ORIGINAL ARTICLE
Inosine-specific cleavage of RNA for microarray analysis
of RNA A-to-I editing targets
Hurng-Wern Huang
a, Hsueh-Wei Chang
b,c,d, Hui-Chun Wang
b, Chiu-Jin Lu
a,
Shaio-Wen Chen
c, Szu-Cheng Yi
c, Hay-Yan Wang
e, Chung-Lung Cho
e,
Cheng-Hsuan Wu
f,g, Jyuer-Ger Yang
e,f,h, Chao-Neng Tseng
b,c,i,*
a
Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung, Taiwan
b
Graduate Institute of Natural Products, Kaohsiung Medical University, Kaohsiung, Taiwan
c
Department of Biomedical Science and Environmental Biology, Kaohsiung Medical University, Kaohsiung, Taiwan
d
Center of Excellence for Environmental Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan
e
Department of Biological Sciences, National Sun Yat-Sen University, Kaohsiung, Taiwan
f
Reproductive Medicine Center, Changhua Christian Hospital, Changhua, Taiwan
gInstitute of Medicine, Chung Shan Medical University, Taichung, Taiwan h
Department of Obstetrics and Gynecology, Changhua Christian Hospital, Changhua, Taiwan
i
Cancer Center, Kaohsiung Medical University Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan
Received 16 March 2012; accepted 25 April 2012 Available online 28 December 2012
KEYWORDS ADAR; Adenosine deamination; I-specific cleavage; Microarray; RNA editing
Abstract Dysregulations of RNA A-to-I editing are associated with developmental defects in mouse and human diseases. Although several methods of identifying RNA A-to-I editing sites are currently available, most of the critical editing targets responsible for the important bio-logical functions of adenosine deaminases that act on RNA (ADARs) remain unknown. Here we report a modified I-specific cleavage method that improves the quality of the RNA product. Preliminary microarray comparison of RNAs subjected to I-specific cleavage or mock digestion reported 165 genes that showed more than 0.2-fold reductions due to the cleavage. Six of the 165 genes were randomly selected for further verification, and three were verified to be targets of I-specific cleavage. This method may provide an alternative method of identifying novel RNA A-to-I editing sites using a microarray and facilitate the inquiry into the roles of RNA A-to-I editing in various biological processes.
Copyrightª 2012, Kaohsiung Medical University. Published by Elsevier Taiwan LLC. All rights reserved.
RETRACTED
* Corresponding author. Graduate Institute of Natural Products, Kaohsiung Medical University, 100 Shih-Chuan 1st Road, Kaohsiung City 807, Taiwan.
E-mail address:[email protected](C.-N. Tseng).
1607-551X/$36 Copyrightª 2012, Kaohsiung Medical University. Published by Elsevier Taiwan LLC. All rights reserved. http://dx.doi.org/10.1016/j.kjms.2012.08.031
Available online atwww.sciencedirect.com
Introduction
RNA adenosine-to-inosine (A-to-I) editing mediated by the adenosine deaminase that acts on RNA (ADAR) protein family is the most predominant form of single base editing in the animal kingdom (Fig. 1A) [1e3]. ADAR1 and ADAR2 are ubiquitously expressed, whereas ADAR3 is only found in the brain[4e6]. Because the base pairing property of I is similar to G, RNA A-to-I editing has equivalent effects of A-to-G mutations at the DNA level (Fig. 1B)[7]. Depending on the position of the edited sequence, the effects of A-to-I editing includes recoding of protein sequence, altering splicing signals, changing RNA secondary structure and stability, as well as the specificity of miRNAs[8].
