PROSITE documentation PDOC51569

DOT1 domain profile

The Dot1 protein (Dot1p) is an histone-lysine N-methyltransferase (EC 2.1.1.43) [E1] that methylates lysine 79 (Lys-79) of histone H3. It was first identified as a Disruptor Of Telomeric silencing in yeast where Dot1p is implicated in gene silencing and localization of the Silent Information Regulator (SIR) complex; in higher eukaryotes the methylation carried out by this enzyme may be used for differentiating chromatin domains. Unlike other histone-lysine methyltransferases (HKMTs), Dot1p displays a Rossmann-like (Class I) S-adenosyl-L-methyionine (SAM)-dependent MT fold (see <PDOC51555>) while other HKMTs contain the SET domain (see <PDOC50280>) and hence belong to a whole different structural class (see <PDOC51565>) [1,2].

Whereas most HKMTs, such as Suvar3-9 methylate Lys on the N-terminal tails of histones that stick out from the nucleosome, Dot1p substrate (Lys-79 of histone H3) is located in the conserved histone core, in a short turn connecting the first and second helices, exposed on the nucleosome disk surface [1,2]. In order for Lys-79 of H3 to be methylated by Dot1p, another lysine, Lys-123 of histone H2B, needs to be ubiquitinated. A possible reason put forward for this requirement is that the ubiquitination may create a space between adjacent nucleosomes, permitting access of Dot1p to its substrate [1,2]. In yeast, different states of methylation on Lys-79 of histone H3 (unmodified, mono-, di- and trimethylated) co-exist at the same time, but no clear function is associated with these different methylation states [1].

The strucure of the evolutionary conserved core of Dot1p, the DOT1 domain, has first been described for the yeast Dot1p in complex with S-adenosyl-L-homocysteine (AdoHcy) (see <PDB:1U2Z>) and then for the human Dot1-like protein (Dot1Lp) in complex with SAM (see <PDB:1NW3>). The DOT1 domain is about 300-350 amino acids long and is usually located at either of the extremities of the protein sequence: it stands at the C-terminus of the yeast Dot1p and at the N-terminus of the human Dot1Lp [1,2]. DOT1 displays a rather elongated structure and can be subdivided into two parts: the N- and the C-terminal subdomains [2]. The N-terminal part is made up of five α helices and two pairs of short β strand hairpins. The C-terminal part displays a Rossmann-like fold (see <PDOC51555>): it consists in a seven-stranded β sheet tucked by five α helices (three helices on one side of the sheet and two on the other), the sheet contains a central topological switchpoint resulting in a deep pocket where SAM is bound. The two subdomains are linked covalently by a loop. Altogether the SAM binding pocket is formed by five segments of the DOT1 domain of which four are located in the C-terminal substructure of the DOT1 domain and one in the loop connecting both parts; two of these segments are conserved across different Class I SAM-dependent MTs (see <PDOC51555>) [2].

The profile we developed covers the entire DOT1 domain.