Sin3A recruits Tet1 to the PAH1 domain via a highly conserved Sin3-Interaction Domain

Abstract

The Sin3A complex acts as a transcriptional hub, integrating the function of diverse transcription factors with histone modifying enzymes, notably, histone deacetylases (HDAC) 1 and 2. The Sin3A protein sits at the centre of the complex, mediating multiple simultaneous protein-protein interactions via its four paired-amphipathic helix (PAH) domains (PAH1-4). The PAH domains contain a conserved four helical bundle, generating a hydrophobic cleft into which the single-helix of a Sin3-interaction domain (SID) is able to insert and bind with high affinity. Although they share a similar mode of interaction, the SIDs of different repressor proteins bind to only one of four potential PAH domains, due to the specific combination of hydrophobic residues at the interface. Here we report the identification of a highly conserved SID in the 5-methylcytosine dioxygenase, Tet1 (Tet1-SID), which interacts directly with the PAH1 domain of Sin3A. Using a combination of NMR spectroscopy and homology modelling we present a model of the PAH1/Tet1-SID complex, which binds in a Type-II orientation similar to Sap25. Mutagenesis of key residues show that the 11-amino acid Tet1-SID is necessary and sufficient for the interaction with Sin3A and is absolutely required for Tet1 to repress transcription in cells.

Figure 1

Identification of a conserved Sin3 Interaction Domain (SID) in Tet1 and Tet3. (A) Schematic diagram of Tet1, 2 and 3 showing the presence of a putative helical SID in Tet1/3, but not Tet2. (B) The putative Tet1-SID is highly conserved across multiple species; Hs- Homo sapiens, Mm – Mus musculus, Ne - Notamacropus eugenii (Wallaby), Ps - Pelodiscus sinensis (Chinese soft shell turtle), Dr - Danio rerio. (C) Co-immunoprecipitation followed by western blotting reveals the requirement of the Tet1-SID for association with exogenous and endogenous Sin3A. This image is cropped, the uncropped version of the blot is shown in Supplementary Fig. S5A.

Figure 2

The Tet1-SID binds to the PAH1 domain of Sin3A. (A) Schematic diagram showing the boundaries of Sin3A deletion constructs and relevant domains, PAH domains 1, 2, 3 and 4 are indicated by boxes. PAH4 is shaded in grey as it is likely non-functional. HID – HDAC interaction domain. Lower panel, GST-pulldown with the SID domains of Sap25 and Tet1 using ³⁵S-Met using the indicated Sin3A truncations. Short and long exposures of the gel were taken to visualize the binding of Tet1 to PAH1 as indicated. (B) GST-pulldown experiment using the indicated GST-fusion proteins with ³⁵S-Met labelled full-length Sin3A. (C) GST-pulldown with GST-Tet1 and full-length ³⁵S-Met labelled Sin3A. Mutations in individual PAH domains (P1–P4) to either Proline (PP) or Alanine (AA) are indicated. This image is cropped, the uncropped version of the gel is shown in Supplementary Fig. S5B. (D) GST-pulldown with GST-PAH domains (1–3) and Tet1 746–951 labelled with ³⁵S-Met.

Figure 3

The Tet1-SID forms an amphipathic helix with key hydrophobic residues required for interaction with Sin3A. (A) Helical wheel of the Tet1-SID residues 893–900 demonstrates hydrophilic and hydrophobic faces. (B) GST pulldown using GST-Tet1 878–911 wild-type and mutations, as indicated, with ³⁵S-Met labelled Sin3A. This image is cropped, the uncropped version of the gel is shown in Supplementary Fig. S5C. (C) Column fractionation of purified GB1-Tet1 and PAH1-His (mixed at a 1:1 ratio) demonstrates their ability to form a binary-complex in solution, which is dependent on the presence of key hydrophobic residues.

Figure 4

NMR and structural modelling of the PAH1:Tet1 complex. (A) Overlay of the ¹⁵N-HSQC spectra of apo-PAH1 (black) and wild-type PAH1:Tet1 complex (red). All NMR data was recorded at 303 K. (B) Results of minimal chemical shift mapping of various PAH1-Tet1 complexes. Value for the WT complex (coloured black) is computed relative to the apo state and establishes the magnitude of the effect. Values for mutant complexes are computed relative to the WT-complex and thus are a measure for the importance of the mutated residue for complex formation. (C) Model of the PAH1:Tet1 complex derived from the PAH1:Sap25-SID NMR ensemble (see methods). PAH1 is shown in surface representation; the Tet1 peptide in ribbon presentation with crucial residues as sticks: A893, I894 (green), L897 (red), T898, L900 (orange), and E902. The latter is potentially capable of an electrostatic interaction with PAH1-R120. (D) Close-ups of the PAH1:Tet1 complex; crucial residues shown and highlighted as in (C).

Figure 5

Tet1 L897 and L900 are key residues for interaction with Sin3A in cells. (A) Co-immunoprecipitation followed by western blotting reveals the requirement of L897 and L900 for the interaction with endogenous Sin3A in mouse ES cells. This image is cropped, the uncropped version of the blot is shown in Supplementary Fig. S5D. (B) Transcription of a luciferase reporter assay is driven by a simple TATA based promoter enhanced by addition of a LexA-VP16 fusion protein. Transcription can be repressed by co-expression of either the Mxd1-SID (MadN35), or Tet1-SID fused to a Gal4 DNA binding domain (DBD). Repression is dependent upon the interaction with Sin3A, since mutation of the Tet1-SID causes loss of repression.