containing the R/G site. The R/G site is marked in bold. The arrows denote the mutations, nucleotide and position downstream of the R/G site, made to investigate the importance of internal loops for selective R/G editing.

MATERIALS AND METHODS

Preparation of RNA substrates

Part of the rat GluR-B gene, consisting of exon 13, including the R/G site, and a shortened intron 13, was T/A cloned into the pGEM-T vector system (Promega) and named pGluR-R/G short intron (GRG-SI) (21). The GRG-SI RNA was transcribed using the vector T7 promoter and gel purified as described below. Mutants of the conserved R/G stem–loop were produced by site-directed mutagenesis on the pGRG-SI plasmid using the QuikChange Site-Directed Mutagenesis kit (Stratagene). The following complementary dideoxy oligonucleotides (Interactiva) were used: C42U46 mutant, 5′C42U46 (5′-GCTCAATGTTGTTATACTATTCCACCTACCCTGTG-3′) and 3′C42U46 (5′-CACATCAGGGTAGGTGGAATAGTATAACAACATTGAGC-3′); G56 mutant, 5′G56 (5′-CCACCTACCGTGATGTGTCTTTAAGACTCTAACGG-3′) and 3′G56 (5′-CCGTTAGAGTCTTAAAGACACATCACGGTAGGTGG-3′); U56 mutant, 5′U56 (5′-CCACCTACCTTGATGTGTCTTTAAGACTCTAACGG-3′) and 3′U56 (5′-CCGTTAGAGTCTTAAAGACACATCAAGGTAGGTGG-3′); C42U46U56 mutant, U56 oligos (above) on the C42U46 mutated pGRG-SI plasmid; C42U46G56 mutant, G56 oligos (above) on the C42U46 mutated pGRG-SI plasmid.

Oligoribonucleotides used in the fluorescence measurements were synthesized by Dharmacon Research Inc. Three oligos were used to form wild-type and G56 stem duplexes. The two duplexes had one strand in common, where the fluorescent nucleic base analog 2-aminopurine (2-AP) replaced the adenosine at the R/G site. The opposite strand oligo had either wild-type sequence or the G56 mutation. The 2-AP containing strand and the two complementary strands, wild-type and G56, were deprotected in 100 mM acetic acid adjusted to pH 3.8 with TEMED at 60°C for 30 min according to the protocol provided by Dharmacon Research. The deprotected oligoribonucleotides were purified on a 20% polyacrylamide, 7 M urea PAGE gel, visualized using UV shadowing and eluted by crushing the gel slice and soaking overnight in 0.5 M NH4OAc, 0.1 mM EDTA, 0.1% SDS. The RNA was phenol/chloroform extracted and precipitated. Duplex annealing was performed with 42 pmol of the 2-AP containing strand and 62 pmol of the complementary strand in 30 µl of 1 mM Tris–HCl pH 8, 6 mM NaCl and 10 µM EDTA. The annealing reaction was heated to 65°C followed by slow cooling to room temperature. Duplex formation was analyzed by electrophoresis on a native 8% polyacrylamide gel (Fig. 4B; see below).

Figure 4.

Figure 4

Figure 4

Open in a new tab

(A) Fluorescence emission spectra of RNA samples. RNA oligonucleotides, 27 nt long, were used to make R/G substrate analogs with a 2-AP replacing the R/G site adenosine. Fluorescence intensity is plotted as a function of scanned emission wavelength (λex 310 nm) for 0.8 µM single-stranded 2-AP, 0.8 µM duplex 2-AP wild-type (2-AP WT) and 0.8 µM duplex G56 mutant (2-AP G56). The spectra are the average of nine scans. The buffer spectrum has been subtracted in all spectra. (B) Gel mobility shift assays showing RNA duplex formation and ADAR2 binding. Lane 1, single-stranded 2-AP; lane 2, duplex of 2-AP WT; lane 6, duplex of 2-AP G56; lanes 2–5, increasing amounts of ADAR2 were added to 20 fmol of wild-type 2-AP RNA duplex; lanes 7–9, ADAR2 added to 2-AP G56 duplex, as described for lanes 2–5. Gel retardation was analyzed on 8% native polyacrylamide gels.