A-to-I editing plays common and essential roles in the normal biological processes such as neural transmission, erythropoiesis, immune response, and energy metabo-lism. The nervous system contains the highest inosine content in brain mRNA and expresses the majority of the known editing targets[9]. Gria2, which encodes an AMPA (a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) receptor subunit GluRB contains the most prevalent editing site, the Q/R site. A-to-I editing of the Q/R site recodes the glutamine in the pore-forming reentrant loop to an arginine at greater than 99% efficiency [10]. The editing of the Q/R site regulates targeting, unitary conductance, and calcium permeability of AMPA recep-tors, as well as the neuronal susceptibility to hypoxia-induced excitotoxicity [11e14]. Lack of Q/R site editing in ADAR2/ mice results in progressive seizures and death shortly after birth [15]. Underediting of GluR2 in spinal motor neurons was also found in human cases of sporadic amyotrophic lateral sclerosis[16]. Htr2c encodes the serotonin 5-HT2C receptor that contains five closely spaced editing sites on the cytoplasmic loop interacting with downstream G proteins [17]. The five editing sites can recode three amino acids and affect the downstream
signaling[18]. ADAR1/knockout mouse embryos die at E13.5 because of a failure of erythropoiesis and wide-spread apoptosis [19e21], probably due to unchecked activation of the interferon pathway that induces apoptosis of hematopoitic stem cells[22,23]. Endotoxin-induced microvascular lung injury and systematic inflam-mation dramatically increases ADAR1 expression, causing 5% of adenosines in the thymic mRNA to be edited to inosines in the mouse [24,25]. ADAR2 expression is increased by a high fat diet and reduced by fasting in the pancreatic islets in mice [26]. Ubiquitous overexpression of ADAR2 results in adult-onset hyperphagia-mediated obesity [27].
Aberrant RNA A-to-I editing has also been implicated in human diseases and cancer development. The altered editing pattern of Htr2C greatly reduces serotonin signaling in depression suicide victims [28]. ADAR1 muta-tions have been found in human cases of dyscromatosis symmetrica hereditaria [29,30]. Reduced A-to-I editing activity has been reported in human brain, prostate, kidney, and testis tumors [31]. A reduction in ADAR2 expression is strongly correlated with the malignancy of glioblastoma multiforme [32]. Overexpression of ADAR reduces the proliferation of astrocytoma cell lines [33], but accelerates the cell cycle by up-regulating CDK2 in HEK293 cells[34].
Deciphering these roles of RNA A-to-I editing depends on the identification of the editing targets and the resulting changes of protein functions. Experimental approaches that can systematically identify and analyze novel A-to-I editing sites will be necessary to elucidate their function during development and disease. In 1997, Morse and Bass
[35]developed an specific cleavage method to cut the I-containing mRNAs. Guanidine nucleotides are modified by the formation of stable glyoxal/borate adducts and pro-tected from RNase T1 digestion that recognizes both guanidine and inosine residues under normal conditions
Figure 1. Identification of A-to-I editing target by I-specific cleavage. (A) ADAR (adenosine deaminase that acts on RNA) cata-lyzes the deamination reaction that converts adenosine to inosine. (B) A-to-I editing changes the original A:U pairing to I:C pairing. (C) The stable modification of guanidine with glyoxal and borate makes guanidine resistant to RNase T1 digestion.
(Fig. 1C). Using differential display techniques to analyze the RNA samples subjected to I-specific cleavage, Morse and Bass [36] identified five and 19 new editing sites in Caenorhabditis elegans and the human brain, respectively. In this work, we report the result of a microarray anal-ysis of RNAs subjected to a modified I-specific cleavage method. Microarray comparison of RNAs subjected to I-specific cleavage or mock digestion reported 165 genes that showed more than 0.2-fold reductions due to the cleavage. Six of the 165 genes were randomly selected for further verification, and three were verified to be targets of I-specific cleavage and therefore potential RNA A-to-I editing targets. This result suggests that microarray analysis of I-specifically cleaved RNA can be used to identify potential RNA A-to-I editing genes.
Materials and methods
Mouse brain RNA
Brain total RNA was extracted from 1-month-old ICR mice. The mouse brains were surgically removed and homoge-nized in 10 volumes of Trizol by 20 strokes in a Dounce homogenizer. Total RNA was then purified by following the manufacturer’s instructions. An additional Trizol extraction step was added in the end to remove any residual lipids or proteins.
