Adenosine (A) to inosine (I) RNA editing is a widespread post-transcriptional mechanism in mammals that contributes to increase the protein diversity. Adenosine deaminases that act on RNA (ADARs) are the enzymes catalyzing RNA editing. ADARs are particularly active within the brain where they act on transcripts involved in neurotransmission. In this work the editing efficiency of all known site-selectively edited substrates have been analyzed during development of the mouse brain. We show that there is a global regulation of RNA editing, where editing levels of sites increase as the brain matures. This increase in editing efficiency cannot be explained by an increase in ADAR protein expression. During differentiation of primary cells from the mouse brain, editing levels increases similar to what we observe in vivo. Interestingly, the subcellular localization of the ADAR enzymes of cultured neurons show a different distribution in immature compared mature neurons. An accumulation of the ADAR enzymes in the nucleus may explain elevated A-to-I editing during brain development. Furthermore, we find that certain adenosines work as principal sites where editing of the transcript is initiated. Presumably, these sites are kinetically favored and are hypothesized to recruit the ADAR enzymes to the RNA substrate. Editing is then coupled to sites located in multiples of 12 nucleotides from each other. Interestingly, these sites reside on the same side in the 3D helix structure. The Gabra-3 transcript is site-selectively edited at a single position changing an isoleucine codon for a methionine upon editing. Gabra-3 encodes the a3 subunit of the GABA_A receptor. We show that receptors assembled with edited a3 are less stable at the cell surface than the non-edited a3. We propose that the amino acid change upon editing, could affect protein interactions important for trafficking and stability of the GABA_A receptors. Further, the editing event in a3 may have the function to reduce the number of a3 subunits in favor of other a subunits.

Adenosine (A) to inosine (I) RNA editing is a widespread post-transcriptional mechanism in eukaryotes that increases the protein diversity. Adenosine deaminases acting on RNA (ADARs) are the enzymes that catalyze this conversion. The diversity generated by ADAR enzymes occurs mainly in the brain where they target transcripts coding for proteins in the central nervous system (CNS).

We have determined the editing frequency of known ADAR substrates during development of the mouse brain using the large-scale 454-sequencing method. We show in paper I that editing is regulated during development of the brain, where it increases along with the maturation of the brain. We propose that the unedited isoform of proteins are required for the undeveloped brain while the edited isoforms are more suitable for the mature brain.

In paper II we show that substrates with multiple editing sites, one specific principle adenosine is favored for initial editing. We demonstrate that within these substrates, editing is coupled when adenosines are located in multiples of twelve nucleotides. These edited adenosines reside on the same side in the tertiary RNA helical structure. A model is suggested where kinetically favored structures at principle editing sites attract ADAR to the substrate, followed by editing at sites that are structurally adjacent to the initiation site.

Editing of the mammalian Gabra-3 transcripts coding for the GABA_A receptor α3 subunits recodes an isoleucine (I) to a methionine (M) referred as the I/M site. In paper III we demonstrate that receptors containing edited α3 subunits have altered trafficking properties compared to receptors containing unedited α3 subunits. We suggest that the amino acid residue change, affects protein interactions required for stability and trafficking of GABA_A receptors. We propose that the biological function of editing is to reduce the number of α3 subunits in favor of other α subunits.

The dsRNA structure at the I/M site in the Gabra-3 transcript is formed within the exon 9 sequence. We show in paper IV that a conserved intronic dsRNA structure in the downstream intron is required for editing to occur at the I/M site. We demonstrate that in the context of this intronic duplex also non-ADAR substrates can be edited. We propose that the intronic dsRNA stabilize the short I/M stem structure, thereby increasing the ability of ADAR to bind and edit the I/M site. These discoveries have expanded the knowledge in how ADAR editing is employed to supply the development of the brain as well as the RNA structure requirement for editing to occur.