Limits...
ADAR regulates RNA editing, transcript stability, and gene expression.

Wang IX, So E, Devlin JL, Zhao Y, Wu M, Cheung VG - Cell Rep (2013)

Bottom Line: To study the role of ADAR proteins in RNA editing and gene regulation, we sequenced and compared the DNA and RNA of human B cells.The results uncovered over 60,000 A-to-G editing sites and several thousand genes whose expression levels are influenced by ADARs.Our results also reveal that ADAR regulates transcript stability and gene expression through interaction with HuR (ELAVL1).

View Article: PubMed Central - PubMed

Affiliation: Department of Genetics, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA. Electronic address: ixwang@umich.edu.

Show MeSH
Features of A-to-G and RDD Sites(A) The nucleotide 5′ to A-to-G sites is depleted of G, and the nucleotide 3′ to A-to-G sites is enriched for G. In contrast, the nucleotide 3′ to G-to-A sites is enriched for T, and the nucleotide 3′ to T-to-C sites is enriched for G. Sequences for 10 nt upstream and downstream of A-to-G or RDD sites were analyzed and the frequencies of A, C, T, and G at each position are shown. The horizontal line at a frequency of 0.25 indicates the expected frequency if the four nucleotides are represented equally.(B) A-to-G and other RDD sites are found in different genomic regions. Upper panel: genome-wide distribution (“Mixed” indicates regions with multiple or ambiguous annotation). Lower panel: distribution in exonic regions.(C) Sequence motifs for editing targets pulled down in anti-ADAR RNA-IP assays. The MEME program was used to analyze DNA sequences corresponding to 100 nt upstream and downstream of editing sites. The four motifs that are most significantly enriched in input sequences are shown (p < 10−10, Fisher’s exact test). Scrambled sequences were used as negative-control sequences.(D and E) Expression levels of transcripts do not correlate with editing levels. RPKM values of transcripts measured in an ADAR1 knockdown sample and a negative-control sample (NTC) are plotted. Edited and nonedited transcripts are indicated in different colors.(D) All transcripts.(E) Genes encoding zinc-finger proteins whose expression levels changed by ≥20%.See also Table S8.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3935819&req=5

Figure 5: Features of A-to-G and RDD Sites(A) The nucleotide 5′ to A-to-G sites is depleted of G, and the nucleotide 3′ to A-to-G sites is enriched for G. In contrast, the nucleotide 3′ to G-to-A sites is enriched for T, and the nucleotide 3′ to T-to-C sites is enriched for G. Sequences for 10 nt upstream and downstream of A-to-G or RDD sites were analyzed and the frequencies of A, C, T, and G at each position are shown. The horizontal line at a frequency of 0.25 indicates the expected frequency if the four nucleotides are represented equally.(B) A-to-G and other RDD sites are found in different genomic regions. Upper panel: genome-wide distribution (“Mixed” indicates regions with multiple or ambiguous annotation). Lower panel: distribution in exonic regions.(C) Sequence motifs for editing targets pulled down in anti-ADAR RNA-IP assays. The MEME program was used to analyze DNA sequences corresponding to 100 nt upstream and downstream of editing sites. The four motifs that are most significantly enriched in input sequences are shown (p < 10−10, Fisher’s exact test). Scrambled sequences were used as negative-control sequences.(D and E) Expression levels of transcripts do not correlate with editing levels. RPKM values of transcripts measured in an ADAR1 knockdown sample and a negative-control sample (NTC) are plotted. Edited and nonedited transcripts are indicated in different colors.(D) All transcripts.(E) Genes encoding zinc-finger proteins whose expression levels changed by ≥20%.See also Table S8.

Mentions: The results from ADAR knockdown and RNA-IP suggest that although ADARs mediate A-to-G editing, they do not mediate other types of RDDs. The levels of other types of differences were largely unaffected by ADAR knockdown, and the transcripts that showed those differences were not bound by ADAR. This prompted us to compare the genomic features surrounding the A-to-G editing sites and other types of RDDs. First, the sequence contexts of A-to-G and non-A-to-G sites are different. The base 5′ adjacent to the adenosine in A-to-G sites is depleted of guanosine (G) and the base 3′ to A-to-G editing sites is enriched for G (Figure 5A), consistent with previous reports (Lehmann and Bass, 2000). This sequence feature is specific to A-to-G editing because it is not present in random adenosines within nonedited Alu repeats (data not shown). This sequence motif was also not found for any of the RDDs. We identified sequence motifs for G-to-A and T-to-C sites, and they differed from the motif around the A-to-G sites (Figure 5A). Second, the A-to-G sites were more clustered than the non-A-to-G sites (67% of A-to-G sites were found within 25 nt of each other, compared with 14% of non-A-to-G RDDs). Third, most of the A-to-G sites were within or near inverted repeats, which form dsRNA and are preferentially recognized and bound by ADAR enzymes. Nearly 45% of the A-to-G sites resided within inverted repeats and another 30% were found near inverted repeats (<1 kb). In contrast, very few (0.9%) of the non-A-to-G sites were found in inverted repeats. Lastly, A-to-G sites and RDD sites were found in different regions of genes. A-to-G sites were found mostly in the 3′ UTRs, whereas RDDs were found mainly in the 5′ UTRs and in coding exons. Only 4% of the A-to-G sites (compared with 35% of RDDs) were in coding exons (Figure 5B). The differences between A-to-G editing sites and the other types of RDDs suggest that they are mediated by different mechanisms. Biochemically, this is expected since some of the RDDs are transversion events that cannot be explained simply by deamination.