Fluorescence measurements

Fluorescence measurements of two RNA duplexes, 2-AP WT and 2-AP G56, and the single-stranded 2-AP strand were performed on a Perkin Elmer LS-50B fluorometer. The excitation wavelength was 310 nm and the emission spectra were scanned between 335 and 430 nm. Slit widths of 10 nm were used for both excitation and emission. Ultra micro cuvettes from Hellma were used. In a total volume of 55 µl, 0.8 µM RNA was used in 25 mM Tris–HCl pH 8, 3 mM NaCl, 2 mM EDTA, 10% glycerol, 0.5 mM DTT and 100 mM KCl. The fluorescence from the buffer alone was subtracted from the different RNA samples.

In vitro editing assay

The ADAR enzymes were expressed as previously described (15). An aliquot of 300 fmol of wild-type or mutated GRG-SI RNA was incubated with 100 ng recombinant rat ADAR1 or ADAR2a in 100 µl containing 40 mM Tris–HCl pH 8, 5% glycerol, 25 mM KCl, 1.1 mM MgCl2, 1 mM DTT, 5 mM EDTA, 0.2 mM ATP, 0.16 U/µl RNasin (Promega). The reaction was incubated at 30°C for 30 min. Edited RNA was phenol/chloroform extracted and precipitated. Using AMV Reverse Transcriptase (Boehringer Mannheim) and primer Intron 13 (5′-CATACTTGTCAGACAGGGTGAGC-3′) RNA was reverse transcribed and the population of DNA sequences was PCR amplified using Taq polymerase (Life Technologies), primer A and Intron 13. The PCR products were purified using PCR Clean Up (Qiagen) and cloned into the pGEM-T Easy vector system (Promega). Separate clones were sequenced using primer A (DYEnamic ET terminator cycle sequencing kit; Amersham) at the KISeq sequence facility at Karolinska Institute (Stockholm, Sweden).

Gel mobility shift assay

Various amounts of rADAR2 (1.2–4.8 nM) were incubated with 20 fmol of [γ-32P]ATP 5′-end-labeled 2-AP WT or 2-AP G56 RNA in 10 mM Tris–HCl pH 8, 25 mM KCl, 10 mM NaCl, 1 mM MgCl, 0.5 mM DTT, 10% glycerol, 0.1 mg/ml BSA and 0.2 mM ATP. The reaction was incubated at 25°C for 10 min in a total volume of 10 µl. Reactions were analyzed by electrophoresis on native 8% polyacrylamide gels using a FLA 3000 Phosphorimager.

RESULTS

Editing selectivity of ADAR1 and ADAR2 at the R/G stem–loop

Selective in vitro editing was analyzed on an RNA substrate derived from glutamate receptor subunit B (GluR-B) pre-mRNA. For this study the GluR-B RNA used contains the R/G editing site situated in a 67 nt long stem–loop structure (GRG-SI) (21). The editing selectivity of ADAR1 and ADAR2 was compared on the GluR-B R/G substrate in an in vitro editing assay. The extent of editing in single isolated substrates was determined by sequencing of cloned RT–PCR products. Site-selective editing is here defined as a substrate edited only at the R/G site. Out of 21 sequenced clones, three were selectively edited at the R/G site by ADAR1 (Fig. 2A). The majority, 11 clones, had multiple edited sites in various combinations, not always including the R/G site position and with a distinct preference for the +39 site (the R/G site is 0). In contrast, ADAR2 editing resulted in a majority of the edited molecules (15 out of 19) being selectively edited specifically at the R/G site (Fig. 2A). Site-selective editing versus editing at other sites by ADAR1 and ADAR2 are compared in Figure 2B. In substrates edited by ADAR1 21% of the edited clones were selectively edited only at the R/G site. Although efficiency of editing at the R/G site was high for ADAR1, accompanying editing at other sites was also high, in particular at +39. The efficiency of editing at the R/G site was similar for ADAR1 and ADAR2. However, selective editing at the R/G site dominated the ADAR2 edited products (79%) (Fig. 2B). Few substrates were edited at other adenosines and interestingly no preference for editing at +39 was seen. The increased number of substrates promiscuously edited by ADAR1 compared to ADAR2 cannot be explained by a saturated concentration of ADAR1 since modification was totally absent in one-third of the analyzed substrates. Endogenously encoded GluR-B R/G RNA from rat brain was determined to be selectively edited at the R/G site, by us and others (data not shown; 22). Our data indicate that ADAR2 can perform site-selective in vitro editing, similar to what occurs in vivo, while ADAR1 shows a more promiscuous in vitro editing.

