PMC

Fig. 2. Deletion map and sequence alignments of the CaM regulatory domain. Schematic representation of DRP-1 deletion mutants and sequence alignments to other related kinases. Multiple sequence alignment of the CaM regulatory (CaM Reg.) and flanking regions of DAPk, DRP-1, smMLCK, CaMKI and CaMKIIa is shown. Identical amino acids are boxed; homologous amino acids are shown in grey. Arrowheads point to candidate phosphorylation sites in the CaM regulatory domain of DRP-1.

Fig. 5. Analysis of the oligomerization status of DRP-1 mutants by gel filtration. (A) Extracts prepared from cells transfected with Δ40, wild-type and S308A DRP-1 forms were run on a Superdex 75 column and fractions were analysed by western blotting using anti-HA antibodies. Arrowheads indicate the estimated molecular weights of the various fractions accomplished by running molecular weight markers under the same conditions. (B) A graph showing the relative amount of the various mutants in the different fractions (the peak value in each fractionation is considered as 100%) after densitometric quantitation of the western blot shown in (A).

Fig. 6. Induction of apoptosis by DRP-1 mutants; both dephosphoryl ation at position 308 and homodimerization are necessary conditions. (A) Percentage of apoptotic cells. The bars show the percentage of apoptotic cells resulting from cotransfections of 293 cells with DRP-1 mutants and GFP. Quantitation was performed 22 h post-transfection. This experiment was repeated three times with reproducible results. (B) Protein expression of DRP-1 mutants in 293 transfected cells. Proteins (40 µg) extracted from the transfected cells were assessed using anti-HA antibodies to visualize the level of expression. (C) The same blot was reacted with anti-GFP antibodies to normalize for the amount of transfected cell lysates in each lane.

Fig. 4. Substitution of Ser308 to alanine elevates the apoptotic activity and the homodimerization of DRP-1. (A) Percentage of apoptotic cells. Bars showing the percentage of apoptotic cells among the GFP-positive cells resulting from cotransfections of 293 cells with DRP-1 (wild type or point mutations) and GFP. A vector driving the expression of luciferase was used as a non-relevant background control (luc). Quantitation was perfomed 18 h post-transfection. (B) Protein expression of DRP-1 point mutations in 293 transfected cells. Protein extracts (30 µg) from the transfected cells shown in (A) were assessed using anti-HA antibodies to visualize the level of DRP-1 protein expression. The positions of DRP-1 monomers and dimers are marked by arrows. (C) Long exposure of the upper part of the blot shown in (B).

Fig. 10. (A) Three-dimensional model structure of the kinase domain of DRP-1 with a bound peptide derived from the CaM regulatory segment. The water-accessible surface of the kinase domain is coloured according to the electrostatic potential: red for negative and blue for positive potential. The peptide is shown as a stick diagram with Ser308 emphasized in green and the phosphate moiety in magenta. The arrowheads point to the ATP-binding P-loop and to Lys141. (B) A scheme of DRP-1 structural motifs. The scheme provides a summary of the different features that change by the point mutations or the deletions including the dimerization, CaM-binding and apoptotic functions. The various domains are marked by KD (kinase domain), CBD (CaM-binding domain) and Tail (the C-terminal 40-amino-acid peptide). The catalytic cleft is marked by a V-shaped structure in which the phosphate residue on Ser308 resides. N.D., not done.

Fig. 1. The in vitro kinase activity of DRP-1 is inversely correlated with its autophosphorylation. (A) DRP-1, DRP-1 Δ40 and DRP-1 Δ73 mutant proteins were assayed in vitro for kinase activity in the presence of Ca²⁺/CaM or EGTA as described in Materials and methods. Autophosphorylated DRP-1 proteins (between the 30 and 46 kDa markers) and exogenous MLC substrate are shown. (B) The autophosphorylation of wild-type DRP-1 protein is compared with the catalytically inactive K42A mutant (upper part); western blot analysis with anti-FLAG antibodies showing comparable wild-type and mutant protein levels within the immunoprecipitates (lower part). (C and D) Bars showing the relative amount of autophosphorylation and MLC phosphorylation by DRP-1 deletion mutants in the presence or absence of Ca²⁺/CaM, after normalizing to protein expression levels. The latter was determined by incubating the same membranes with anti-HA antibodies followed by ECL detection.

Fig. 9. Ser308 is dephosphorylated in vivo in response to activated Fas and TNF-α receptors. (A) HeLa cells ectopically expressing Fas receptors (HFB) were transfected with wild-type DRP-1, in vivo labelled and treated with anti-Fas agonistic antibodies or with TNF-α as described in Materials and methods. The levels of ³³P incorporated into DRP-1 protein were quantitated in the cell lysates; the values in the untreated culture were set as 100%. (B) The same HeLa cells were transfected with S308A DRP-1, pulse labelled in vivo, subjected to activation of Fas or TNF-α receptors, and processed for DRP-1 labelling as in Figure B. Again the values in the untreated culture were set as 100%. (C) The different DRP-1 forms (wild type, S308A and S308D) were transfected into HeLa cells; in the case of transfec tion with wild-type DRP-1, cells were treated for the last 2.5 h with TNF-α before their extraction. Proteins were separated on a TSK-G3000 column and the pattern of elution was assessed by western blotting with anti-HA antibodies. The densitometric quantitation of the western blots is shown; the relative amounts of DRP-1 protein in each fraction out of the total eluted protein are presented. The positions of the monomer and dimer are indicated by arrows.