I-specific cleavage
Glyoxyl/borate modification
First, 10mg of mouse brain RNA was glyoxylated in 100 mL of 10 mM sodium phosphate (pH 7.5), 50% dimethyl sulfoxide (DMSO), and 0.6% deionized glyoxal at 37C for 45 minutes. The RNA was further modified with borate by mixing with equal volume of 1 M sodium borate (pH 7.5) and precipi-tated with 500mL ethanol. The RNA was then redissolved in 15mL of 1 M sodium borate buffered with 10 mM Tris (pH 7.5).
RNase T1 digestion and removal
RNase T1 (100 U for exon array and 1000 U for oligo and expression array analysis) was added, and the nuclease reaction was carried out at 37C for 30 minutes. RNase T1 was then inactivated and removed by extraction with 10 volumes of Trizol. The resulting RNA pellet was dissolved in 400mL nuclease-free water.
Glyoxal/borate removal
To remove borate as completely as possible, the RNA was subjected to two rounds of NaOAc/ethanol precipitation at room temperature. The centrifugation steps for ethanol precipitation of RNA were carried out in a microcentrifuge at 12 000 rpm for 15 minutes. To remove glyoxal, the RNA pellet was resuspended in 400mL of 10 mM sodium phos-phate (pH 7.5) containing 50% DMSO and incubated at 60C for 2 hours, then precipitated by NaOAc/ethanol and dis-solved in 10 mL of nuclease-free water ready for down-stream application. Fifty to eighty percent of the starting RNA was routinely recovered at the completion of I-specific cleavage procedure.
Verification of I-specific cleavage on known editing targets with RT-PCR
One-step reverse transcription-polymerase chain reaction (RT-PCR) was performed using SuperScript III One-Step RT-PCR System with Platinum Taq (Invitrogen) with 10 ng RNA template. The primers designed to cover the editing sites in known editing targets of mouse nervous system were as follows: Gria2 forward, 50-TGGACTTATATGAGGAGTGCAG-30; Gria2 reverse, 50-GGATGTAGAATACTCCAGCAAC-30; Htr2c forward, 50-GTCCCTAGCCATTGCTGATATG-30; Htr2c reverse, 50-GCTTTCGTCCCTCAGTCCAATC-30; Kcna1 forward, 50-GGGT CATCCGCTTGGTAAGG-30; Kcna1 reverse, 50-CACTATCGGCAA TGAGCGGTTCC-30. Negative controls of I-specific cleavage were:b-actin forward, 50-CACTATCGGCAATGAGCGGTTCC-30; b-actin reverse, 50-TGCATCCTGTCAGCAATGCCTG-30 GAPDH forward, 50-GCACAGTCAAGGCCGAGAAT-30; GAPDH reverse: 50-GCCTTCTCCATGGTGGTGAA-30.
Double-strand cDNA synthesis and amplification
Double-stranded cDNA was synthesized from 200 ng total RNA and amplified using the Microarray Target Amplifica-tion Kit (Roche). Briefly, first and second strand cDNA were synthesized using the TAS-T7 Oligo(dT)24 primer and the TAS-(dN)10 primer, respectively, resulting in ds-cDNA with TAS tags on both ends. The ds-cDNA was purified using the Microarray Target Purification Kit, and one-quarter of the purified product was used as the template for a 100-mL PCR reaction that underwent 21 cycles of amplification. The amplified ds-cDNA was then again purified with the Micro-array Target Purification Kit.
Microarray analysis
Probe preparation, hybridization, and image acquisition for mouse expression arrays were carried out by the Affymetrix Gene Expression Service Lab of Academica Sinica (http://ipmb.
sinica.edu.tw/affy/). Five micrograms of amplified ds-cDNA
with T7 tag was submitted for whole transcriptome analysis using Mouse Gene Chip Array 430.2. Data normalization and comparison of expression arrays were performed using DNA-Chip Analyzer (dDNA-Chip)[37]. Probe set annotations and gene ontology information were retrieved from NetAffx Analysis Center (http://www.affymetrix.com/analysis/index.affx).