Figure 2.

Figure 2

Figure 2

Open in a new tab

Comparison of R/G site-selective editing by recombinant rat ADAR1 and ADAR2. RNA substrates were sequenced after in vitro editing, RT–PCR and cloning. (A) Distribution of edited sites in the clones. I represents an edited adenosine. The number of clones with the same editing pattern is indicated to the right. (B) Black bars represent the percentage of edited clones displaying selective editing at the R/G site only. Gray bars represent the percentage of edited clones with R/G site editing and additional edited sites. Light gray bars represent the percentage of edited clones with no R/G site editing but with editing at other sites.

The effect of internal loops on editing selectivity

Since ADAR2, unlike ADAR1, edits the wild-type R/G substrate with high selectivity for the R/G site in vitro, structural requirements for ADAR2 selective editing were focused upon. Site-selective editing has been found in RNA structures interrupted by bulges or loops. It has been hypothesized that these mismatched regions are used to create recognition sites for selective editing. The wild-type R/G stem–loop substrate harbors three internal loops. A mutational analysis was performed to investigate the importance of mismatched regions for selective editing (Figs 1 and 3A). In the C42U46 mutant two of the internal loops in the stem distant to the R/G site were closed, creating a completely paired stem in the vicinity of the mismatched R/G site (Figs 1 and 3A). This mutant RNA was incubated with ADAR2 and analyzed for editing as described above. As previously seen, the efficiency of editing at the R/G site in this mutant was decreased compared to the wild-type RNA (15). However, as illustrated in Figure 3, the selective R/G editing persisted; out of the nine clones with edited sites, eight had A→G changes only at the R/G site.

Figure 3.

Figure 3

Figure 3

Open in a new tab

The importance of internal loops for selective R/G editing. ADAR2 editing of wild-type RNA is compared to editing of four mutant RNA. (A) Distribution of edited sites in the clones. I represents an edited adenosine. The number of clones with the same editing pattern is indicated to the right. (B) Black bars represent the percentage of edited clones displaying selective editing at the R/G site only. Gray bars represent the percentage of edited clones with R/G editing and additional edited sites. Light gray bars represent the percentage of edited clones with no R/G site editing but with editing at other sites.

To further investigate the structural requirements for selective editing, a mutant was analyzed where the R/G site together with the two other internal loops were paired (C42U46U56). This results in a long perfectly paired stem (Figs 1 and 3A). Unlike the previously tested constructs, this mutant substrate was not dominated by editing specifically at the R/G site (Fig. 3B). Two novel sites (+11 and +44) appeared to be favored for ADAR2 editing. The lack of selective R/G editing of the C42U46U56 mutant could be due to pairing of the R/G site or the formation of a completely paired substrate. To investigate this, editing of a substrate with a single mutation (U56), pairing the R/G site, was analyzed. The mutation restored selective editing, although a large reduction in editing efficiency was observed (Fig. 3). This result indicates that the decrease in selective R/G editing found with the C42U46U56 mutant RNA was due to the extended helix structure, not the paired R/G site. Taken together, the two distal internal loops in the stem–loop structure seems to have a minor effect on selective R/G editing, although the efficiency of ADAR2 editing is negatively affected.