Fig. 8. Ser308 is auto-phosphorylated in vivo. (A) 293 cells, transfected with wild type, S308A and K42A DRP-1 mutants (left), and HeLa cells, transfected with wild type and S308A mutants (right), were subjected to in vivo labelling with [³³P]orthophosphate as described in Materials and methods. Proteins were extracted and the levels of ³³P incorporated into DRP-1 were quantitated after normalization to the protein expression levels. (B and C) Mass spectrometric analysis of ectopically expressed DRP-1 protein extracted from growing Hela cells. (B) DRP-1 tryptic peptides were analysed by liquid chromatography–mass spectrometry. Full mass spectrum at retention time 60 min. The masses corresponding to singly and doubly charged ions of peptide 307–320 in the unphosphorylated (m/z = 1589 and 795.6) and the phosphorylated (m/z = 1669.4 and 835.3) forms are indicated by arrows. (C) CID of the m/z = 835.3 peptide and comparison with the simulated CID of phosphorylated 307–320 DRP-1 peptide. The major fragment m/z = 795.6 (underlined) correlates with the doubly charged mass of the full peptide after HPO₃ removal (indicating the existence of a single phosphate residue on this peptide). The relevant CID fragments are marked in bold and their amino acid sequence and masses are shown at the bottom. The phosphoserine is marked as S*.

Fig. 3. Identification of Ser308 as the major autophosphorylation site of DRP-1 in vitro. (A) DRP-1 and its point mutation constructs were expressed and imunoprecipitated with anti-HA antibodies. The proteins were then assayed in vitro for kinase activity (autophosphorylation and MLC phosphorylation) in the presence or absence of Ca²⁺/CaM as described in Figure A. (B and C) Bars showing the relative amount of autophosphorylation and MLC phosphorylation by DRP-1 point mutations in the presence or absence of Ca²⁺/CaM, after normalizing to protein expression levels as described in Figure C and D. (D and E) Mapping autophosphorylation sites by liquid chromatography–mass spectrometry. (D) Full mass spectrum at retention time 53 min. The masses corresponding to peptide 305–320 in the unphosphorylated (m/z = 980.2), the phosphorylated (m/z = 1021.6) and the dephosphorylated forms (m/z = 972.8) are marked by arrows. (E) Collision-induced dissociation (CID) of the m/z = 1021.6 peptide and comparison with the simulated CID of phosphorylated 305–320 DRP-1 peptide. The major fragment m/z = 972.2 (underlined) correlates with the doubly charged mass of the full peptide after H₃PO₄ removal (indicating the existence of a single phosphate residue on this peptide). The relevant CID fragments are marked in bold and their amino acid sequence and mass are shown. It indicates that the phosphate residue resides exclusively on Ser308 and not on Ser310, Ser313 or Thr319. The CID simulation of the phosphorylated 305–320 DRP-1 peptide is shown at the bottom. The phosphoserine is marked as S*.

Fig. 7. DRP-1 mutant forms differ in the binding affinity and activation by CaM and in their dimerization status. (A) Differential activation by CaM of DRP-1 mutant forms. Δ40, wild-type, S308A and S308D DRP-1 constructs were transfected into 293 cells, immunoprecipitated and eluted from the beads with an excess of HA peptide. Equal amounts of the various DRP-1 mutants (estimated by western blotting) were then assayed in an in vitro kinase assay towards MLC in the presence of various concentrations of CaM (0–1000 nM). (B) Differential affinity to CaM by DRP-1 mutant forms. Δ40, wild-type, S308A and S308D DRP-1 constructs were transfected into 293 cells and immunoprecipitated with anti-HA antibodies. Equal amounts of the various DRP-1 mutants were then subjected to CaM binding assay with 200 or 2 nM ³⁵S-labelled CaM as described in Materials and methods. For each CaM concentration the highest binding value was set as 100% (25 227 and 1055 c.p.m. for 200 and 2 nM CaM, respectively); the remaining values were calculated accordingly. (C) The immunoprecipitated beads carrying the various DRP-1 mutants were fractionated on gels and reacted with anti-HA antibodies. The position of DRP-1 monomers and dimers as well as of the immunoglobulin heavy chain is indicated by arrows. (D) Samples of clarified cell extracts (20 µg; 2 µg/µl) were subjected to cross-linking with glutardialdehyde as described in Materials and methods, then separated on gels and western blotted. Another sample from each extract (10 µg) was run in parallel without a prior cross-linking (TCL, total cell lysate). The immunoblots were reacted with anti-HA antibodies. The cross-linked dimer runs faster than predicted due to the globular conformation maintained by the cross-linking agent.