Protein kinase A phosphorylation modulates transport of the polypyrimidine tract-binding protein

Abstract

The heterogeneous nuclear ribonucleoprotein particle (hnRNP) proteins play important roles in mRNA processing in eukaryotes, but little is known about how they are regulated by cellular signaling pathways. The polypyrimidine-tract binding protein (PTB, or hnRNP I) is an important regulator of alternative pre-mRNA splicing, of viral RNA translation, and of mRNA localization. Here we show that the nucleo-cytoplasmic transport of PTB is regulated by the 3',5'-cAMP-dependent protein kinase (PKA). PKA directly phosphorylates PTB on conserved Ser-16, and PKA activation in PC12 cells induces Ser-16 phosphorylation. PTB carrying a Ser-16 to alanine mutation accumulates normally in the nucleus. However, export of this mutant protein from the nucleus is greatly reduced in heterokaryon shuttling assays. Conversely, hyperphosphorylation of PTB by coexpression with the catalytic subunit of PKA results in the accumulation of PTB in the cytoplasm. This accumulation is again specifically blocked by the S16A mutation. Similarly, in Xenopus oocytes, the phospho-Ser-16-PTB is restricted to the cytoplasm, whereas the non-Ser-16-phosphorylated PTB is nuclear. Thus, direct PKA phosphorylation of PTB at Ser-16 modulates the nucleo-cytoplasmic distribution of PTB. This phosphorylation likely plays a role in the cytoplasmic function of PTB.

Fig. 1.

Endogeneous PTB is a phospho-protein. (A) Immunoprecipitation using anti-PTB (BB7) or anti-Flag control antibody from HEK 293 cells labeled in ³²P-orthophosphoric acid. An SDS/PAGE gel of the immunoprecipitated proteins is shown. The molecular mass markers are to the left (M). (B) PTB domains and the position of the PKA target Ser-16 (Left). S16A and control S21A mutants are shown below the sequence. The alignment of the Ser-16 region from several species is shown (Right). The peptide sequence used for making antibodies is boxed.

Fig. 2.

PKA phosphorylates PTB at Ser-16 in vitro and in vivo. (A) Purified recombinant PTB, nPTB, and mutant proteins were incubated with purified PKA and [γ-³²P]ATP and separated on the gel shown. (Upper) Bands for phosphorylated PTB and nPTB, autophosphorylated PKA are indicated. (Lower) Coomassie blue-stained PTB is shown. (B) Phospho-amino acid analysis by two-dimensional electrophoresis. Positions of standard phospho-serine (pS), phospho-tyrosine (pY), or phospho-threonine (pT) are shown as ellipses. (C)(Left) Phospho-peptide mapping of the PKA-phosphorylated PTB and its mutants. (Right) TLC of the phospho-peptides immunoprecipitated with the pS16-specific antibody is shown. The wild-type and anti-pS16 IP samples were run simultaneously in the same tank to allow accurate alignment of the plates. (D) Dot blot showing the specificity of the pS16 and S16 antibodies for phosphorylated or unphosphorylated peptides. (E) Stimulation of the PKA pathway increases phosphorylation of PTB Ser-16 in PC12 cells. Total protein of PC12 cells treated with or without forskolin (10 μM) for 6 h, was probed with the antibodies as indicated. The arrow indicates the PTB band. The arrowhead indicates a band weakly recognized by the anti-pS16 antibody but not by the PTB antibody BB7. This band may be the neuronal PTB homologue nPTB.

Fig. 3.

Ser-16 is required for nuclear export of PTB. (A) A Western blot of the total protein from HEK 293 cells either not transfected (NT), or transfected with vector (V), wild-type Myc-PTB, or the S16A or S21A mutants. This was probed with anti-Myc antibodies to control for equal expression of the transfected clones as compared with the endogeneous U1 70K protein. (B) Confocal microscopy of HEK 293 cells transfected with Myc-PTB plasmids and stained with anti-Myc. All of these proteins show strongly nuclear staining. (C) Heterokaryon assay of wild-type Myc-PTB or its mutants expressed in human HEK 293 cells that were fused with mouse NIH 3T3 cells. Mouse nuclei are distinguished from human nuclei by their punctate Hoechst staining. Actin staining of the fused cells with phalloidin-TRITC confirmed the cell fusion by determining that the actin cytoskeleton surrounded the multiple nuclei. Arrowheads indicate the positions of the mouse nuclei in heterokaryons. (D) Bar graph of the average (±SD) intensity of the Myc-PTB staining in the mouse nuclei relative to the human nuclei of the same heterokaryons (n = 4, 10, and 9 heterokaryons, respectively).

