|
FIGURE 1.
Expression of Gbx2 depends on NuRD recruitment. Alignment of the extreme N termini of mammalian Sall family members 1â4, and the Sal orthologs, Sem-4 (C. elegans) and XsalF (Xenopus), reveals the sequence conservation within the first 12 amino acids of the N terminus. The conserved residues are shown in boldface type. *, denotes the critical residues in Sall1 that are required for repression and NuRD recruitment (10). Although amino acid 12 is not conserved, it is always hydrophobic and is required for Sall1-mediated repression. An alternative splice form of Sall2 (Sall2Alt) encodes a different 5â² exon, resulting in a dissimilar N terminus. Sequences were obtained from the NCBI data base and have the following accession numbers: M. musculus, NP_056587 Sall1; C. elegans Sem-4, AABO3333; M. musculus Sall2Alt, AJ007396; and X. laevis XsalF, AA579483. 400 pg of capped RNA (XsalF or XsalFÎ12) was injected into two left animal blastomeres. Co-injection of β-galactosidase verified the site of injection, and embryos were harvested at the neurula stage (AâD). Whole mount in situ hybridization for Gbx2 revealed reduced expression in the XsalF-injected side (left) compared with the uninjected (right) side in 82% of embryos (B, 82%). Deletion of the SRM (XsalFÎ12) abolished this repression (D, 60%) or resulted in increased Gbx2 expression (C, 32%). 6% could not be scored. Control embryos were not injected (A, con). Arrows indicate the site of injection.
|
|
FIGURE 2.
Sall1 requires DNA binding sites, ATAAAA and ATAATT, for interaction with the Gbx2 locus. Top, Gbx2 wild-type and mutant probes used in EMSAs. The Gbx2 mutant probe contains point mutations in the Sall1 DNA binding sites which are underlined. Bottom, EMSA showing Sall1 binding to the Gbx2 probe. Lane 1 contained the designated wild-type Gbx2 probe alone, and lane 2 (mock) indicates the same probe with nuclear extract expressing empty Flu expression plasmid (pcDNA3) as a control. Lanes 3â7 indicate nuclear extract prepared from COS-1 cells transfected with full-length Flu-Sall1. Supershift assays were performed using affinity-purified antibodies that recognize different epitopes of Sall1 (lanes 4 and 5), and a cold competition assay was performed with 10-fold molar excess Gbx2 wild-type probe (lane 6) or 10-fold molar excess of Gbx2 mutant probe (lane 7).
|
|
FIGURE 3.
Sall1 and NuRD directly bind Gbx2 in vivo to mediate repression. ChIP assays were performed on P19 cells with anti-Sall1 (gray bars), anti-MBD3 (horizontal lines), or a nonspecific IgG control antibody (black bars). A, immunoprecipitated DNA was analyzed by qPCR amplification. ChIP analysis revealed Sall1 and MBD3 binding to a Gbx2 region at â0.7 kb; the â4.6 kb and â5.4 kb regions of Gbx2 and Gapdh served as negative controls. The qRT-PCR results represent ±S.D. derived from three independent experiments. B, COS-1 cells were transfected with a pGL3-promoter reporter containing a 960-bp promoter region of Gbx2 (â1620 bp to â660 bp) and the indicated Flu-epitope tagged Sall1 constructs consisting of wild-type Sall1, Sall2Sall1 chimera where the first 12 amino acids are altered, and a Sall1 construct (Î12) where the first 12 amino acids are deleted. COS-1 cells were also transfected with mutant pGL3-promoter reporters and Flu-Sall1. ATmut1 contains a mutation in the first AT recognition sequence (ATAAAA to ACAAAA), and ATmut2 contains a mutation in the second AT recognition sequence (ATAATT to ACAACC). The -fold repression was calculated by dividing the normalized luciferase activity of COS-1 cells expressing Flu alone by the activity of the Sall1 expression plasmids and is expressed as the mean ± S.D. for triplicate transfections of three independent experiments. Sall1 immunoblot reveals equal expression of each Sall1-flu fusion protein for each transfection condition.
|
|
FIGURE 4.
S2E phosphomimetic mutation of the Sall1 repression motif prevents repression and NuRD recruitment. In A: Left panel, COS-1 cells were transfected with the indicated GAL4DB-Sall1 fusion proteins and a reporter with luciferase under control of the SV40 promoter and 5 upstream GAL4 binding sites. The -fold repression was calculated as luciferase activity relative to GAL4DB alone and is expressed as the mean ± S.D. for triplicate transfections of three independent experiments. Right panel, COS-1 cells were also transfected with GST fusions corresponding to the GAL4 fusions of Sall1. GST pulldown of GST Sall1N-(2â136) or GST-Sall1N-(2â136) with S2A and S2E point mutations were analyzed for association with selected NuRD components by Western blot using antibodies specific to HDAC1, HDAC2, RbAp46, RbAp48, and Mi-2β. B, P19 cell extracts were immunoprecipitated (IP) with a control FLAG or monoclonal Sall1 antibody and separated by SDS-PAGE. The gel was then either transferred and probed with a polyclonal Sall1 or phosphospecific antibody or stained with Pro-Q Diamond Phosphoprotein in Gel Stain to reveal that Sall1 exists as a serine phosphoprotein in vivo.
|
|
FIGURE 5.
