pubmed.ncbi.nlm.nih.gov

Identification of the in vivo casein kinase II phosphorylation site within the homeodomain of the cardiac tisue-specifying homeobox gene product Csx/Nkx2.5 - PubMed

Identification of the in vivo casein kinase II phosphorylation site within the homeodomain of the cardiac tisue-specifying homeobox gene product Csx/Nkx2.5

H Kasahara et al. Mol Cell Biol. 1999 Jan.

Abstract

Csx/Nkx2.5, a member of the homeodomain-containing transcription factors, serves critical developmental functions in heart formation in vertebrates and nonvertebrates. In this study the putative nuclear localization signal (NLS) of Csx/Nkx2.5 was identified by site-directed mutagenesis to the amino terminus of the homeodomain, which is conserved in almost all homeodomain proteins. When the putative NLS of Csx/Nkx2.5 was mutated a significant amount of the cytoplasmically localized Csx/Nkx2.5 was unphosphorylated, in contrast to the nuclearly localized Csx/Nkx2.5, which is serine- and threonine-phosphorylated, suggesting that Csx/Nkx2.5 phosphorylation is regulated, at least in part, by intracellular localization. Tryptic phosphopeptide mapping indicated that Csx/Nkx2.5 has at least five phosphorylation sites. Using in-gel kinase assays, we detected a Csx/Nkx2.5 kinase whose molecular mass is approximately 40 kDa in both cytoplasmic and nuclear extracts. Mutational analysis and in vitro kinase assays suggested that this 40-kDa Csx/Nkx2.5 kinase is a catalytic subunit of casein kinase II (CKII) that phosphorylates the serine residue between the first and second helix of the homeodomain. This CKII site is phosphorylated in vivo. CKII-dependent phosphorylation of the homeodomain increased Csx/Nkx2. 5 DNA binding. Serine-to-alanine mutation at the CKII phosphorylation site reduced transcriptional activity when the carboxyl-terminal repressor domain was deleted. Although the precise biological function of Csx/Nkx2.5 phosphorylation by CKII remains to be determined, it may play an important role, as this CKII phosphorylation site within the homeodomain is fully conserved in all known members of the NK2 family of the homeobox genes.

PubMed Disclaimer

Figures

FIG. 1
FIG. 1

NLS of Csx/Nkx2.5. (A) Schematic of NLS of wild-type (WT) Csx/Nkx2.5 and mutant Csx/Nkx2.5 at the amino terminus of the homeodomain (HD). WT Csx/Nkx2.5 contains a seven-amino-acid stretch of arginine-lysine-rich sequence (RRRRKPR). An arrow indicates the homeodomain starting site (R137). Seven amino acids were mutated into polyalanines in PM1. Either the last three amino acids (KPR) (PM2) or the first four amino acids (RRRR) (PM3) are also changed to alanine (underlined). (B) These plasmids were transiently transfected into COS cells and stained with anti-Csx/Nkx2.5 MAb 2D10 (a, c, e, g, and i) and Hoechst dye (b, d, f, h, and j). WT Csx/Nkx2.5 (a) colocalized with Hoechst nuclear staining (b). PM1 and PM2 mutant proteins were predominantly localized in the cytoplasm (c and e). PM3 protein localized in the nucleus (g). (C) Western blotting of the transfected COS cells with 2D10 MAb. Compared with WT Csx/Nkx2.5, which migrated as a single band at approximately 37 kDa (lane 1), cytoplasmically localized PM1 (lane 3) and PM2 (lane 5) migrated as two separate bands. However, after CIAP treatment, these two bands migrated as a single band (lanes 4 and 6), like WT Csx/Nkx2.5 (lane 2) and PM3 (lane 8). Molecular masses (in kilodaltons) are given to the left of the gel.

FIG. 2
FIG. 2

Csx/Nkx2.5 phosphorylation in vivo. (A) Western blot of COS cells transiently transfected with the expression vector encoding the wild-type Csx/Nkx2.5. Cell lysates were incubated with (lanes 2 and 3) or without (lane 1) CIAP. Dephosphorylated Csx/Nkx2.5 protein migrated to 35 kDa (native Csx/Nkx2.5, 37 kDa). When the CIAP reaction was inhibited by Na2HPO4 (lane 3), the band shift was completely abolished. (B) Western blot of mouse neonatal heart cells. Cardiac cell lysates were immunoprecipitated with Csx/Nkx2.5 MAb (lanes 1 and 2) or control mouse IgG1 (lane 3) and incubated with (lane 2) or without (lanes 1 and 3) CIAP. After CIAP treatment, the Csx/Nkx2.5 band migrated to 35 kDa (native Csx/Nkx2.5, 37 kDa).