Results
Glyoxal/borate modification of RNA
I-specific cleavage consists of three major steps: (1) modi-fication of G with glyoxal and borate, (2) RNase T1 cleavage of inosine, and (3) removal of RNase T1, borate, and glyoxal. Because it involves high salt, heat, nuclease, and lengthy incubation times, this process often compromises the material for downstream applications.
To examine the status of the RNA during the process of modification and the removal of glyoxal and borate, aliquots of RNAs were removed at different steps of the procedure and analyzed by electrophoresis. The starting
mouse brain total RNA showed two major bands of 18 s and 28 s rRNAs (Fig. 2A, lane 1), which became single-stranded and difficult to visualize by EtBr staining after being denatured by glyoxal (Fig. 2A, lane 2). Total RNA modified by both glyoxal and borate showed a slower migration pattern, indicating increased molecular weight as a result of modification (Fig. 2A, lane 3). The recovered RNA showed smearing and trailing on agarose gel (Fig. 2A, lane 4) and failed as the template for one-step RT-PCR (data not shown). To improve the quality of RNA, we enhanced the removal of borate abducts from the modified RNA by adding two rounds of ethanol precipitation carried out at room temperature. The recovered RNA did not show signs of smearing and trailing on agarose gel after this enhanced removal of borate (Fig. 2B). The major bands of the rRNA reappeared in roughly the same migration position but diffused, likely due to the formation of random duplex structures during renaturation.
I-specific cleavage
We then examined the change in RNA size distribution caused by I-specific cleavage. The original step of removing RNase T1 by proteinase K digestion followed by phenol/ chloroform extraction was replaced by one simple Trizol
extraction. I-specific cleavage with 1000 U RNase T1 slightly reduced the overall size of mouse brain mRNA compared to control RNA with mock digestion, indicating that the modification of G residues was effective and the removal of RNase T1 by Trizol was complete (Fig. 3A). We routinely recover 50e80% of the starting RNAs with both 260/280 and 260/230 ratios exceeding 1.9 using the modified I-specific cleavage protocol.
The specific cleavage of known ADAR editing targets was verified with one-step RT-PCR using primers designed to cover the known editing site of each gene. After I-specific cleavage, no RT-PCR products were detected in highly edited genes such as Gria2 and Htr2c. In the case of Kcna1, whose editing efficiency is lower, the band of the RT-PCR product was weaker but still visible, whereas the b-actin and GAPDH bands were unchanged after I-specific cleavage
(Fig. 3B). These results indicate that the RNase T1 cleavage
was specific to inosine-containing RNA, and the quality of recovered RNA was sufficient for RT and PCR reactions.
Synthesis and amplification of double-strand cDNA
To ensure that the modified I-specific cleavage method generated high-quality RNAs suitable for microarray analysis, we then compared the length of the PCR-amplified cDNAs that
Figure 2. Glyoxal/borate modification of mouse brain total RNA. Aliquots of RNAs at different steps of the glyoxal/borate modification procedure were analyzed by 1% agarose gel electrophoresis using (A) the original protocol described by Moss et al. or (B) the modified protocol. Lane 1: mouse brain total RNA; lane 2: total RNA denatured by glyoxal; lane 3: total RNA modified by both glyoxal and borate; lane 4: total RNA after the removal of borate and glyoxal.
were reverse-transcribed from I-specifically cleaved RNA prepared by the original and the modified protocols, respec-tively. To facilitate the generation of aRNA probes for microarray hybridization, the Poly(dT)24 oligomers tagged with a T7 promoter sequence were used to prime the cDNA synthesis. The yields of cDNAs generated from RNAs treated with original I-specific cleavage protocol were lower, and the cDNA fragment lengths were shorter than 1 kbp (Fig. 3C). Using the modified protocol, both I-specifically cleaved RNA and control, mock digestion RNA yielded ds-cDNAs that showed similar even distributions between 300 bp and 3 kbp
(Fig. 3D). Taken together, these results indicate that the
modification of I-specific cleavage protocol improved the quality of the RNA product, making it more compatible for downstream molecular biology-based applications.