ADAR regulates RNA editing, transcript stability, and gene expression.

Wang IX, So E, Devlin JL, Zhao Y, Wu M, Cheung VG - Cell Rep (2013)

Features of A-to-G and RDD Sites(A) The nucleotide 5′ to A-to-G sites is depleted of G, and the nucleotide 3′ to A-to-G sites is enriched for G. In contrast, the nucleotide 3′ to G-to-A sites is enriched for T, and the nucleotide 3′ to T-to-C sites is enriched for G. Sequences for 10 nt upstream and downstream of A-to-G or RDD sites were analyzed and the frequencies of A, C, T, and G at each position are shown. The horizontal line at a frequency of 0.25 indicates the expected frequency if the four nucleotides are represented equally.(B) A-to-G and other RDD sites are found in different genomic regions. Upper panel: genome-wide distribution (“Mixed” indicates regions with multiple or ambiguous annotation). Lower panel: distribution in exonic regions.(C) Sequence motifs for editing targets pulled down in anti-ADAR RNA-IP assays. The MEME program was used to analyze DNA sequences corresponding to 100 nt upstream and downstream of editing sites. The four motifs that are most significantly enriched in input sequences are shown (p < 10−10, Fisher’s exact test). Scrambled sequences were used as negative-control sequences.(D and E) Expression levels of transcripts do not correlate with editing levels. RPKM values of transcripts measured in an ADAR1 knockdown sample and a negative-control sample (NTC) are plotted. Edited and nonedited transcripts are indicated in different colors.(D) All transcripts.(E) Genes encoding zinc-finger proteins whose expression levels changed by ≥20%.See also Table S8.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3935819&req=5

Figure 5: Features of A-to-G and RDD Sites(A) The nucleotide 5′ to A-to-G sites is depleted of G, and the nucleotide 3′ to A-to-G sites is enriched for G. In contrast, the nucleotide 3′ to G-to-A sites is enriched for T, and the nucleotide 3′ to T-to-C sites is enriched for G. Sequences for 10 nt upstream and downstream of A-to-G or RDD sites were analyzed and the frequencies of A, C, T, and G at each position are shown. The horizontal line at a frequency of 0.25 indicates the expected frequency if the four nucleotides are represented equally.(B) A-to-G and other RDD sites are found in different genomic regions. Upper panel: genome-wide distribution (“Mixed” indicates regions with multiple or ambiguous annotation). Lower panel: distribution in exonic regions.(C) Sequence motifs for editing targets pulled down in anti-ADAR RNA-IP assays. The MEME program was used to analyze DNA sequences corresponding to 100 nt upstream and downstream of editing sites. The four motifs that are most significantly enriched in input sequences are shown (p < 10−10, Fisher’s exact test). Scrambled sequences were used as negative-control sequences.(D and E) Expression levels of transcripts do not correlate with editing levels. RPKM values of transcripts measured in an ADAR1 knockdown sample and a negative-control sample (NTC) are plotted. Edited and nonedited transcripts are indicated in different colors.(D) All transcripts.(E) Genes encoding zinc-finger proteins whose expression levels changed by ≥20%.See also Table S8.
Mentions: The results from ADAR knockdown and RNA-IP suggest that although ADARs mediate A-to-G editing, they do not mediate other types of RDDs. The levels of other types of differences were largely unaffected by ADAR knockdown, and the transcripts that showed those differences were not bound by ADAR. This prompted us to compare the genomic features surrounding the A-to-G editing sites and other types of RDDs. First, the sequence contexts of A-to-G and non-A-to-G sites are different. The base 5′ adjacent to the adenosine in A-to-G sites is depleted of guanosine (G) and the base 3′ to A-to-G editing sites is enriched for G (Figure 5A), consistent with previous reports (Lehmann and Bass, 2000). This sequence feature is specific to A-to-G editing because it is not present in random adenosines within nonedited Alu repeats (data not shown). This sequence motif was also not found for any of the RDDs. We identified sequence motifs for G-to-A and T-to-C sites, and they differed from the motif around the A-to-G sites (Figure 5A). Second, the A-to-G sites were more clustered than the non-A-to-G sites (67% of A-to-G sites were found within 25 nt of each other, compared with 14% of non-A-to-G RDDs). Third, most of the A-to-G sites were within or near inverted repeats, which form dsRNA and are preferentially recognized and bound by ADAR enzymes. Nearly 45% of the A-to-G sites resided within inverted repeats and another 30% were found near inverted repeats (<1 kb). In contrast, very few (0.9%) of the non-A-to-G sites were found in inverted repeats. Lastly, A-to-G sites and RDD sites were found in different regions of genes. A-to-G sites were found mostly in the 3′ UTRs, whereas RDDs were found mainly in the 5′ UTRs and in coding exons. Only 4% of the A-to-G sites (compared with 35% of RDDs) were in coding exons (Figure 5B). The differences between A-to-G editing sites and the other types of RDDs suggest that they are mediated by different mechanisms. Biochemically, this is expected since some of the RDDs are transversion events that cannot be explained simply by deamination.

Bottom Line: To study the role of ADAR proteins in RNA editing and gene regulation, we sequenced and compared the DNA and RNA of human B cells.The results uncovered over 60,000 A-to-G editing sites and several thousand genes whose expression levels are influenced by ADARs.Our results also reveal that ADAR regulates transcript stability and gene expression through interaction with HuR (ELAVL1).

View Article: PubMed Central - PubMed

Affiliation: Department of Genetics, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA. Electronic address: ixwang@umich.edu.

Show MeSH