FIG. 3
FIG. 3

Csx/Nkx2.5 is phosphorylated in vivo and in vitro. (A) COS cells transfected with Csx/Nkx2.5 expression vector (lanes 1 and 3) or pcDNA3 vector alone (lanes 2 and 4) were 32P labeled, lysed, and immunoprecipitated (IP) with Csx/Nkx2.5 MAb (lanes 1 and 2) or control IgG1 (lanes 3 and 4). The 37-kDa Csx/Nkx2.5 band was specifically immunoprecipitated with Csx/Nkx2.5 MAb (lane 1) but not with control IgG1 (lane 3). (B) MBP alone (MBP), MBP-homeodomain Csx/Nkx2.5 (HD), and MBP-full length Csx/Nkx2.5 (Csx) were phosphorylated with [γ-32P]ATP and 2 μg of cytoplasmic (Cyt) or nuclear (Nuc) extract of NIH 3T3 cells. Samples were subjected to SDS-PAGE, and phosphorylated proteins were visualized by autoradiography. Both cytoplasmic (lanes 1 to 3) and nuclear lysates (lanes 4 to 6) approximately equally phosphorylated Csx/Nkx2.5 proteins (lane 2 versus lane 5; lane 3 versus lane 6). MBP was not phosphorylated by either cytoplasmic (lane 1) or nuclear (lane 4) extracts.

FIG. 4
FIG. 4

Phosphorylation of Csx/Nkx2.5 by CKII within the homeodomain. (A) In-gel kinase assays. NIH 3T3 cell nuclear extracts (25 μg) were separated in a gel containing either the homeodomain-MBP fusion protein (HD), the full-length Csx/Nkx2.5-MBP fusion protein (Csx), or MBP. After denaturation and renaturation, gels were incubated in kinase buffer containing [γ-32P]ATP. One kinase with a molecular mass of ∼40 kDa (arrowheads) phosphorylated the Csx/Nkx2.5 homeodomain, Csx/Nkx2.5 (lane 1), and the full-length Csx/Nkx2.5 (lane 2), but not MBP (lane 3). (B) Schematic representation of Csx/Nkx2.5 mutants carrying the consensus CKII sites and serine-to-alanine substitutions. (C) In vitro CKII phosphorylation of Csx/Nkx2.5. Three different Csx/Nkx2.5 fusion proteins described in panel B were incubated with [γ-32P]ATP in the presence (+) or absence (−) of purified CKII. Each sample was subjected to SDS-PAGE and transferred to a PVDF membrane. Autoradiography (upper panel) revealed that CKII phosphorylated the full-length Csx/Nkx2.5 (lane 1) and the homeodomain (lane 3) but neither the C-terminally deleted mutant (lane 5) nor MBP (lane 7). Loaded fusion proteins are shown by Western blotting of the same PVDF membrane using anti-MBP antibody (lower panel). (D) Serine 163 was mutated into alanine (Csx/Nkx2.5163S-A) and the kinase assay was performed. Lane 1 contains the wild-type Csx/Nkx2.5 and lane 2 contains the mutant Csx/Nkx2.5163S-A. Equal amounts of proteins were subjected to the kinase reaction as shown in the MBP Western blot (lower panel).

FIG. 5
FIG. 5

Phosphopeptide and phosphoamino acid analysis of Csx/Nkx2.5. (A) Tryptic phosphopeptide mapping of wild-type Csx/Nkx2.5 (a), CKII mutant (b), and in vitro CKII-phosphorylated MBP-homeodomain protein (c). Metabolically 32P-labeled Csx/Nkx2.5 and the CKII mutant (Csx/Nkx2.5163S-A) were immunoprecipitated with anti-Csx/Nkx2.5 MAb (2D10), resolved by SDS-PAGE, transferred to a PVDF membrane, and autoradiographed. In vitro CKII-phosphorylated MBP-homeodomain protein was also electrophoresed and autoradiographed. These protein bands were cut out, digested with trypsin, and resolved on TLC plates by electrophoresis in the first dimension and chromatography in the second dimension. In vitro CKII-phosphorylated peptide (c [arrow]) which corresponded to peptide 1 in the wild type (a [arrow]), was markedly reduced in the CKII mutant (b). Arrowheads, sample loaded points; X, lysine marker. (B) Phosphoamino acid analysis of wild-type Csx/Nkx2.5 showed Csx/Nkx2.5 is phosphorylated predominantly in S and weakly in T.

FIG. 6
FIG. 6

Csx/Nkx2.5 phosphorylation increases DNA binding affinity. Csx/Nkx2.5 protein expressed in COS cells was dephosphorylated with CIAP, and DNA binding activity was compared to that of phosphorylated Csx/Nkx2.5. (A) Transfected-cell lysates were separated by SDS-PAGE and blotted with Csx/Nkx2.5 MAb. The majority of Csx/Nkx2.5 protein was dephosphorylated by CIAP treatment (lane 2). A 5-μl aliquot of each cell lysate (lanes 1 and 2) and 5 ng of MBP-Csx/Nkx2.5 fusion protein contained equivalent amounts of Csx/Nkx2.5 protein. (B) EMSAs with ANF Csx/Nkx2.5 binding site (TGAAGTG). The DNA-Csx/Nkx2.5 protein complex (arrowhead) was supershifted by anti-Csx/Nkx2.5 MAb (arrow) in a dose-dependent manner. ANF probe was end-labeled in this experiment. (C) EMSA with ANF Csx/Nkx2.5 binding site. Lanes CIAP+ contain twofold serially diluted CIAP-treated Csx/Nkx2.5 protein; lanes CIAP− contain untreated Csx/Nkx2.5 protein. The shifted bands were scanned and plotted against protein concentration (lower panel). The plotted line was shifted downward by CIAP treatment, indicating phosphorylated Csx/Nkx2.5 has higher DNA binding affinity than dephosphorylated protein. (D) DNA mobility shift assay (similar to that shown in panel C) using A20 Csx/Nkx2.5 binding site (AGTTAATTG). Phosphorylated protein (lanes CIAP−) showed higher DNA binding affinity than dephosphorylated protein.

