Click here to close
Hello! We notice that you are using Internet Explorer, which is not supported by Xenbase and may cause the site to display incorrectly.
We suggest using a current version of Chrome,
FireFox, or Safari.
CHMP1 is a novel nuclear matrix protein affecting chromatin structure and cell-cycle progression.
Stauffer DR
,
Howard TL
,
Nyun T
,
Hollenberg SM
.
???displayArticle.abstract???
The Polycomb-group (PcG) is a diverse set of proteins required for maintenance of gene silencing during development. In a screen for conserved partners of the PcG protein Polycomblike (Pcl), we have identified a new protein, human CHMP1 (CHromatin Modifying Protein; CHarged Multivesicular body Protein), which is encoded by an alternative open reading frame in the PRSM1 gene and is conserved in both complex and simple eukaryotes. CHMP1 contains a predicted bipartite nuclear localization signal and distributes as distinct forms to the cytoplasm and the nuclear matrix in all cell lines tested. We have constructed a stable HEK293 cell line that inducibly overexpresses CHMP1 under ecdysone control. Overexpressed CHMP1 localizes to a punctate subnuclear pattern, encapsulating regions of nuclease-resistant, condensed chromatin. These novel structures are also frequently surrounded by increased histone H3 phosphorylation and acetylation. CHMP1 can recruit a PcG protein, BMI1, to these regions of condensed chromatin and can cooperate with co-expressed vertebrate Pcl in a Xenopus embryo PcG assay; this is consistent with a role in PcG function. In combination, these observations suggest that CHMP1 plays a role in stable gene silencing within the nucleus.
Fig. 1.
CHMP1 is a conserved eukaryotic protein. (A) Conceptual translation products CHMP1 and PRSM1 from alternative reading frames of the human PRSM1 mRNA. (B) Phylogenetic tree for CHMP1 and CHMP1.5 from human (H.s.), and CHMP1 from Drosophila melanogaster (D.m.), C. elegans (C.e.) and S. cerevisiae (S.c.). ClustalX was used to align proteins and calculate the distance between proteins for tree construction (Thompson et al., 1997). (C) All CHMP1 proteins from the tree are aligned. Amino acids that are identical in at least three proteins are highlighted.
Fig. 2.
Distinct forms of mammalian CHMP1 protein are found in cytoplasmic and nuclear compartments. (A) The subcellular distribution of CHMP1 was determined by western blot analysis with purified anti-CHMP1 antiserum after 18% SDS-PAGE. (B) Human cells lines derived from cervical (HeLa), prostate (LNCaP), breast (MCF7), colon (HT-29), and chorionic (JAR) carcinomas were compared with primary human umbilical vein endothelial cells (HUV-EC-C) for CHMP1 expression levels and subcellular distribution. (C) Expression levels of the nuclear and cytoplasmic forms of CHMP1 are compared in a mouse cell line (C2C12), mouse embryo fibroblasts and the indicated adult mouse tissues. Tissue extracts are unfractionated and represent total protein. Blots were stripped and re-probed with antibody to alpha-tubulin (cytoplasm-specific) and matrin 4 (nuclear matrix-specific) to demonstrate the accuracy of the fractionation. * indicates a nonspecific band. c, cytoplasm; ds, nuclear proteins released by DNaseI-digestion plus 800 mM NaCl-extraction; m, nuclear matrix solubilized by detergent; w, whole cell with complete detergent solubilization (see Materials and Methods).
Fig. 3.
Exogenous CHMP1 is tightly regulated and enters the nucleus. All experiments use IND-CHMP1 cells and exogenous CHMP1 induction with 10 μM muristerone. (A) Total RNA was harvested at the indicated times of induction and analyzed by northern blot for CHMP1, with CHO-B as an internal control. Endogenous CHMP1 RNA does not comigrate with exogenous and is not detected with this shorter exposure. (B) Cells were harvested at 0 hours and 24 hours of induction, fractionated and analyzed as in Fig. 2. CHMP1 species that comigrate with endogenous (closed arrows) or are novel (open arrow) are indicated. (C) Cells were induced for the indicated times, fixed and double labeled with anti-CHMP1 (red) and Hoechst 33258 (blue). Note the cytoplasmic fibers at 3 hours and 9 hours. (D) Cells were induced for 24 hours, including treatment with colcemid (200 ng/ml) for the final 10 hours to enrich for mitotic cells. Metaphase spreads were prepared (Chandler and Yunis, 1978) and stained for CHMP1 and DNA. The inset shows an enlarged region from the spread. c, cytoplasm; m, nuclear matrix solubilized by detergent; w, whole cell with complete detergent solubilization.
Fig. 4.
Exogenous CHMP1 blocks cell-cycle progression in late S-phase. IND-CHMP1 cells were induced with 10 μM muristerone for the indicated times. (A) Cells were scored for exogenous CHMP1 expression (triangles) and histone H3 phosphorylation (Ser10) with specific antibodies (see Methods). The phospho-H3 (P-H3) pattern was scored as punctate nuclear (open squares), or uniform nuclear and therefore mitotic (closed squares). DNA replication was measured by BrdU incorporation (asterisk) and apoptosis by TUNEL labeling (open circles). (B) The IND-CHMP1 line and its parental activator line (Invitrogen) were induced with muristerone for the indicated times, fixed, stained with propidium iodide and analyzed by flow sorting. The percentage of cells in each cycle phase is indicated in the corner of each panel.
Fig. 5.
Exogenous CHMP1 locally condenses chromatin and alters histone modification. IND-CHMP1 cells were induced with muristerone (24 hours mur) or left uninduced (0 hours mur). (A) Cells were fixed, then triple-labeled with anti-CHMP1, Hoechst 33258 and propidium iodide (PI) or (B) double-labeled for DNA and antibody to modified forms of histone H3 (Upstate Biologicals). Note the strong staining only in the mitotic cell at 0 hours. Ac-H3, acetylated (Lys9,14); P-H3, phosphorylated (Ser10).
Fig. 6.
CHMP1 subnuclear domains are nuclease-resistant. IND-CHMP1 cells were induced for 24 hours with muristerone, fixed and digested with the indicated amount of DNaseI for 1 hour at room temperature as described (Nickerson et al., 1997). Nuclei were double-labeled for CHMP1 or phosphorylated (Ser10) histone H3 (P-H3), and DNA (Hoechst 33258) and visualized by fluorescence microscopy.
Fig. 7.
CHMP1 subnuclear domains contain the PcG protein BMI1. Human U2OS cells were transfected with pIND-CHMP1 and the fusion trans-activator expression vector pVgRXR (Invitrogen) and induced with muristerone. After 48 hours cells were fixed and triple-stained for CHMP1 (green), BMI1 (red) and DNA (blue). Images were collected by laser scanning confocal microscopy. Note that the BMI1 signal is completely internal to CHMP1 (merge) and overlaps with condensed chromatin in the transfected cell.
Fig. 8.
CHMP1 produces a PcG phenotype upon misexpression in Xenopus. (A) Four-cell-stage embryos were unilaterally injected (arrowhead) with the indicated synthetic, capped RNA plus a marker RNA encoding β-galactosidase and processed to measure the side of injection (in situ β-galactosidase assay) and En-2 expression pattern (Yoshitake et al., 1999). (B) Embryos were injected with 0.1 ng of each indicated RNA. At stage 20 embryos were fixed and analyzed for En-2 expression by RNA in situ hybridization. The percentage of embryos with the indicated effects is the average of three independent experiments. Standard error is indicated. Control embryos show less than 3% repression and 0.5% shift.