Analysis and preliminary verification of potential editing targets
We then tested if the targets of I-specific cleavage can be detected by expression microarray analysis. Brain RNA was extracted from four female mice and subjected to I-specific cleavage or mock digestion individually. Then T7-tagged ds-cDNAs were synthesized as described above, and pooled into two samples: one was subjected to I-specific cleavage and the other to mock digestion. They were submitted to the Affymetrix Gene Expression Service Lab of Academica Sinica for whole transcriptome analysis using Affymetrix Mouse Gene Chip Arrays 430.2. The data were normalized and compared using dCHIP software [37]. A total of 165 genes (4 of them were identified twice by different probe sets; Supplementary Table 1) were selected as potential editing targets based on these criteria: showing more than 0.2-fold reductions (1.2-fold change) and a difference
larger than 700 in signal intensity after I-specific cleavage, declared present before I-specific cleavage.
To verify the targets of I-specific cleavage identified by this microarray screening, six of the 169 potential editing targets (designated as ETs) were randomly selected for preliminary verification using a strategy that did not require sequencing the full length of cDNAs. The poly Aþ fragments of I-specific cleavage products were first purified using an mRNA purification kit and analyzed by one-step RT-PCR using primers designed to amplify the 50-most region of potential editing targets. The fragments 50 to the editing site would be lost during the mRNA purification process resulting in a reduction of RT-PCR templates in edited genes (Fig. 4A). We observed varying degrees of reduction of RT-PCR products after I-specific cleavage in all six potential editing targets screened in this manner (Fig. 4B). The densitometric analysis of results from replica experi-ments using brain RNAs from four female mice showed that three out of six potential editing targets showed a signifi-cant reduction of RT-PCR products after I-specific cleavage (Fig. 4C). This result suggests that microarray analysis of I-specifically cleaved RNA can be used to identify potential genes subject to RNA A-to-I editing.
Discussion
Experimental methods that can systematically identify A-to-I editing target genes can greatly increase our under-standing of the roles played by ADAR in biological processes including stress response, synaptic transmission, metabolic regulation, and tumorigenesis. Here, we report a modified inosine-specific cleavage method that, in combination with microarray analysis, allows for a systematic screening of potential editing targets.
Figure 3. I-specific cleavage of mouse brain total RNA and the synthesis and amplification of double-strand cDNA. (A) RNAse T1 cleavage of mouse brain RNA modified with glyoxal and borate. Agarose gel electrophoresis indicated that after I-specific cleavage the size distribution was slightly lower. (B) Reverse transcription-polymerase chain reaction (RT-PCR) verification of I-specific cleavage of known ADAR editing targets Gria2, Htr2c and Kcna1. RT-PCR was performed with primers designed to cover the known editing sites. The three known A-to-I editing targets showed clear reduction of RT-PCR products after I-specific cleavage but not the two housekeeping genesb-actin and GAPDH. The total RNA subjected to (C) the original I-specific cleavage protocol described by Moss et al. or (D) the modified protocol was reverse transcribed by using T7-Poly(dT) oligomer and amplified by PCR. The size distribution and the yield of the amplified cDNA showed clear improvement in the modified protocol.
The new protocol improved the quality of treated RNAs as judged by UV spectrometry and the efficiency for cDNA synthesis. The downstream RT became more efficient, generating cDNAs with a length of up to 3 kbp. RT-PCR analyses indicated that the cleavage of RNase T1 on RNA modified with glyoxal and borate was specific to inosine-containing RNAs, cutting previously known editing targets such as Gria2, 5HT2c, and Kcna1 but notb-actin or GAPDH. The microarray comparison of RNAs subjected to I-specific cleavage or mock digestion reported that 165 genes showed more than 0.2-fold reductions due to the cleavage. Gene ontology analysis showed that the major biological processes that these genes participated in included cell adhesion, cell cycle, metabolic process, RNA processing, transcription, and transport. Six of the 165 genes were
randomly selected for further verification, and three were shown to be targets of I-specific cleavage and therefore potential RNA A-to-I editing targets.