FIG. 7
FIG. 7

CKII phosphorylation increases DNA binding affinity. (A) EMSA with ANF Csx/Nkx2.5 binding site. Csx/Nkx2.5 homeodomain fusion protein (MBP-HD) (1 ng) incubated with either CKII (lanes CKII+) or heat-inactivated CKII (lanes CKII−) and their twofold serially diluted proteins were assayed for DNA binding. The shifted bands were scanned and plotted against the protein concentration (lower panel). CKII treatment shifted the plotted line upward, indicating that CKII-phosphorylated protein has higher DNA binding affinity. Slower-migrating bands might represent the dimerized protein. The first lane contains the gel shift with non-fused MBP protein; the last lane contains the free probe without added fusion protein. (B) A result similar to that shown in panel A is observed with A20 Csx/Nkx2.5 binding site.

FIG. 8
FIG. 8

Effect of mutation of CKII phosphorylation site in Csx/Nkx2.5 on transcriptional activation. Transcriptional activation by wild-type Csx/Nkx2.5, CKII mutant, and their carboxyl-terminal deletion mutants. 10T1/2 cells were transfected with expression vectors encoding the full-length or carboxyl-terminal deletion mutants [Csx/Nkx2.5(1-230) and Csx/Nkx2.5(1-199)] with the reporter gene ANF-luc (expression plasmid dose, 0.25 and 0.5 μg) and A20x3-luc (plasmid dose, 0.5 μg). In the wild-type Csx/Nkx2.5 (open bars), carboxyl-terminal deletion mutants (1-230) and (1-199) markedly increased the transcriptional activity on both ANF-luc and 3xA20-luc. When 0.5 μg of Csx/Nkx2.5(1-230) was transfected with the ANF reporter gene (middle panel), luciferase activity increased up to 350-fold, which was 15 times higher than that for the full-length Csx/Nkx2.5. In the CKII mutants (solid bars), similar transactivation by the carboxyl-terminal deletion was observed; however, the fold activation was almost half that of wild-type Csx/Nkx2.5. In full-length Csx/Nkx2.5, wild type and CKII mutants transactivated the reporter gene approximately equally. The fold activation value of luciferase activity is a relative value compared to the cells transfected with the empty expression vector (pcDNA3) and the reporter gene ANF-luc or A20x3-luc. Bars represent means + standard errors.

FIG. 9
FIG. 9

Sites of CKII phosphorylation and NLS. The CKII phosphorylation site of Csx/Nkx2.5 is located between the first and second helix of the homeodomain. The consensus CKII site ([S/T]XX[E/D]) is fully conserved in all NK class homeobox proteins (top set) but not in other classes of homeobox proteins with the exception of msh class (msh and mMsx-1) and POU class (mOct-6 and mPit-1) homeoproteins (middle and bottom sets). The NLS of Csx/Nkx2.5 requires the amino-terminal KPR residues of the homeodomain. The consensus NLS is the stretch of arginine and lysine residues and usually includes proline (1). Three other reported NLSs, those of the homeoproteins rat TTF-1/Nkx2.1 (24), mOct-6 (54), and yeast MAT-α2 (27) (underlined), are shown. Tyrosine residues (Y54) characteristic of the NK2 family are also marked. Note that NK-3 and NK-4 belong to the NK2 family whereas NK-1 and Nkx5.1 do not. References for homeodomain sequences are as follows: top set, , , and ; middle and bottom sets, , , and .

Similar articles

Cited by

References

    1. Alberts B, Bray D, Lewis J, Raff M, Roberts K, Watson J D. Molecular biology of the cell. 3rd ed. New York, N.Y: Garland Publishing, Inc.; 1994.
    1. Allende J E, Allende C C. Protein kinases. 4. Protein kinase CK2: an enzyme with multiple substrates and a puzzling regulation. FASEB J. 1995;9:313–323. - PubMed
    1. Aurisicchio L, Di Lauro R, Zannini M. Identification of the thyroid transcription factor-1 as a target for rat MST2 kinase. J Biol Chem. 1998;273:1477–1482. - PubMed
    1. Biben C, Harvey R P. Homeodomain factor Nkx2-5 controls left/right asymmetric expression of bHLH gene eHand during murine heart development. Genes Dev. 1997;11:1357–1369. - PubMed
    1. Biggin M D, McGinnis W. Regulation of segmentation and segmental identity by Drosophila homeoproteins: the role of DNA binding in functional activity and specificity. Development. 1997;124:4425–4433. - PubMed

Publication types

MeSH terms

Substances