The interaction of rDaxx with E2A is selective, direct, and occurs in vivo.
(A) Yeast two-hybrid analysis of the interaction between AD-rDaxx (613–730), a C-terminal fragment of rDaxx fused to a transcription activation domain, and a LexA DNA binding cassette or transcription factors E12, E47, Id3, Oct1, and Max1 fused to LexA. AD-rDaxx (613–730) expression was controlled by a galactose-inducible promoter, and the interactions were assessed on glucose and galactose-containing media. Dark hatches (assayed in quadruplicate) indicate β-galactosidase activity after 48 h of incubation.
(B) Binding of [³⁵S]-labeled proteins to GST fusion constructs retained by glutathione-Sepharose beads. Left, [³⁵S]methionine-labeled rDaxx was incubated with GST or GST-E12(477–654). Right, [³⁵S]methionine-labeled E47 was incubated with GST, GST-E12, or the indicated GST-rDaxx deletion constructs. Retained proteins were resolved by 15% SDS-PAGE, with equivalency of loading confirmed by Coomassie Blue staining (not shown). Binding activity was quantified by phosphor imaging.
(C) Nuclear protein extracts from C2C12 myoblasts (left), NIH 3T3 fibroblasts, and U266 B-lymphoblastoid cells (right), were analyzed by immunoprecipitation and immunoblotting. Lane 1 (immunoblotting for Daxx protein), 10 % of input used for lanes 2 and 3; lanes 2–5, immunoprecipitation using control antiserum (lane 3) or anti-E47 antiserum (lanes 2, 4, 5), followed by immunoblotting for Daxx.

Daxx inhibits transcription through a histone deacetylase-dependent mechanism.
(A) Nuclear extracts from the stable 3T3 lines were tested in an electromobility shift assay. Arrow indicates specific E-box binding activity; ns, a non-specific band. Lane 1, free μE5 probe; lanes 2–6, nuclear extracts from: lane 2, 3T3; lane 3, 3T3-GFP-RV; lane 4, 3T3-rDaxx-RV; lane 5, 3T3-rDaxxC-RV; lane 6, 3T3-GFP-RV, with 100-fold excess of cold E-box competitor.
(B) Transient transfection assays of μE-luc reporter activity in NIH 3T3 cells. Open bar, with E12 alone; solid bars, with E12 and rDaxx expression plasmids. Trichostatin A (TSA, 100 nM) or sodium butyrate (SB, 5 mM) were added 4 h after transfection, and cells were harvested after 24 h.

Decreased Daxx expression potentiates muscle differentiation.
(A), Left, Immunofluorescent micrographs of C2C12 cells transfected with expression plasmids pBB14 (encoding US9-GFP to identify transfected cells) and either pSupressorRetro (vector control, top) or pDaxx siRNA −8 (bottom). DAPI (Nuc), US9-GFP (GFP), and Daxx signals are indicated. Right, Immunoblot (20 ug total protein per lane) for Daxx expression in retroviral vector and pDaxx siRNA-transduced cultures. Alpha-tubulin (α-tub) is shown as a loading control. (B), Myotube count after 3 days of differentiation in cells transduced with vector (open bar) or pDaxx siRNA (solid bars) retroviruses. *, p <.0025. 6 randomly selected high power fields (hpf) were counted in each of 3 independent experiments.

Daxx represses E2A-dependent transcription.
(A) NIH 3T3 fibroblasts were transfected with E47 expression plasmid (0, 3, 10, or 30 ng), μE-luc (E-box) reporter construct (200 ng), and rDaxx, rDaxxC (aa 619–730), or control expression plasmids (400 ng). Bars indicate the standard error. *, p ≤ .0001 vs. respective control. Maximal reporter activity in the absence of rDaxx was set to 100.
(B) Left, NIH 3T3 fibroblasts were transfected with an E47 expression construct (10 ng), E-box reporter construct (200 ng), and a range of rDaxx or rDaxxC expression constructs (0, 3, 10, 30, 100, 200, or 400 ng). Activity with E47 alone was set to 100. Activities for rDaxx concentrations of 3 ng or greater were all significantly different (p < .0005) from E47 alone. Right, Stable NIH 3T3 lines generated with control (open bar), full-length rDaxx (black), or C-terminal fragment rDaccC (gray) retroviruses, were transfected with E-box reporter (300 ng) and E12 expression plasmid (10 ng) and subjected to luciferase assays.
(C) NIH 3T3 fibroblasts were transfected as indicated with expression plasmids for E12 and MyoD (100 ng each), with or without rDaxx (200 ng), along with the −3300 MCK-luc reporter (200 ng). Maximal reporter activity was set to 100. *, p ≤ .0003 vs. activity without rDaxx.

Increased Daxx expression opposes myogenic differentiation and muscle gene expression.
(A–D) C2C12 cells were transduced with GFP-RV (control) or rDaxx-derived retroviruses (Daxx, DaxxC, and DaxxΔ613–654) and differentiated in culture. Bar, 200 μm. (A), Left, Phase contrast images after 4 days of differentiation. Right, Myotubes after 3 days of differentiation were counted in 6 randomly-selected high power fields (hpf) per condition. Open bars, control; solid bars, transduced with Daxx retrovirus. *, p < .0004, **, p < .0001. (B) Immunofluorescent micrographs after 4 days of differentiation. Upper panels, nuclei stained with DAPI (blue); lower panels, same fields stained for myosin heavy chain (red). (C) Muscle gene expression. Total RNA from control GFP-RV or Daxx-RV (+) cells was harvested at the timepoints indicated and subjected to Northern analysis (10 μg per lane). The MCK probe identified a specific MCK band in day 2 and day 4 samples and a slightly smaller non-specific band throughout the time course. GAPDH was evaluated to assess equivalency of loading. (D) Daxx inhibits activation of the MCK promoter in C2C12 cells. GFP-RV (control, left) and Daxx-transduced (right) C2C12 cells were transfected with the MCK-luc reporter (200 ng) and pcDNA3 vector (500 ng, black bars) or MyoD and E12 expression constructs (250 ng each, gray bars). Luciferase activity was determined after 1 or 4 days of differentiation. (E) ChIP analysis of Daxx interaction with MCK and MEF2C regulatory elements during differentiation. C2C12 myoblasts were exposed to 10% horse serum for the indicated number of days prior to crosslinking, chromatin extraction, and immunoprecipitation. PCR was performed to identify the indicated promoter/enhancer sequences. Controls shown include no antibody (no Ab), chromatin input (1%) prior to immunoprecipitation (In), and immunoprecipitation for histone H3 (IP H3). PCR for GAPDH promoter sequences is shown as a reference.