The importance of the editing complementary sequence for editing efficiency and site selectivity

Endogenously edited adenosines have been found in double-stranded regions with an opposing pyrimidine (U or C). To investigate the effect on editing selectivity, the wild-type cytosine opposing the R/G site was changed to a guanosine (G56) (Figs 1 and 3A). This mutation has previously been seen to have a negative effect on the efficiency of ADAR2 editing (20). Confirming previous results, few molecules of this transcript were edited. Further, only 1 out of the 22 sequenced clones were edited at the R/G site (Fig. 3A). Four other edited adenosines were detected in four independent clones, indicating that ADAR2 does not favor the R/G site over other adenosines. As illustrated in Figure 3B, site selectivity was severely affected by this mutation. Therefore, we conclude that the structure in close proximity to the R/G site is important for both editing efficiency and selectivity. However, we cannot exclude that the lack of selective editing is influenced by the decrease in editing efficiency.

The opposing base affects the structural conformation at the R/G site

Fluorescence measurements were used to determine if the G in the G56 mutant is causing a change in the immediate structure surrounding the R/G site, making it less favorable for editing. The adenosine at the R/G site was replaced by a fluorescent nucleic acid base analog, 2-AP. The electronic environment surrounding the 2-AP influences the quantum yield and emission maximum of the fluorescence (23). The 2-AP is highly fluorescent in single-stranded RNA, while it is quenched when stacked or base paired in a helix. Fluorescence measurements have previously been used to study conformational changes at the R/G site during catalysis of ADAR2 editing (24) and in studies of other nucleic acid modifying enzymes (25–28). In our analysis, the 27 nt long single-stranded RNA with a 2-AP at the R/G site showed an emission maximum at 364 nm (Fig. 4A). When the 2-AP RNA was hybridized to the editing complementary sequence (2-AP WT) the emission maximum decreased more than 12-fold compared to the single-stranded substrate, indicating that the adenosine at the R/G site is stacked within the helix. The fluorescence spectrum of a duplex 2-AP G56 substrate showed a 10 times higher intensity than that of the wild-type duplex (Fig. 4A). These results strongly suggest that the G mutation in the edited complementary sequence of G56 does not allow the 2-AP to be stacked within the helix, indicating that the G56 mutation creates a conformational change affecting the R/G site adenosine.

To confirm that the 2-AP G56 substrate can form a stable duplex structure, the migration of this substrate was compared with wild-type double-stranded (2-AP WT) and single-stranded (2-AP) RNAs. The 2-AP G56 and 2-AP WT substrates show similar migration in a non-denaturing polyacrylamide gel (Fig. 4B, lanes 2 and 6) that is retarded compared to the single-stranded 2-AP RNA (Fig. 4B, lane 1), indicating that not only the two strands in 2-AP WT but also in the 2-AP G56 substrate create a duplex conformation. Further, ADAR2 binding of the 2-AP WT and 2-AP G56 duplexes were investigated by gel mobility shift assay. An increasing amount of ADAR2 protein was added to the RNA duplexes and analyzed on a native polyacrylamide gel (Fig. 4B). The ADAR2 protein binds both duplexes, with no apparent difference. This implies that the conformational change induced by the G56 mutation does not affect ADAR2 binding.

DISCUSSION

The R/G site is efficiently edited by both ADAR1 and ADAR2, while other sites like the GluR-B Q/R site are only edited by ADAR2 (8,20,29,30). However, the ability of the ADAR enzymes to selectively edit these natural substrates in vitro has not been investigated. In an attempt to investigate how efficiency of editing relates to site selectivity we have used a GluR-B derivative to analyze editing at the R/G site. Selective editing is here defined as modification of single adenosines that has been verified as endogenously edited. Selective editing at the R/G site was compared to editing at other sites within the substrate using ADAR1 and ADAR2. To study selective editing we believe that it is important to analyze a natural substrate. Phylogenetic analysis of the sequence required for R/G editing reveals an exceptionally high conservation between organisms (30). This and the size of the R/G site hairpin makes the R/G site a suitable in vitro substrate to study the selectivity of A→I editing.

