Figure 1.
TEL binds RAR and RXR in vitro and in vivo.
(A) TEL and MN1-TEL bind RAR and RXR whereas MN1 does not. In vitro generated, 35S-labeled proteins, as shown in the input lanes, were used for GST pulldowns with GST-RAR, GST-RXR and GST (as control). (B) TEL binds the DBD domain of RARα. Upper panel: Schematic overview of the RARα and RXRα constructs. The domains within RARα and RXRα are indicated with letters commenly used to indicate the domains within the nuclear receptor family. At the bottom the different regions are indicated with respect to their function. AF1, Activator function 1; DBD, DNA binding domain; CTE, C-terminal extension; LBD, ligand-binding domain; AF2, activator function 2. Lower panel: GST pulldowns with in vitro transcribed/translated TEL and TIF2. Binding of TEL to RAR and RXR differs from the binding of TIF2 to RARα and RXRα. Binding of TIF2 was detected for LB-containing RARα constructs whereas TEL bound to the DBD region of RARα. No binding of TEL to the AF1 and first part of DBD was detected (GST-RAR-N2). (C) HLH and DBD domains of TEL are crucial for binding to RARα. Upper panel: Schematic overview of deletion constructs of TEL. Numbers indicate the amino acid positions. Middle left: ITT products were used for GST pulldown assays with GST-RAR-N3. Middle right: Only deletion constructs that contain both HLH and DBD domains were detected in the in vitro binding assay. Lower panel: WT TEL showed binding to GST-RARα, whereas HLH, DBD or double mutant TEL proteins failed to bind GST-RARα. (D) bio-precipitation of bioV5-tagged RAR-N3 proteins. Only in the presence of bioV5-tagged RARN3, TEL-HA proteins were detectable in the precipitations, showing the in vivo binding between TEL and RAR. Right panel shows input lysates. TEL proteins were detectable with both a-TEL and a-HA antibodies.
Figure 2.
TEL represses transcription that is mediated by the heterodimer RAR/RXR.
(A) Schematic overview of different luciferase reporter constructs RARE-luc, MSV-luc and RARE3MSV-luc. (B) The M43-TEL isoform inhibits the RARE luc reporter whereas the M1-TEL isoform does not. (C) The viral promoter MSV is repressed by both isoforms of TEL. M43-TEL is more efficient. (D) Both TEL isoforms repress the RARE3 MSV-luc construct. M43-TEL is more efficient. All transfections were performed in the presence of all-trans retinoic acid (ATRA).
Figure 3.
Deletion mutants of TEL have no repressive activity.
(A) Three different reporters were tested with TEL deletion mutants. The repression of TEL on the RAR-RXR dimer (i.e. RARE luc and RARE3-MSV luc) was only detected using full length M43 TEL. Although deletion mutant HD was able to bind RAR in vitro (figure 1), it was unsuccessful in repression. The MSV-luc reporter contains not only an RAR/RXR responsive element but also many possible ETS binding sites (figure 2E). The HD deletion construct of TEL repressed the MSV-luc reporter efficiently. (B) Both modes of repression by TEL (i.e. on RAR/RXR and on ETS sites) are disabled in the HLH mutant of TEL (TEL-V112A/L113A).
Figure 4.
Repression of TEL on RAR/RXR-mediated transcription does not involve histone deacetylase activity.
Increasing amounts of TSA were tested on RARE luc reporter in the presence of ATRA, and with and without M1-TEL, M43-TEL or MN1-TEL (120 ng). RAR/RXR-mediated transcription was stimulated by TSA to high levels, whereas M43-TEL prevents this. Intermediate effects were observed with MN1-TEL and M1-TEL.