Fig. 4.

PKA coexpression relocalizes PTB to the cytoplasm and the S16A mutation blocks this effect. (A) A Western blot of protein from HEK 293 cells transfected with Flag-tagged PKA. (B) Confocal microscopy of anti-Myc-stained PTB or anti-Flag-stained PKA in HEK 293 cells. (C) To distinguish cells with strong nuclear staining from those with stronger cytoplasmic Myc-PTB staining, anti-Myc immunoflourescence intensity was plotted along a line drawn across the cells. Cells showing a single peak of flourescence centered on the nucleus were counted as predominantly nuclear stainings. Cells showing a bimodal distribution with a dip over the nucleus were counted as predominantly cytoplasmic. These data are presented in a bar graph at the right as the percentage of cells with predominantly nuclear staining of Myc-PTB with PKA coexpression. These data are from the cotransfection of 2 μg of Myc-PTB with 6 μg of PKA plasmid, where n for each bar is 178, 107, and 78 cells, respectively. (D) Activating the PKC pathway does not relocalize Myc-PTB to cytoplasm. Shown are the confocal images of Myc-PTB cotransfected with pPKCγ-EGFP. The cells were stimulated with TPA (40 ng/ml) to activate PKCγ. Note that the activated PKCγ proteins are typically localized at the cell membrane.

Fig. 5.

The N terminus of PTB is sufficient to mediate PKA-dependent relocalization. (A) Diagrams of the PTB (amino acids 1–55)-EGFP fusion proteins. (B) The EGFP fluorescence intensity in transfected cells was measured with the programnih image 1.62b7 for the cytoplasm and nucleus of each cell and the ratio determined (C/N). The average increase in the C/N ratio after PKA expression is plotted (n = 6, 5, and 5 cells, respectively). The C/N ratio approximately doubles after PKA expression for the wild-type and S21A PTBs, but is unchanged for the S16A mutant. (C) Confocal images of HEK 293 cells transfected with the constructs in A. These are coexpressed either with (Lower) or without (Upper) PKA. Note that the localization of EGFP itself was not affected by PKA coexpression.

Fig. 6.

Ser-16 is phosphorylated by PKA in vivo. (A) Western blot of immunoprecipitated Myc-PTB from transfected HEK 293 cells. Note that the phospho-Ser-16 band is only detectable in the PKA cotransfected sample. (B and C) Ser-16 is the major site of PKA phosphorylation. (B) Immunoprecipitation of radiolabeled PTB from transfected HEK 293 cells grown in ³²P-orthophosphoric acid (Left). This phosphorylated protein was subjected to phospho-amino acid analysis as in Fig. 2. (C) Phospho-peptide mapping of the immunoprecipitated Myc-PTB showed a major peptide (b) and two less intense spots (a and c). There were also three minor spots (x, y, and z) that were used to align the panels. The major spot b is eliminated by the S16A mutation. This spot is immunoprecipitated by the pS16 antibody (Right). The wild-type and anti-pS16 IP samples were run in the same tank, and the spot b in both samples is at the same distance from the starting point in the first dimension.

Fig. 7.

Phospho-Ser-16-PTB is restricted to the cytoplasm in X. laevis oocytes. Nuclear or cytoplasmic fractions of stage III/IV oocytes were analyzed by Western blot using either anti-phospho-Ser-16 (pS16), anti-nonphosphorylated PTB (S16), or anti-Xenopus PTB (R3B3) antibodies. Oocytes were manually dissected into nuclear and cytoplasmic fractions. The pS16 and S16 lanes used 10 oocyte equivalents of proteins, and the R3B3 lanes used 5 oocyte equivalents.