Codanin-1, the protein encoded by the gene mutated in congenital dyserythropoietic anemia type I (CDAN1), is cell cycle-regulated

Abstract

Background: Congenital dyserythropoietic anemia type I is an inherited autosomal recessive macrocytic anemia associated with ineffective erythropoiesis and the development of secondary hemochromatosis. Distinct erythroid precursors with internuclear chromatin bridges and spongy heterochromatin are pathognomonic for the disease. The mutated gene (CDAN1) encodes a ubiquitously expressed protein of unknown function, codanin-1. Based on the morphological features of congenital dyserythropoietic anemia type I erythroblasts and data on a role in cell cycle progression of codanin-1 homolog in Drosophila we investigated the cellular localization and possible involvement of codanin-1 during the cell cycle.

Design and methods: Codanin-1 localization was studied by immunofluorescence and immune electron microscopy. Cell cycle expression of codanin-1 was evaluated using synchronized HeLa cells. E2F proteins are the main regulator of G(1)/S transition. An E2F1-inducible cell line (U20S-ER-E2F1) enabled us to study codanin-1 expression following ectopic E2F1 induction. Direct binding of E2F1 to codanin-1 promoter was assessed by chromatin immunoprecipitation. We used a luciferase-reporter plasmid to study activation of CDAN1 transcription by E2F1.

Results: We localized codanin-1 to heterochromatin in interphase cells. During the cell cycle, high levels of codanin-1 were observed in the S phase. At mitosis, codanin-1 underwent phosphorylation, which coincided with its exclusion from condensed chromosomes. The proximal CDAN1 gene promoter region, containing five putative E2F binding sites, was found to be a direct target of E2F1.

Conclusions: Taken together, these data suggest that codanin-1 is a cell cycle-regulated protein active in the S phase. The exact role of codanin-1 during the S phase remains to be determined. Nevertheless this represents the first step towards understanding the function of the proteins involved in congenital dyserythropoietic anemia.

Figure 1.

Codanin-1 is a nuclear chromatin-bound protein. (A) Western blots using anti-codanin-1 against whole cell extracts of asynchronous K562, U2OS, and HeLa cells detected a 130-kDa band. (B) Nuclear localization of codanin-1 in HeLa cells; sparing of the nucleolus is evident. (1) 4′-6′-Diamidino-2-phenylindole (DAPI) DNA stain. (2) Anti-codanin-1 antibody (1:150), followed by secondary antibodies (Cy3-conjugated donkey anti-rabbit). (3) Differential interference contrast image of the cells. The pictures were taken using an Olympus Fluoview-FV500 confocal laser scanning microscope. The objective was UPLAPO60x with N.A. 0.85. (C) Biochemical sub-nuclear fractionation of HeLa cell extracts. WCE -whole cell extract. TOPO II-topoisomerase II. Fractions were immunoblotted with the antibodies indicated. Anti-TOPOII (an abundant nuclear matrix protein) was used as a control for the fractionation. Codanin-1 is present mainly in the chromatin fraction. (D) Immunogold ultrastructural localization of codanin-1 to heterochromatin. (D1) CDA I erythroblast showing the typical spongy appearance of the nuclear heterochromatin and a dilated membrane pore (arrow). (x7000). (D2) Enlarged portion of the nucleus of 1, showing localization of gold particles in the nuclear heterochromatin, but not in the euchromatin (x45000). (D3) An erythroblast from a case of autoimmune hemolytic anemia, without dyserythropoietic features, showing gold particles in the nuclear area (x10000). (D4) Enlarged portion of the nucleus in 3 showing preferential localization of gold particles in the nuclear heterochromatin (x45000). (D5) A typical U2OS cell showing a nucleus composed mainly of euchromatin. A very small amount of heterochromatin can be seen adjacent to the nuclear membrane (x4500). (D6,7) Two enlarged regions of the nucleus of 5 showing preferential localization of gold particles in two small areas of heterochromatin adjacent to the nuclear membrane (x45000). Cells were examined with a Philips 201 electron microscope.

Figure 2.

