RNA Editing by Mammalian ADARs - Advances in Genetics - page 111

particular alternative splicing. As described earlier, ADAR2 edits its own tran-
script within an intron creating an alternative splice acceptor site which leads to
a truncated protein lacking both dsRBDs and the deaminase domain (Rueter
et al.

, 1999). This generates a negative feedback loop to regulate the amount of
active ADAR2, and alternative splicing directly correlates with the activity of
the ADAR2 enzyme (Maas et al. , 2001).

Adar2 / mice have an elevated level of incorrectly processed The

GluR-B , 2000). Intron 11

contains the ECS region that is required for editing at the

transcripts that retain intron 11 (Higuchi et al.

site,
leading to the hypothesis that the processes of editing and splicing are linked. In
addition to editing at the Q/R site, there are “hotspot” regions within intron 11
at the

GluR-B Q/R

þ 60 site (hotspot 1) and the þ 262/263/264 sites (hotspot 2) that are also

edited (Higuchi et al. , 1993). Analysis of editing and splicing of this region of

GluR-B transcript in a cell culture assay revealed that editing at both the Q/R site

and intronic hotspot 2 is required for efficient splicing of intron 11 (Schoft et al. ,

2007). This may account for the observed 100% editing of mature GluR-B

transcript , as unedited transcript is not processed correctly. The intronic

hotspot 2 editing event could potentially alter a splice-repressor signal as muta-
tions which disrupted the duplex structure of this region did not increase splicing
efficiency (Schoft

in vivo

et al. , 2007).

editing site is 2 bp from the 5 0 splice donor site of the
mutually exclusive FLIP/FLOP exons. Unlike the Q/R site, editing at the R/G
site reduced the efficiency of splicing downstream of the editing event; however,
the decreased splicing efficiency correlated with an increase in splicing fidelity
(Schoft

The GluR-B R/G

, 2007). When an “uneditable” construct with a mutated ECS was
used, there was skipping of both downstream FLIP/FLOP exons and this gener-
ated a transcript containing a premature stop codon (Schoft

et al.

, 2007). These
results demonstrate that editing at the R/G site facilitates correct splicing of the
downstream FLIP/FLOP exons. Inhibition of splicing by ADAR2 was also ob-
served with

et al.

transcript containing the R/G
site, although the inhibition was partially alleviated by addition of RNA helicase
A indicating that inhibition of splicing was in part due to the duplex structure
(Bratt and Ohman, 2003). A construct containing an inosine residue at the R/G
site that mimics constitutive editing also showed inhibition of downstream
splicing. Therefore, the inhibitory effect is also due to the presence of inosine
rather than solely due to ADAR2 binding and this may be due to the unedited
sequence being closer to the consensus splice site (Schoft

in vitro splicing reactions on a GluR-B

, 2007). Also, the
edited R/G site shows greater similarity to the consensus binding site for hnRNP
A1 and this could act as a splicing silencer; however, further work is required to
prove this. In adult mouse brain, no correlation was found between editing and
choice of the mutually exclusive FLIP/FLOP exons so editing and alternative
splicing are not linked at this site (Schoft

et al.

et al. , 2007).