The efficiency of R/G editing is similar for ADAR1 and ADAR2. However, we show here that ADAR1 shows poor site selectivity, promiscuously editing other adenosines in the substrate (Fig. 2B). However, in the group of non-selectively edited adenosines there are some preferences for certain sites. In contrast, the majority of the molecules were selectively edited at the R/G site by ADAR2. Our results show the importance of analyzing editing of all adenosines in a substrate when the nature of editing at a specific site is under investigation. Also noteworthy is that the size of the GluR-B R/G derivative used in this analysis is in the range of 400 nt, however, edited adenosines were only detected in the double-stranded region of the R/G stem–loop. The results in this report give further evidence that ADAR1 and ADAR2 might have different roles in the cell. One hypothesis is that ADAR2 has the potential to edit specific nucleotides important for endogenous protein function, while ADAR1 catalyzes a more promiscuous editing that could play a role, for example, in viral defense, since one form of ADAR1 also is induced by γ-interferon (31). However, we cannot exclude that additional factors may be required to achieve site-selective ADAR1 editing in vitro.

Internal bulges and loops have been hypothesized to play an important role in limiting non-selective editing. We wanted to analyze whether there was an effect of structural changes in the R/G stem–loop on ADAR2 editing selectivity. The R/G stem contains three internal loops and the R/G site is situated in one of these mismatches. By changing the structure into a perfect stem using site-directed mutagenesis, selective R/G editing was reduced. Less than 30% of the edited molecules were selectively edited, compared to nearly 80% site-selective editing of the wild-type substrate (Fig. 3B). Therefore, this completely double-stranded molecule appears to no longer have the character of a selectively edited substrate. Recently, ADAR2 has been reported to interact with its substrate as a homodimer (32,33). In Cho et al. (33) it was suggested that the natural substrate induces cooperativity between the two monomers, while this is not the case for a non-selectively edited substrate that is completely double-stranded. Further, our previous analysis of the interaction between ADAR2 and the R/G stem–loop fused to a competing long synthetic dsRNA showed preferential binding to the selectively edited site (34). It is therefore possible that the mechanism of non-selective A→I modification differs from selective editing. If ADAR1 catalyzes a reaction that is exclusively of the non-selective type, this might explain the promiscuity of the enzyme. However, this cannot be determined by comparing the binding properties of ADAR1 and ADAR2. The enzymes have similar dissociation constants and by gel retardation analysis ADAR1 can be shifted to a higher order complex similar to ADAR2 (M.Öhman, unpublished results).

What determines the R/G site selectivity of this substrate? In a mutant substrate in which only two of the mismatches were closed and the R/G site was intact the selectivity persisted, although the efficiency of editing decreased. This suggests that the internal loops of the stem are not required for selective R/G editing and that site recognition and catalysis can be separate events. This is also in accordance with our previous results showing that this mutant substrate binds to the ADAR2 protein with the same affinity as the wild-type substrate (15). In Jaikaran et al. (32) it was suggested that an ADAR2 monomer binds in the vicinity of a selectively edited site but that the catalytic reaction requires association of a second ADAR2 monomer. This would explain how site recognition and catalysis could have different substrate preferences.

