Figure 1.
BAP29 and BAP31 coiled coil predictions and sequence alignments.
(a) Coiled coil prediction. The prediction was carried out on the full-length C-terminal cytoplasmic domains of BAP29 (left) and BAP31 (right) using the Coils server (window size; 21). Similar results were obtained with the Paircoil2 server, except that the N-terminal most coiled coil scored under the default threshold in the case of BAP29 (not shown). (b) Global pairwise sequence alignment of human BAP29 and BAP31. Residues shaded blue are identical. For BAP31, the predicted transmembrane helices are boxed in green and the two coiled coils predicted using the Coils server are labeled CC1 and CC2 and marked with pink bars (a score of 0.9 was used as cut-off). Grey triangles denote the caspase-8 cleavage sites in BAP31, (#) symbols denote start and end residues of the full-length C-terminal cytoplasmic regions used in this study, and (*) symbols denote start and end of the vDED constructs. The ‘a’ and ‘d’ positions of the coiled coil heptad repeats of BAP31 vDED are labeled according to the crystal structure. (c) Multiple BAP31 vDED:DED sequence alignment. BAP31 vDED (here defined as the region between the two caspase-8 cleavage sites) was aligned with six different human DED domains; DED1 and DED2 from caspase-8 (UniProt: Q14790) and from CASH (UniProt: O15519), and the single DED domains of FADD (UniProt: Q13158) and FLAME3 (UniProt: Q8WXF8). The alignment is colored in shades of blue according to sequence identities and annotated with the secondary structure of FADD. BAP31 vDED, shares 10–18% sequence similarity with caspase-8 DED1 (15%), caspase-8 DED2 (18%), FADD DED (18%), CASH DED1 (11%), CASH DED2 (10%) and FLAME3 DED (16%).
Figure 2.
Limited proteolysis of the C-terminal cytoplasmic regions.
Samples of the cytoplasmic regions of BAP31 (left) and BAP29 (right) were treated with chymotrypsin and quenched at the indicated time points. For reference, the SDS gels were run together with the untreated vDED domains and a molecular marker (weight for the bands are indicated in kDa). Bands marked B1–B4 were cut out and subjected to in gel trypsination and MALDI TOF mass spectrometry analysis. The detected peptides are listed and depicted graphically beneath the gels and the vDED constructs are shown for reference.
Figure 3.
CD spectroscopy of the C-terminal cytoplasmic regions and vDED domains.
(a) CD spectra of the four expressed BAP29 and BAP31 constructs. Spectra are shown at both pH 7.0 (stippled lines) and pH 4.2 (solid lines) for BAP31 vDED (top left), BAP31 full-length C-terminal cytoplasmic region (top right), BAP29 vDED (bottom left) and BAP29 full-length C-terminal cytoplasmic region (bottom right). (b) CD melting curves. Curves are shown for BAP31 vDED (top), BAP31 full-length C-terminal cytoplasmic region (middle) and BAP29 vDED (bottom). For BAP31, curves were obtained at both pH 7.0 and pH 4.2 as annotated on the figure. (c) Effect of pH on stability. Plot shows the midpoint of thermal unfolding of the BAP31 C-terminal cytoplasmic region as a function of pH at a protein concentration of 1 µM. (d) Effect of protein concentration on stability. Plot shows the midpoint of thermal unfolding of the BAP31 C-terminal cytoplasmic region as a function of protein concentration at pH 6.0.
Table 1.
CD results.
Figure 4.
Structures of the BAP31 vDED domain.
(a) Structural overview of the two crystal forms. (b) Selected inter-helical interactions. Side chain interactions discussed in the text are shown for the P21 form. Stippled lines indicate hydrogen bonds or ionic interactions. (c) Charge distribution at the surface. The P21 form is colored by electrostatic potential (inset shows color code) and shown in two orientations; the orientation on the left is the same as in the top panel in (a). The stippled box denotes a small predominantly acidic region. (d) Putative tetramer. In the P21 form, two dimers form a tail-to-tail tetramer across a crystal contact, which is predicted by the PISA server to be stable in solution.
Figure 5.
Quaternary structure of the vDED domain.
(a) Gel filtration analysis of BAP31 vDED at pH 7.0. Different concentrations of BAP31 vDED were used; 10 µM (blue), 25 µM (grey), 50 µM (green), 100 µM (grey), 200 µM (pink) and 400 µM (red). As seen in the top right graph, the elution volume was approximately the same for all concentrations (average is 16.54 mL). The column was calibrated with aldolase (Stoke’s radius of 48.1 Å), conalbumin (36.4 Å), ovalbumin (30.5 Å), and carbonic anhydrase (23 Å). Plotting elution volume against Stoke’s radius of these standards (bottom right) allowed us to estimate the Stokes’ radius of BAP31 vDED to 28.97. (b) Chemical cross-linking analysis of BAP29 vDED and BAP31 vDED. BAP29 vDED was cross-linked with glutaraldehyde (labeled GA on the figure) at pH 7.5 and the reaction was quenched at different time points as indicated beneath the gel. The weak dimer band seen is due to spontaneous disulfide bridge formation (it is also present in the untreated sample). Cross-linking of BAP31 vDED with glutaraldehyde resulted in aggregation (not shown). Both BAP29 vDED and BAP31 vDED were also cross-linked with DMS at pH 7.5 at various protein concentrations as indicated in the figure. (c) Native ESI mass spectrometry analysis of BAP31 vDED. A substantial peak was seen for the dimer species (≥35% of the monomer peak height) while peaks for the trimeric and tetrameric species were minuscule (∼3% and <1% respectively of the monomer height).
Figure 6.
BAP29–BAP31 interaction studies.
(a) IMAC pull-down. Left; His-tagged full-length C-terminal cytoplasmic domain of BAP31 was adsorbed to the nickel resin (run fraction; lane PR) and excess protein was washed off (PW). Purified untagged BAP29 full-length cytoplasmic domain was then applied to the column (run fraction; lane R). The column was then thoroughly washed (W1–W4) and eluted with high concentrations of imidazole (E1, E2). No co-elution of BAP29 was observed. Right, as for the left panel, but BAP29 was adsorbed and BAP31 applied to the column afterwards. The dimer bands seen on the gels for BAP29 are due to spontaneous disulfide bridge formation. (b) Analytical gel filtration. Runs were conducted at both pH 7.0 (left) and pH 4.2 (right) with BAP31 vDED alone (red), BAP29 vDED alone (blue) or both proteins together (green). No shifts were observed, indicating that a complex is not formed. Similar gel filtration runs were also conducted for the full-length C-terminal cytoplasmic domains and combinations of the vDED and full-length domains. Also here, no complex formation was observed (results not shown).
Table 2.
Data processing and refinement statistics.