The sensitivity of detecting an inosine-containing site could be influenced by several factors including editing frequency, distance of the editing site to the probe set, and degrees of RNase T1 digestion. The candidate genes iden-tified by this microarray study are very likely edited in the 30 UTR. The probe sets of expression arrays are usually designed to target the 30regions of transcripts. Because the hybridization probes of expression array are generated by RT from the 30 end, RNAs that have been cut in the 50 regions would not show a significant reduction of the hybridization signal in probe sets targeting regions 30end of the transcript. In fact, recoding of an amino acid sequence
Figure 4. Preliminary verification of potential editing targets identified by mouse expression microarray analysis. (A) To verify if a gene is an editing target, the poly Aþ fragments are recovered after I-specific cleavage of total RNA and subjected to RT-PCR. Primers (indicated by arrows) are designed to amplify near the 50-most regions as possible by one-step RT-PCR. I-specific cleavage results in loss of fragments 50to the editing site and reduction of the RT-PCR products. (B) Six potential editing targets (designated as ETs) were randomly selected based on the analysis results using Mouse Gene Chip Array 430.2. These potential editing targets showed different degrees of reduction of RT-PCR products after I-specific cleavage. (C) Densitometric analysis of these six editing targets by using brain RNAs from four female mice. Three out of six potential editing targets showed significant reduction of RT-PCR product after I-specific cleavage (mean SD, *p < 0.05).
by A-to-I editing only occurs in a small number of genes, and most editing sites are found in noncoding regions[38]. In the expression array analysis of this study, out of the total 45,101 probe sets, 35,726 (79.2%) were declared marginal or absent in both arrays. The percentage of present probe sets for mock digestion and I-specific cleavage samples was 16.4% and 16.2%, respectively. The 30/50ratio ofb-actin and GAPDH was more than 20, indicating that the modified I-specific cleavage protocol still resulted in certain degrees of deterioration in RNA quality. Extension of cDNA synthesis to the 50 end of mRNA was inefficient probably because of random breaks in the mRNA or the incomplete removal of modification. Further optimization of the I-specific cleavage protocol may help to strike a balance between degrees of modification, cleavage, and the quality of the RNA sample. Possible improvements may also be made by using new strategies of I-specific cleavage, for example, by developing alternative methods of chemical modification of guanidine that are efficient and can be easily removed, or by engi-neering RNase T1 to redirect its specificity toward inosine. RNase T1 has been shown to be amenable to directed mutagenesis that alters its nucleotide specificity[39].
In summary, we have established a modified I-specific cleavage method that improves the quality of the RNA product. The preliminary analysis of the cleavage sites by using expression microarray analysis has indentified 165 potential targets of I-specific cleavage. This method provides an additional, alternative strategy to current methods of identifying novel RNA A-to-I editing sites using microarray[40]and will facilitate the inquiry into the roles of RNA A-to-I editing in various biological processes.
Acknowledgments
This work was supported in part by the Kaohsiung Medical University Research Foundation under the grants KMU-M100001, KMU-M110010 (to C.N.T.), NSYSU-KMU Joint Research Project #NSYSUKMU 101-02 (to C.N.T. and C.L.C.), and KMU-EM-99-1.4 (to H.W.C.); and by the Department of Health, Executive Yuan, R.O.C. (Taiwan) under the grant DOH100-TD-C-111-002 (to C.N.T.).
Appendix A. Supplementary data
Supplementary data related to this article can be found at
http://dx.doi.org/10.1016/j.kjms.2012.08.031.
References
[1] Bass BL. How does RNA editing affect dsRNA-mediated gene silencing? Cold Spring Harb Symp Quant Biol 2006;71:285e92. [2] Bass BL, Weintraub H. An unwinding activity that covalently modifies its double-stranded RNA substrate. Cell 1988;55: 1089e98.
[3] Wagner RW, Smith JE, Cooperman BS, Nishikura K. A double-stranded RNA unwinding activity introduces structural alter-ations by means of adenosine to inosine conversions in mammalian cells and Xenopus eggs. Proc Natl Acad Sci USA 1989;86:2647e51.