With an opposing C, the R/G site is not expected to be base paired. When the A:C mismatch was changed to an A:U base pair, site selectivity persisted, although the efficiency of editing was also reduced in this mutant substrate. Since the pyrimidines (C and U) are small compared to the more bulky purine bases this mutation is not likely to have a significant influence on the overall structure of the R/G stem–loop. When the C opposing the R/G adenosine was changed to a G, a 20-fold reduction in R/G editing efficiency was observed. This is also consistent with previous data showing that a purine (A or G) as the opposing base to the R/G adenosine gives low editing efficiency (20). The total number of edited molecules was low, with other adenosines in the stem edited to the same extent as the R/G site, indicating that the selectivity for R/G editing was lost. The low efficiency of editing at the R/G site in the G56 mutant can therefore, at least partly, be explained by the low selectivity. Further, other endogenous sites of editing frequently have A:C mismatches, but A:A or A:G have never been observed. Nucleoside deaminases such as adenosine deaminase (ADA) and cytidine deaminase share a consensus sequence similar to the deamination domain in the ADAR family. It has been proposed that ADAR2 uses a base flipping mechanism during catalysis, similar to what has been described for nucleoside deaminases and DNA methyltransferases (24,35). In a previous report an increased conformational flexibility of the bases opposite the R/G site during ADAR2 binding has been shown, possibly to facilitate base flipping of the adenosine (36). Binding and catalysis was also shown to induce conformational changes within the ADAR2 protein. Our fluorescent data shows that the opposing base is influencing the conformation of the R/G site adenosine. When a G replaces the C, the R/G site adenosine is unstacked, which is different from the wild-type substrate that allows the A to be stacked in the helix. When the larger guanosine opposes the R/G adenosine in a mutant substrate the necessary conformational changes in ADAR2 during catalysis might not occur, and editing will be inefficient. However, as seen in Figure 4B, the mutation has no effect on binding. Further, in a recent report we showed that the G56 mutant substrate competes less well with long synthetic dsRNA than the wild-type R/G stem–loop in preferential binding (34). This observation might be due to another mechanism of editing used on the G56 substrate, similar to the non-selective editing found in long completely double-stranded substrates.

In summary we can envision at least two reasons why purines are not found as the opposing base for edited sites: (i) they have an inhibitory effect on base flipping that causes inefficient editing; (ii) the mechanism of editing is non-selective and allows editing in the adjacent base paired regions. Further, we propose that the structure in close proximity to the edited site determines the nature of the catalytic mechanism and that bulges and internal loops in the adjacent region have a minor effect on site-selective editing.

The results in this report provide further knowledge on why only a few RNAs are selectively edited in vivo and how non-selective A→I editing of dsRNA is prevented.

Acknowledgments

ACKNOWLEDGEMENTS

We thank Britt-Marie Sjöberg and Patrick Young for critically reading this manuscript. Thanks also to Lars Wieslander and Eva Bratt for helpful discussions. This work was supported by grants from Swedish Natural Science Research Council, Carl Tryggers Stiftelse, Magnus Bergvalls Stiftelse, Nilsson-Ehle and Åke Wibergs Stiftelse.

