Three isoforms have been identified for Chd8.

<p>Chd8-S arises from passage of transcription through the end of exon 9 into intron 10, where it terminates. <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-RodriguezParedes1" target="_blank">[32]</a>, <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Rodenberg1" target="_blank">[34]</a> Chd8-S contains a single chromodomain (C1). Two longer isoforms contain the tandem chromodomains (C1 and C2) and helicase domain characteristic to Chd proteins <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Marfella1" target="_blank">[91]</a>, <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Hall1" target="_blank">[92]</a>. Chd8-L1 and Chd8-L2 result from two different start sites for transcription <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-RodriguezParedes1" target="_blank">[32]</a>, with the Chd8-L1 transcript encoding an amino terminus extension that encompasses a p53 binding domain <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Nishiyama2" target="_blank">[36]</a>. All three isoforms contain a series of five nuclear localization signals (NLS) <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Kobayashi1" target="_blank">[38]</a> and a β-catenin binding domain (β) <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Thompson1" target="_blank">[31]</a>, <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Sakamoto1" target="_blank">[37]</a>, <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Nishiyama3" target="_blank">[39]</a> that is also required for histone H1 <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Nishiyama1" target="_blank">[35]</a> binding <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Nishiyama1" target="_blank">[35]</a> and STAT3 <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Yamashina1" target="_blank">[28]</a> binding. Chd8-L1 and Chd8-L2 also contain a pair of BRK domains, which mediate chromatin interaction via the histones and is required for interaction with CTCF <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046316#pone.0046316-Ishihara1" target="_blank">[29]</a>. All three isoforms contain the AKAP domain (RII) characterized in this study.