[4] Kim U, Wang Y, Sanford T, Zeng Y, Nishikura K. Molecular cloning of cDNA for double-stranded RNA adenosine
deaminase, a candidate enzyme for nuclear RNA editing. Proc Natl Acad Sci USA 1994;91:11457e61.
[5] Kim U, Garner TL, Sanford T, Speicher D, Murray JM, Nishikura K. Purification and characterization of double-stranded RNA adenosine deaminase from bovine nuclear extracts. J Biol Chem 1994;269:13480e9.
[6] Chen CX, Cho DS, Wang Q, Lai F, Carter KC, Nishikura K. A third member of the RNA-specific adenosine deaminase gene family, ADAR3, contains both single- and double-stranded RNA binding domains. RNA 2000;6:755e67.
[7] Basilio C, Wahba AJ, Lengyel P, Speyer JF, Ochoa S. Synthetic polynucleotides and the amino acid code, V. Proc Natl Acad Sci USA 1962;48:613e6.
[8] Nishikura K. Editor meets silencer: crosstalk between RNA editing and RNA interference. Nat Rev Mol Cell Biol 2006;7: 919e31.
[9] Paul MS, Bass BL. Inosine exists in mRNA at tissue-specific levels and is most abundant in brain mRNA. EMBO J 1998;17: 1120e7.
[10] Sommer B, Kohler M, Sprengel R, Seeburg PH. RNA editing in brain controls a determinant of ion flow in glutamate-gated channels. Cell 1991;67:11e9.
[11] Hume RI, Dingledine R, Heinemann SF. Identification of a site in glutamate receptor subunits that controls calcium perme-ability. Science 1991;253:1028e31.
[12] Burnashev N, Monyer H, Seeburg PH, Sakmann B. Divalent ion permeability of AMPA receptor channels is dominated by the edited form of a single subunit. Neuron 1992;8:189e98. [13] Peng PL, Zhong X, Tu W, Soundarapandian MM, Molner P,
Zhu D, et al. ADAR2-dependent RNA editing of AMPA receptor subunit GluR2 determines vulnerability of neurons in forebrain ischemia. Neuron 2006;49:719e33.
[14] Liu S, Lau L, Wei J, Zhu D, Zou S, Sun HS, et al. Expression of Ca(2þ)-permeable AMPA receptor channels primes cell death in transient forebrain ischemia. Neuron 2004;43:43e55. [15] Higuchi M, Maas S, Single FN, Hartner J, Rozov A, Burnashev N,
et al. Point mutation in an AMPA receptor gene rescues lethality in mice deficient in the RNA-editing enzyme ADAR2. Nature 2000;406:78e81.
[16] Kawahara Y, Ito K, Sun H, Aizawa H, Kanazawa I, Kwak S. Glutamate receptors: RNA editing and death of motor neurons. Nature 2004;427:801.
[17] Burns CM, Chu H, Rueter SM, Hutchinson LK, Canton H, Sanders-Bush E, et al. Regulation of serotonin-2C receptor G-protein coupling by RNA editing. Nature 1997;387:303e8. [18] Price RD, Weiner DM, Chang MS, Sanders-Bush E. RNA editing
of the human serotonin 5-HT2C receptor alters receptor-mediated activation of G13 protein. J Biol Chem 2001;276: 44663e8.
[19] Hartner JC, Schmittwolf C, Kispert A, Muller AM, Higuchi M, Seeburg PH. Liver disintegration in the mouse embryo caused by deficiency in the RNA-editing enzyme ADAR1. J Biol Chem 2004;279:4894e902.
[20] Wang Q, Khillan J, Gadue P, Nishikura K. Requirement of the RNA editing deaminase ADAR1 gene for embryonic erythro-poiesis. Science 2000;290:1765e8.
[21] Wang Q, Miyakoda M, Yang W, Khillan J, Stachura DL, Weiss MJ, et al. Stress-induced apoptosis associated with null mutation of ADAR1 RNA editing deaminase gene. J Biol Chem 2004;279:4952e61.
[22] XuFeng R, Boyer MJ, Shen H, Li Y, Yu H, Gao Y, et al. ADAR1 is required for hematopoietic progenitor cell survival via RNA editing. Proc Natl Acad Sci USA 2009;106:17763e8.