REFERENCES

  • 1.Gerber A.P. and Keller,W. (2001) RNA editing by base deamination: more enzymes, more targets, new mysteries. Trends Biochem. Sci., 26, 376–384. [DOI] [PubMed] [Google Scholar]
  • 2.Burns C.M., Chu,H., Rueter,S.M., Hutchinson,L.K., Canton,H., Sanders-Bush,E. and Emeson,R.B. (1997) Regulation of serotonin-2C receptor G-protein coupling by RNA editing. Nature, 387, 303–308. [DOI] [PubMed] [Google Scholar]
  • 3.Higuchi M., Single,F.N., Kohler,M., Sommer,B., Sprengel,R. and 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. [DOI] [PubMed] [Google Scholar]
  • 4.Egebjerg J. and Heinemann,S.F. (1993) Ca2+ permeability of unedited and edited versions of the kainate selective glutamate receptor GluR6. Proc. Natl Acad. Sci. USA, 90, 755–759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Lomeli H., Mosbacher,J., Melcher,T., Hoger,T., Geiger,J.R., Kuner,T., Monyer,H., Higuchi,M., Bach,A. and Seeburg,P.H. (1994) Control of kinetic properties of AMPA receptor channels by nuclear RNA editing. Science, 266, 1709–1713. [DOI] [PubMed] [Google Scholar]
  • 6.Seeburg P.H., Higuchi,M. and Sprengel,R. (1998) RNA editing of brain glutamate receptor channels: mechanism and physiology. Brain Res. Brain Res. Rev., 26, 217–229. [DOI] [PubMed] [Google Scholar]
  • 7.Sprengel R., Higuchi,M., Monyer,H. and Seeburg,P.H. (1999) Glutamate receptor channels: a possible link between RNA editing in the brain and epilepsy. Adv. Neurol., 79, 525–534. [PubMed] [Google Scholar]
  • 8.Melcher T., Maas,S., Herb,A., Sprengel,R., Seeburg,P.H. and Higuchi,M. (1996) A mammalian RNA editing enzyme. Nature, 379, 460–464. [DOI] [PubMed] [Google Scholar]
  • 9.Yang J.H., Sklar,P., Axel,R. and Maniatis,T. (1997) Purification and characterization of a human RNA adenosine deaminase for glutamate receptor B pre-mRNA editing. Proc. Natl Acad. Sci. USA, 94, 4354–4359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Higuchi M., Maas,S., Single,F.N., Hartner,J., Rozov,A., Burnashev,N., Feldmeyer,D., Sprengel,R. and Seeburg,P.H. (2000) Point mutation in an AMPA receptor gene rescues lethality in mice deficient in the RNA-editing enzyme ADAR2. Nature, 406, 78–81. [DOI] [PubMed] [Google Scholar]
  • 11.Gerber A., O’Connell,M.A. and Keller,W. (1997) Two forms of human double-stranded RNA-specific editase 1 (hRED1) generated by the insertion of an Alu cassette. RNA, 3, 453–463. [PMC free article] [PubMed] [Google Scholar]
  • 12.Lai F., Chen,C.X., Carter,K.C. and Nishikura,K. (1997) Editing of glutamate receptor B subunit ion channel RNAs by four alternatively spliced DRADA2 double-stranded RNA adenosine deaminases. Mol. Cell. Biol., 17, 2413–2424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Lai F., Drakas,R. and Nishikura,K. (1995) Mutagenic analysis of double-stranded RNA adenosine deaminase, a candidate enzyme for RNA editing of glutamate-gated ion channel transcripts. J. Biol. Chem., 270, 17098–17105. [DOI] [PubMed] [Google Scholar]
  • 14.Fierro-Monti I. and Mathews,M.B. (2000) Proteins binding to duplexed RNA: one motif, multiple functions. Trends Biochem. Sci., 25, 241–246. [DOI] [PubMed] [Google Scholar]
  • 15.Öhman M., Källman,A.M. and 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Bass B.L. (2002) RNA editing by adenosine deaminases that act on RNA. Annu. Rev. Biochem., 71, 817–846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Morse D.P. and Bass,B.L. (1999) Long RNA hairpins that contain inosine are present in Caenorhabditis elegans poly(A)+ RNA. Proc. Natl Acad. Sci. USA, 96, 6048–6053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Morse D.P., Aruscavage,P.J. and Bass,B.L. (2002) RNA hairpins in noncoding regions of human brain and Caenorhabditis elegans mRNA are edited by adenosine deaminases that act on RNA. Proc. Natl Acad. Sci. USA, 99, 7906–7911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Lehmann K.A. and Bass,B.L. (1999) The importance of internal loops within RNA substrates of ADAR1. J. Mol. Biol., 291, 1–13. [DOI] [PubMed] [Google Scholar]