Cell cycle-dependent localization of codanin-1. Asynchronized HeLa cells were fixed and co-stained with purified rabbit anti-codanin-1 and mouse anti-α-tubulin. DNA was stained with DAPI. While codanin-1 associates with DNA during inter-phase (top panel), it is clearly excluded from the mitotic condensing chromosomes (prophase to telophase). Images were photographed using a 60× objective on an Olympus Fluoview-FV500 confocal fluorescence microscope.

Figure 3.

Codanin-1 expression and phosphorylation are cell cycle-regulated. (A) Lysates of HeLa cells released from double thymidine block were prepared at various times and immunoblotted with anti-codanin-1 antibody. β-actin was used as a loading control. Cell cycle distribution was analyzed using FACScan analysis of propidium iodide-stained cells. Codanin-1 concentration increases during the S phase and declines thereafter during G2/M. (B) Codanin-1 is phosphorylated during mitosis. (a) Lysates prepared from interphase (I) or nocodazole-arrested (mitotic, M) HeLa cells were resolved on 8% gel and probed with anti-codanin-1. (b) Mitotic HeLa cell extracts were incubated (30 min, 30 °C) in the absence (lane 1) or presence (lane 2) of λ phosphatase as indicated. Lane 3 was loaded with mitotic extract treated with phosphatase in the presence of phosphatase inhibitors. Immunoblotting was performed with anti-codanin-1.

Figure 4.

Overexpressed E2F1 is associated with high levels of codanin-1 in the S phase followed by a sharp decline. (A) U2OS cells expressing E2F1 fused to estrogen receptor (U2OS-ER-E2F1) were treated with 300 nM 4-hydroxytamoxi-fan (OHT) for the times indicated, to induce E2F1 activation. Lysates were prepared and immunoblotted with anti-codanin-1 antibody. β-actin was used as a loading control. Codanin-1 levels rise 8-12 h following E2F1 activation and sharply decline thereafter. (B) Parental U2OS cells were treated and analyzed as described in A. No change in codanin-1 levels was observed. ( C ) mRNA levels rise with E2F1 induction; U2OS-ER-E2F1 cells were treated with 300 nM OHT for the times indicated. RNA was isolated using Trizol and used for reverse transcriptase PCR. CDAN1 and GAPDH were amplified by PCR and the products were separated on a 1% agarose gel. Relative levels of products indicated below the gel were calculated with the Image J program. (D) Codanin-1 is degraded by the proteasome. U2OS-ER-E2F1 cells were treated with 300 nM OHT for the times indicated. The proteasome inhibitor MG-132 was added to the cells at the time point indicated. Lysates were prepared and immunoblotted with anti-codanin-1 antibody. Tubulin was used as a loading control. MG-132 inhibited a reduction in codanin-1 levels observed after 12 h.

Figure 5.

Codanin-1 is a direct transcriptional target of E2F1. (A) Reverse transcriptase-PCR analysis of CDANI and GAPDH mRNA levels: U2OS-ER-E2F1 cells treated with OHT for 8 h (+) or not treated (−) in the presence or absence of cycloheximide (CHX). (B) Chromatin immunoprecipitation analysis was performed using growing U2OS cells: cross-linked chromatin was immunoprecipated with antibodies to E2F1 and HA, and then CDANI and GAPDH promoter fragments were amplified by PCR. Negative (no DNA) and positive (input DNA representing 0.2% of the total chromatin) control amplifications are shown.

Figure 6.

Codanin-1 promoter is sensitive to E2F1 (A) Schematic representation of the codanin-1 promoter-containing luciferase reporter plasmids. (B) U2OS cells were co-transfected with the empty pGL4 or the codanin-1 promoter constructs described in A and either wild-type E2F1 or a mutated form of E2F1 (E132) that cannot bind to DNA. Cells were also transfected with the pRenilla reporter plasmid to evaluate transfection efficiency. Values obtained for luciferase activity by luminometer were divided by the values obtained for Renilla activity. Values obtained for the pGL4 transfected cells were set at 0 and other samples were calculated as fractions of that value to determine the relative activity. Results shown represent the average of three independent experiments.