Phosphorylation of Sall1 at serine 2. A, COS-1 cells were transfected with GST-Sall1-(2â12) or GST-Sall1-(2â12) S2A or S2E point mutants. The cells were radiolabeled with [32P]orthophosphate for 4 h prior to harvesting the lysates. GST-Sall1 constructs were complexed to glutathione, separated by SDS-PAGE, and analyzed by autoradiography to localize phosphorylation to Ser-2 of the SRM. B, phosphopeptide mapping of Sall1. COS-1 cells were radiolabeled with [32P]orthophosphate after transient transfection with Flu-tagged Sall1-(2â1322) or Flu-tagged Sall1S2A-(2â1322). Purified 32P-labeled Sall1 was digested with trypsin, and the resulting phosphopeptides were separated in two dimensions: by electrophoresis and by ascending chromatography. The autoradiograms of the resulting phosphopeptide maps are shown. The left panel shows a phosphopeptide map of full-length Flu-Sall1-(2â1322) revealing five phosphopeptides (1â5). The right panel shows that point mutagenesis of Ser-2 to alanine eliminates the phosphorylation of peptides 1â3 confirming this site as a phosphorylation site of Sall1. This figure is representative of three independent experiments.
|
|
FIGURE 6.
Mutation of serine 2 of the SRM disrupts the incorporation of 32P. A, COS-1 cells were radiolabeled with [32P]orthophosphate after transient transfection with Flu-tagged Sall1-(2â1322) or Flu-tagged Sall1S2A-(2â1322). An anti-FLAG control immunoprecipitation was performed on a cell lysate expressing Flu-tagged Sall1-(2â1322). Purified 32P-labeled Sall1 and S2A were separated on 5% SDS-PAGE and analyzed by autoradiography. B, after the products were run on 5% SDS-PAGE and transferred to 3MM Whatman paper, the phosphorylated Sall1 and S2A bands were excised, and radioactivity was measured by using a scintillation counter. The specific activity of [γ-32P]ATP in the reaction mixtures was 2000 cpm/pmol ATP. The amounts of Sall1 protein and incorporation of 32P into Sall1 and S2A are included in the table.
|
|
FIGURE 7.
Sall1 is phosphorylated by PKC. A, P19 cell extracts were immunoprecipitated (IP) with a control FLAG or monoclonal Sall1 antibody followed by Western blotting with an antibody to PKC. B, purified Sall-Flu-(2â1322) and Sall1S2A-Flu-(2â1322) were expressed in COS-1 cells and immunopurified. An anti-FLAG control immunoprecipitation was performed on a cell lysate expressing Sall-Flu-(2â1322). The purified proteins were then incubated in the absence or presence of purified PKC for 5 min at 30 °C, followed by the addition of 1 μl of 1 mM dATP for 30 min at 30 °C. The products were then run on 5% SDS-PAGE and analyzed by autoradiogram.
|
|
FIGURE 8.
Phosphorylation of the SRM disrupts Sall1-mediated repression of Gbx2 in vivo. A, the serine point mutants, S2A and S2E, were created in the context of XsalF and expressed as capped RNA for injection into eight-cell Xenopus embryos. The injections and in situ hybridization experiments were performed exactly as described in the legend for Fig. 1. The microinjection manipulations revealed that the S2A mutant does not disrupt XsalF-mediated repression, revealing either a decrease in Gbx2 expression (A, panel II, 53.85%, n = 52) or equal expression (data not shown). In contrast, the phosphomimetic mutant, S2E, reveals a result similar to XsalFÎ12 where disruption of NuRD recruitment results in a failure to suppress Gbx2 expression. The S2E embryos showed no change (A, panel IV, 59.18%, n = 49) or unilateral expansion (A, panel III, 36.73%, n = 49) in Gbx2 expression on the injected side. Control embryos were not injected (A, panel I, con). Arrows indicate site of injection. B, COS-1 cells were transfected with the indicated Flu-Sall1 fusion proteins and the Gbx2 (â1620 bp to â660 bp) luciferase reporter. The -fold repression was calculated as luciferase activity relative to Flu alone and is expressed as the mean ± S.D. for triplicate transfections of three independent experiments. C, table showing total number of injected Xenopus embryos for each mRNA and the effects on Gbx2 expression. The representative images are displayed in Figs. 1 and 8. Wild-type XsalF and S2A were both capable of repressing Gbx2 expression. In contrast, XsalFÎ12- and S2E-injected embryos showed a loss of repression or increased levels of Gbx2 expression.
|