  • 20.Wong S.K., Sato,S. and Lazinski,D.W. (2001) Substrate recognition by ADAR1 and ADAR2. RNA, 7, 846–858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Bratt E. and Öhman,M. (2003) Coordination of editing and splicing of glutamate receptor pre-mRNA. RNA, 9, 309–318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Köhler M., Kornau,H.-C. and Seeburg,P.H. (1994) The organization of the gene for the functionally dominant α-amino-3-hydroxy-5-methyl-isoxazole-4-propionic acid receptor subunit GluR-B. J. Biol. Chem., 269, 17367–17370. [PubMed] [Google Scholar]
  • 23.Ward D.C., Cerami,A., Reich,E., Acs,G. and Altwerger,L. (1969) Biochemical studies of the nucleoside analogue, formycin. J. Biol. Chem., 244, 3243–3250. [PubMed] [Google Scholar]
  • 24.Stephens O.M., Yi-Brunozzi,H.Y. and Beal,P.A. (2000) Analysis of the RNA-editing reaction of ADAR2 with structural and fluorescent analogues of the GluR-B R/G editing site. Biochemistry, 39, 12243–12251. [DOI] [PubMed] [Google Scholar]
  • 25.Allan B.W. and Reich,N.O. (1996) Targeted base stacking disruption by the EcoRI DNA methyltransferase. Biochemistry, 35, 14757–14762. [DOI] [PubMed] [Google Scholar]
  • 26.Holz B., Klimasauskas,S., Serva,S. and Weinhold,E. (1998) 2-Aminopurine as a fluorescent probe for DNA base flipping by methyltransferases. Nucleic Acids Res., 26, 1076–1083. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Stivers J.T., Pankiewicz,K.W. and Watanabe,K.A. (1999) Kinetic mechanism of damage site recognition and uracil flipping by Escherichia coli uracil DNA glycosylase. Biochemistry, 38, 952–963. [DOI] [PubMed] [Google Scholar]
  • 28.Allan B.W., Reich,N.O. and Beechem,J.M. (1999) Measurement of the absolute temporal coupling between DNA binding and base flipping. Biochemistry, 38, 5308–5314. [DOI] [PubMed] [Google Scholar]
  • 29.Maas S., Melcher,T., Herb,A., Seeburg,P.H., Keller,W., Krause,S., Higuchi,M. and 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. [DOI] [PubMed] [Google Scholar]
  • 30.Aruscavage P.J. and Bass,L.B. (2000) A phylogenetic analysis reveals an unusual sequence conservation within introns involved in RNA editing. RNA, 6, 257–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.George C.X. and Samuel,C.E. (1999) Human RNA-specific adenosine deaminase ADAR1 transcripts possess alternative exon 1 structures that initiate from different promoters, one constitutively active and the other interferon inducible. Proc. Natl Acad. Sci. USA, 96, 4621–4626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Jaikaran D.C., Collins,C.H. and MacMillan,A.M. (2002) Adenosine to inosine editing by ADAR2 requires formation of a ternary complex on the GluR-B R/G site. J. Biol. Chem., 277, 37624–37629. [DOI] [PubMed] [Google Scholar]
  • 33.Cho D.S., Yang,W., Lee,J.T., Shiekhattar,R., Murray,J.M. and Nishikura,K. (2003) Requirement of dimerization for RNA editing activity of adenosine deaminases acting on RNA. J. Biol. Chem., 278, 17093–17102. [DOI] [PubMed] [Google Scholar]
  • 34.Klaue Y., Källman,A.M., Bonin,M., Nellen,W. and Öhman,M. (2003) Biochemical analysis and scanning force microscopy reveal productive and non-productive ADAR2 binding to RNA substrates. RNA, 7, 839–846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Hough R.F. and Bass,B.L. (1997) Analysis of Xenopus dsRNA adenosine deaminase cDNAs reveals similarities to DNA methyltransferases. RNA, 3, 1–15. [PMC free article] [PubMed] [Google Scholar]
  • 36.Yi-Brunozzi H.Y., Stephens,O.M. and Beal,P.A. (2001) Conformational changes that occur during an RNA-editing adenosine deamination reaction. J. Biol. Chem., 276, 37827–37833. [DOI] [PubMed] [Google Scholar]

Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press

RESOURCES