[23] Iizasa H, Nishikura K. A new function for the RNA-editing enzyme ADAR1. Nat Immunol 2009;10:16e8.
[24] Rabinovici R, Kabir K, Chen M, Su Y, Zhang D, Luo X, et al. ADAR1 is involved in the development of microvascular lung injury. Circ Res 2001;88:1066e71.
[25] Yang JH, Luo X, Nie Y, Su Y, Zhao Q, Kabir K, et al. Wide-spread inosine-containing mRNA in lymphocytes regulated by ADAR1 in response to inflammation. Immunology 2003;109: 15e23.
[26] Gan Z, Zhao L, Yang L, Huang P, Zhao F, Li W, et al. RNA editing by ADAR2 is metabolically regulated in pancreatic islets and beta-cells. J Biol Chem 2006;281:33386e94. [27] Singh M, Kesterson RA, Jacobs MM, Joers JM, Gore JC,
Emeson RB. Hyperphagia-mediated obesity in transgenic mice misexpressing the RNA-editing enzyme ADAR2. J Biol Chem 2007;282:22448e59.
[28] Niswender CM, Herrick-Davis K, Dilley GE, Meltzer HY, Overholser JC, Stockmeier CA, et al. RNA editing of the human serotonin 5-HT2C receptor. Alterations in suicide and implications for serotonergic pharmacotherapy. Neuropsychopharmacology 2001;24:478e91.
[29] Xing Q, Wang M, Chen X, Qian X, Qin W, Gao J, et al. Identi-fication of a novel ADAR mutation in a Chinese family with dyschromatosis symmetrica hereditaria (DSH). Arch Dermatol Res 2005;297:139e42.
[30] Miyamura Y, Suzuki T, Kono M, Inagaki K, Ito S, Suzuki N, et al. Mutations of the RNA-specific adenosine deaminase gene (DSRAD) are involved in dyschromatosis symmetrica hered-itaria. Am J Hum Genet 2003;73:693e9.
[31] Paz N, Levanon EY, Amariglio N, Heimberger AB, Ram Z, Constantini S, et al. Altered adenosine-to-inosine RNA editing in human cancer. Genome Res 2007;17:1586e95.
[32] Maas S, Patt S, Schrey M, Rich A. Underediting of glutamate receptor GluR-B mRNA in malignant gliomas. Proc Natl Acad Sci USA 2001;98:14687e92.
[33] Cenci C, Barzotti R, Galeano F, Corbelli S, Rota R, Massimi L, et al. Down-regulation of RNA editing in pediatric astrocy-tomas: ADAR2 editing activity inhibits cell migration and proliferation. J Biol Chem 2008;283:7251e60.
[34] Zhang F, Rabinovici R. Adenosine deaminase acting on RNA 1 accelerates cell cycle through increased translation and activity of cyclin-dependent kinase 2. Shock 2007;27:214e9. [35] Morse DP, Bass BL. Detection of inosine in messenger RNA by
inosine-specific cleavage. Biochemistry 1997;36:8429e34. [36] Morse DP, Bass BL. Long RNA hairpins that contain inosine are
present in Caenorhabditis elegans poly(A)þ RNA. Proc Natl Acad Sci USA 1999;96:6048e53.
[37] Li C, Wong WH. Model-based analysis of oligonucleotide arrays: expression index computation and outlier detection. Proc Natl Acad Sci USA 2001;98:31e6.
[38] Hundley HA, Bass BL. ADAR editing in double-stranded UTRs and other noncoding RNA sequences. Trends Biochem Sci 2010;7:377e83.
[39] Struhalla M, Czaja R, Hahn U. Addressing the challenge of changing the specificity of RNase T1 with rational and evolu-tionary approaches. Chembiochem 2004;5:200e5.
[40] Ohlson J, Enstero¨ M, Sjo¨berg BM, Ohman M. A method to find tissue-specific novel sites of selective adenosine deamination. Nucleic Acids Res 2005;33:e167.
