Fig 1.
The cFLIPL/calmodulin interaction is mediated by the DED1 domain.
Calmodulin conjugated on sepharose resin was used as bait in all pull-down assays and purified recombinant cFLIPL or its DED domains were used as prey (10 g each). Pull-down lanes consisted of 50 μL resin and supernatant lanes of 500 μL buffer, rendering a dilution factor of 10 for non-interacting prey proteins in the SDS/Coomassie gel. (A) Full-length cFLIPL pulls down on calmodulin-conjugated resin (lane 2), with DED1 strongly interacting with calmodulin (lane 3) while DED2 exhibits almost no interaction (lane 4). GST alone, used as a control, displayed no binding to calmodulin (lane 1). (B) The pull-down was repeated for DED1 and full-length cFLIPL in the presence (lanes 1 and 3) or absence (lanes 2 and 4) of Ca2+. Neither cFLIP nor its DED1 interact with calmodulin in the absence of Ca2+ (lanes 2 and 4). Only pull-down lanes are shown. (C) To further confirm the specificity of the DED1/calmodulin interaction, a competition assay was carried out in the pull-down format, illustrated conceptually in this schematic. Recombinant GST-DED1 is incubated with calmodulin-conjugated resin in the presence of increasing amounts of free calmodulin, which is expected to compete with the resin for binding to DED1. (D) Results of assay described in C. Free calmodulin displaces DED1 from the resin in a dose-dependent manner.
Fig 2.
The interactions identified in pull-downs are recapitulated in ELISA experiments.
In all ELISA experiments recombinant calmodulin was used as bait and immobilized, while GST-tagged DED1 or DED2 constructs were used as prey. (A) GST-DED1 was titrated and the affinity of the interaction with calmodulin was quantified. The two proteins interact with an affinity of 2 μM. (B) Binding of GST-DED2 is much weaker compared to DED1. (C) When GST-DED1 is kept at the KD concentration, the R4 peptide competitively displaces DED1 from calmodulin. GST-DED1 was allowed to bind calmodulin, and then the R4 peptide was titrated. The R4 peptide binds with an IC50 of 63 μM. (D) and (E): Swap of the R2R4 regions correlates with a loss of binding for DED1 and a gain of binding for DED2. DED1-(DED2-R2R4) exhibits diminished calmodulin binding, indicating a loss in activity (D) and DED2-(DED1-R2R4) displays a clear gain in calmodulin-binding activity (E). Error bars denote standard deviation of three replicates.
Fig 3.
The interaction with calmodulin is mediated by the C-terminal part of cFLIP DED1.
(A) Schematic representation of the C-terminal swaps between DED1 and DED2. DED1 segments are illustrated in blue and DED2 segments in red. (B) Constructs illustrated in A were used as prey in a pull-down assay on calmodulin-conjugated resin. DED1 constructs containing DED2 segments lose the ability to bind calmodulin (lanes 1–3), while DED2 constructs containing DED1 segments show a gain in binding ability for calmodulin (lanes 4–6). (C) Synthetic R2, R3, R4 peptides corresponding to the DED1 regions shown in A were titrated into the pull-down assay to displace GST-DED1 from calmodulin-conjugated resin; R2 and R4 show inhibitory activity, while R3 is ineffective. Only pull-down fractions are shown for simplicity. (D) Homology model of DED1 highlights the position of the R2, R3, and R4 regions. Amino acid sequences for the three peptides are shown next to the model.
Fig 4.
The DED1 R4 peptide interacts with calmodulin.
(A)1H, 15N-HSQC spectrum of calmodulin before (blue) and after addition of R4 (red), in the presence of Ca2+. The peak shifts between the two spectra are indicative of a binding event. (B) Chelation of Ca2+ with EDTA abolishes the interaction. The spectra shown in blue and red are in the absence or presence of excess R4, respectively. The high degree of spectral overlap indicates there is no interaction between R4 and the protein. (C)-(F) Enlarged views of corresponding insets from panel A, showing examples of residues directly involved in binding to R4. Spectra depict the following R4:calmodulin ratios: blue—0:1; cyan—1:1; green—1.5:1; magenta 2:1; red—5:1. Other titration points were excluded for ease of visualization. Arrows indicate the direction in which the peaks shift.
Fig 5.
Calmodulin binding epitope for R4 peptide.
Chemical shift perturbations (CSP) observed on the amino acid residues of 15N-labeled calmodulin upon R4 peptide binding. The magnitude of the CSP is plotted for every calmodulin residue. Eight residues experience shifts larger than two standard deviations (red). Eighteen additional residues experience shifts larger than one standard deviation (orange-yellow). Negative bars indicate residues that could not be unambiguously assigned.
Fig 6.
Model of R4 peptide/calmodulin complex.
The solution structure of R4 was docked into calmodulin and the model was refined by molecular dynamics. Colored on calmodulin are the residues which undergo significant chemical shifts in the NMR titration experiments, ranging from red (strongest) to light yellow (weakest). The color scheme corresponds to that in Fig 5. The R4 hydrophobic anchors F65 and I71 are shown, as well as the calmodulin residues that experience the strongest CSPs, M111 and F21. (A) The model is oriented to facilitate visualization of the peptide in the pocket (the N-lobe of calmodulin is facing upward and the C-lobe downward). (B) 90° rotation of the model in A.