Histone deacetylase 2 (HDAC2) protein-dependent deacetylation of mortality factor 4-like 1 (MORF4L1) protein enhances its homodimerization

Abstract

Histone acetyltransferase mortality factor 4-like 1 (MORF4L1) is a relatively new histone acetyltransferase component that exists as a homodimer to exert its epigenetic function. The mechanism of MORF4L1 self-assembly is unknown. Here we report that Lys-148 deacetylation is indispensable for facilitating MORF4L1 self-assembly into a homodimeric unit. Among a stretch of ∼10 amino acids in the NH2 terminus between the chromodomain and MORF4-related gene (MRG) domain within MORF4L1, Lys-148 is normally acetylated. Substitution of Lys-148 with arginine augments MORF4L1 self-assembly. However, acetylation mimics of MORF4L1, including K148L and K148Q, abolished its self-assembly of the histone acetyltransferase component. HDAC2, a deacetylase, interacts with and keeps MORF4L1 in a deacetylation status at Lys(148) that triggers MORF4L1 self-assembly. Knockdown of HDAC2 reduces MORF4L1 self-assembly. HDAC2-dependent deacetylation of MORF4L1 enhances MORF4L1 homodimerization, thus facilitating the functionality of complex formation to repress cell proliferation.

Keywords: Epigenetics; Histone Acetylase; Histone Deacetylase; Posttranslational Modification; Proliferation; Protein Complexes; Protein Structure; Site-directed Mutagenesis.

FIGURE 1.

MORF4L1 is a homodimer. A, MLE cells (6 × 100 mm dish) were collected, lysed, and then cellular proteins were applied to a size exclusion spin column. The fractions were collected and separated by SDS-PAGE followed by immunoblotting analysis. B, V5- and Myc-tagged MORF4L1 were coexpressed in cells. Cell lysates were subjected to V5 immunoprecipitation (IP). The precipitates were analyzed by immunoblotting as indicated. C, V5- and Myc-tagged MORF4L1 were synthesized in vitro. The recombinant proteins were mixed with V5 antibody-conjugated agarose beads overnight. The washed beads were analyzed by immunoblotting as indicated. Input was 10% of the recombinant protein used in the pulldown (PD) assays. These data are representative of three separate experiments.

FIGURE 2.

Deletion of the NH₂ terminus of MORF4L1 ablates its homodimerization. A, MORF4L1 cysteine residue mutants tagged with V5 were coexpressed with Myc-tagged, wild-type MORF4L1 in the cells. The cell lysates were used for immunoprecipitation (IP) studies as indicated. Input was 10% of the cell lysate used in the studies. B, schematic presentation of the deletion mutations. C, NH₂-terminal deletion mutants were expressed in MLE cells. The cell lysates were subjected to V5 immunoprecipitation, and the immunoprecipitates were analyzed by immunoblotting as indicated. NH₂-terminal deletional mutant expression in the cells was checked by V5 immunoblotting. D, purified NH₂-terminal truncated MORF4L1 forms a dimer efficiently. NH₂-terminal truncated MORF4L1 was expressed and purified. Elution products were separated by electrophoresis, and the proteins were visualized by Coomassie Blue stain. These data are representative of three separate experiments.

FIGURE 3.

Reversible MORF4L1 acetylation at Lys-148 modulates its dimerization. A and B, lysine mutants and myc-tagged, wild-type MORF4L1 were coexpressed in cells. Cell lysates were subjected to myc immunoprecipitation (IP), and the precipitates were analyzed by V5 and myc immunoblot analysis. The cell lysates were analyzed with V5 immunoblotting to verify MORF4L1 mutant expression. The densitometry results of A were plotted in B. C, K148R and the acetylation mimic K148L and K148Q mutants were coexpressed with myc-tagged, wild-type MORF4L1 in the cells. The cell lysates were used for myc immunoprecipitation, followed by immunoblotting analysis as indicated. MORF4L1 mutant expression levels were analyzed by V5 immunoblotting. D, wild-type of MORF4L1 tagged with V5 or myc were expressed in cells for 40 h, and cells were treated with anacardic acid (final concentration of 50 μm) or trichastatin A (final concentration of 100 nm) for 6 h. Cell lysates were subjected to myc immunoprecipitation and followed by immunoblot analysis as indicated. Tagged MORF4L1 expression was verified by immunoblotting. These data are representative of three separate experiments.

FIGURE 4.

HDAC2 interacts with MORF4L1. A, cell lysates were used for MORF4L1 immunoprecipitation (IP). The precipitates were analyzed by HDAC2 immunoblotting. B, reciprocally, cell lysates were immunoprecipitated with HDAC2 antibody, and the precipitates were immunoblotted using MORF4L1 antibody. These data are representative of three separate experiments.

FIGURE 5.

HDAC2 is necessary for MORF4L1 self-assembly. A, HDAC2 was depleted in cells by introducing shRNA retroviral constructs for 48 h. Cell lysates were analyzed with HDAC2 immunoblotting. A scrambled retroviral plasmid was used as a shRNA specificity control. β-actin was used as a loading control. B, HDAC2 was first knocked down by shRNA for 24 h in the cells, and then V5- or Myc-tagged MORF4L1 were expressed for another 24 h. Cell lysates were analyzed by V5 immunoprecipitation (IP), followed by immunoblotting. C, V5- and Myc-tagged MORF4L1 and HDAC2 were synthesized in vitro and the MORF4L1 recombinant proteins were mixed with V5 antibody-conjugated agarose beads in the presence or absence of HDAC2 overnight. The washed beads were analyzed by immunoblotting as indicated. Input was 10% of the recombinant protein used in the pulldown (PD) assays. These data are representative of three separate experiments.

FIGURE 6.

HDAC2-dependent deacetylation of MORF4L1 at Lys-148. A, wild-type or a variety of lysine-mutant MORF4L1 plasmids were expressed in the cells. Cell lysates were subjected to V5 immunoprecipitation (IP), and the precipitates were analyzed by acetyl lysine immunoblotting. B, HDAC2 was depleted by introducing shRNA retroviral constructs into the cells for 24 h. Wild-type or the K148R mutant MORF4L1 plasmid was expressed in cells for another 24 h. Cell lysates were analyzed with V5 immunoprecipitation, followed by immunoblot analysis. A scrambled retroviral plasmid was used as a shRNA specificity control. C, wild-type, K148R, or acetylation mimic mutant MORF4L1 plasmids were expressed in cells. The cell lysates were fractionated by size exclusion spin columns, and all fractions were analyzed by immunoblotting. These data are representative of three separate experiments.

FIGURE 7.

Dimerization augments MORF4L1 complex formation and cellular viability. A, HDAC2 was depleted in cells by introducing shRNA retroviral constructs for 48 h. Cell lysates were immunoprecipitated (IP) with PF1 antibody, and the precipitates were analyzed by immunoblotting. A scrambled retroviral plasmid was used as a shRNA specificity control. B, wild-type and K148R mutant MORF4L1 plasmids were coexpressed in cells for 48 h. Cell lysates were used for PF1 immunoprecipitation, followed by immunoblot analysis. Wild-type and K148R mutant MORF4L1 plasmids in the cells were validated by V5 immunoblotting. C, wild-type and K135R, K148R, and K150R mutant MORF4L1 plasmids were expressed in the cells for 48 h. Viable cells were counted in each transfected cell line. The expression levels of the proteins were analyzed by immunoblotting. The data are representative of three separate experiments.