Fig 1.
Pfn1 can be phosphorylated by PKA on multiple residues.
A) His-tagged Pfn1 was treated with buffer containing ATP with or without PKA followed by 2D electrophoresis and visualized by silver staining (isoelectric focusing was conducted using a pH 4–7 IPG strip). Phosphorylation spots are marked by arrows. B) Mass-spectrometry based detection of phosphopeptides from PKA-treated His-Pfn1 reveled Pfn1 phosphorylation on three residues (S56/S57, T89, and S91/T92).
Fig 2.
Pfn1 can be post-translationally modified in a PKA-dependent manner in HEK-293 cells.
A) Lysates prepared from myc-Pfn1-expressing HEK-293 cells were resolved on 4–7 pH gradient IEF gels and then immunoblotted with anti-myc antibody to reveal the isoelectric profile of myc-Pfn1 in serum-starved (control: S-) vs FSK-stimulated conditions—two different exposure blots are shown. B) A bar graph summarizing the relative average intensity of the spot representing the most negatively charged form of myc-Pfn1 (spot #4) in control vs FSK-stimulated conditions. C-D) Isoelectric profiles of myc-Pfn1 in FSK-stimulated cells with or without H89 treatment (panel C), and the corresponding quantifications of the intensity of spot #4 (panel D). N denotes the number of independent experiments (*: p<0.05).
Fig 3.
Molecular interactions of T89 of Pfn1.
A) Actin-interacting surface of Pfn1 (Pfn1: blue; actin: grey); phosphorylatable residues of Pfn1 are indicated as sticks. B) In unbound Pfn1 (PDB 1PFL) T89 forms an intramolecular hydrogen bond with F98. C) In the Pfn1-actin complex (PDB 2BTF) an intermolecular hydrogen bond (dashed yellow line) is observed between T89 of Pfn1 (blue) and Y166 of actin (green).
Fig 4.
Molecular Dynamics Simulation predicting the effect of T89 phosphorylation on the stability of intramolecular 98N-89O bond.
The stability of this intramolecular 98N-89O bond over the course of three 300ns molecular dynamics simulations of monomeric Pfn1 are shown for homology models based on the bound structure for WT, T89A, T89D and phosphorylated T89 (THP89) variants of Pfn1 (each color represents an individual simulation). The histograms alongside the line graphs summarize the results of all three simulations for each variant of Pfn1.
Fig 5.
Effect of site-specific phosphorylation on biochemical characteristics of Pfn1.
A) Pull-down assays of indicated GST-tagged Pfn1 constructs with HEK-293 cell lysate were run on an SDS-PAGE and immunoblotted with anti-actin and anti-VASP antibodies (GST was used as a negative control). Coomassie stain in parallel confirms comparable amounts of GST-tagged proteins in the pull-down assay. Note that virtually negligible amount of T89D-Pfn1 was found to be immobilized on glutathione-linked agarose beads. B) Bacteria expressing indicated GST-tagged Pfn1 constructs were lysed with either non-denaturing (containing 1% NP-40) or denaturing (containing 1% NP-40, 2% SDS for one buffer and the other with 6M urea in addition) extraction buffers. Bacterial lysates were immunoblotted with anti-Pfn1 antibody to demonstrate that T89D-Pfn1 is insoluble in non-denaturing lysis buffer. C) HEK-293 cells expressing indicated EGFP-fused Pfn1 constructs were lysed with either non-denaturing (containing 1% NP-40) or denaturing (containing 1% NP-40, 2% SDS for one buffer and the other with 6M urea in addition) extraction buffers. HEK-293 lysates were immunoblotted with anti-GFP antibody to demonstrate that GFP-T89D-Pfn1 is also insoluble in non-denaturing lysis buffer. Note that endogenous Pfn1 level is not affected by expression of any of the ectopic Pfn1 constructs and extractable completely in non-denaturing lysis buffer. D) Fluorescence images of HEK-293 cells expressing indicated EGFP-fused Pfn1 constructs show that EGFP-Pfn1-T89D causes clustering of this fusion protein as indicated by the arrows. Scale bar represents 20 μm. E) HEK-293 cells expressing indicated EGFP-fused Pfn1 constructs were treated with CHX for up to 8 hours. Cell lysates prepared at different time-points after CHX addition were immunoblotted with the indicated antibodies. T89D-Pfn1 undergoes rapid protein degradation while WT- and T89A-Pfn1 are stable over that period of time, similar to the characteristic of endogenous Pfn1 (degradation of p27kip1, a cell-cycle protein that undergoes rapid turnover, validates CHX efficacy). Tubulin blot serves as the loading control. F) The bar graph summarizes quantification of the time-dependent changes in the expression of the indicated GFP-Pfn1 constructs following CHX treatment in HEK-293. Data was summarized from 3 independent experiments (** indicates p < .01).
Fig 6.
T89D-Pfn1 co-clusters with actin in cells.
A) Lysates prepared from MDA-231 cells expressing indicated EGFP-fused Pfn1 constructs were immunoblotted with anti-GFP antibody to show the relative levels of various EGFP-Pfn1 constructs (tubulin blot serves as the loading control). B) Fluorescence images of MDA-231 cells expressing indicated EGFP-fused Pfn1 and mCherry-actin demonstrate mCherry-actin clustering at the sites of T89D-Pfn1 aggregates (arrows; Scale bar—20 μm). C) Fluorescence images of MDA-231 cells co-expressing mCherry-actin and EGFP-Pfn1-T89D, and treated with Pfn1 siRNA reveal mCherry-actin:T89D-Pfn1 co-clusters (arrows; Scale bar—20 μm). Pfn1 immunoblot alongside confirms near complete loss of endogenous Pfn1 expression in Pfn1-siRNA treated cells when compared against control siRNA transfected cells (tubulin blot—loading control).
Fig 7.
Overexpression of T89D-Pfn1 has a robust effect on actin cytoskeleton in MDA-231 cells.
A) Fluorescence images of MDA-231 cells expressing indicated EGFP-fused Pfn1 constructs and stained with rhodamine-phallodin staining. Scale bar represents 20 μm. B) A bar graph summarizing the average rhodamine-phalloidin fluorescence intensity/cell for T89A-Pfn1 and T89D-Pfn1 expressors relative to that of WT-Pfn1 expressing and untransfected cells. ‘N’ indicates number of cells analyzed for each group pooled from 3 independent experiments (** indicates p < .01; * indicates p < .05).
Fig 8.
Effect of non-phosphorylatable amino-acid substitution at T89 on post-translational modification profile of Pfn1 in cells.
2D-GE of lysates from HEK-293 cells expressing either myc-Pfn1 or myc-Pfn1-T89A demonstrate that alanine substitution at T89 results in a basic shift of a small portion of myc-Pfn1.