Figure 6. The D326K mutation destabilizes the structure of the Munc18-1 loop.

(A) Ribbon diagram of Munc18-1 (violet) bound to syntaxin-1 (SNARE motif in yellow; N-terminal region in orange) (PDB code 3C98) (Burkhardt et al., 2008). The red ellipse shows the location of the loop that connects two helices of domain 3a of Munc18-1. (B) Close-up view of the domain 3a loop region from panel A. Side chains from the loop and the two helices are shown as stick models, and the atoms of the side chains from D326, M316, M324, M330 and M334 are color-coded (carbon green; oxygen red; sulfur yellow). The diagram illustrates that these side chains are well packed within the furled structure of the loop, although they have different degrees of solvent accessibility. (C) Ribbon diagram showing the dimeric structure observed in the crystals of Munc18-1 bound to a syntaxin-4 N-terminal peptide (not shown) (PDB code 3PUJ) (Hu et al., 2011). One Munc18-1 molecules is shown in gray and the other in blue. Note that dimerization involves the two domain 3a helices but the loop is at the end of the dimerization interface. The red ellipse shows the location of the loop for the gray molecule. (D) Close-up view of the loop in the gray Munc18-1 molecule of panel C. Atoms from the loop and the two helices are shown as stick models, and the atoms observed for M324, D326, M330 and M334 are color-coded (carbon green; oxygen red; sulfur yellow). Note that residues 315–323, as well as the side chains of M324, D326 and M330, were not observable, probably because they are not packed against other regions of Munc18-1 and they are thus dynamic. (E) Expansions from the methionine methyl region of ¹H-¹³C HMQC spectra of WT 50%-²H-ILMV-¹³CH₃-Munc18-1 free (black contours) and bound to syntaxin-1 (2–253) (orange contours). The arrows indicate potential shifts caused by syntaxin-1 (2–253) binding on the cross-peaks tentatively assigned to M330 and M334 (Figure 6—figure supplement 2). (F) Expansions from the methionine methyl region of ¹H-¹³C HMQC spectra of D326K 50%-²H-ILMV-¹³CH₃-Munc18-1 free (magenta contours) and bound to syntaxin-1 (2–253) (cyan contours). (G,H) Expansions of the ¹H-¹³C HMQC spectra shown in panels E, F but superimposing the spectra of WT and D326K 50%-²H-ILMV-¹³CH₃-Munc18-1 in isolation (G) or bound to syntaxin-1 (2–253) (H). The spectra were plotted at higher contour levels to emphasize the spectral changes induced by the D326K mutation. Arrows indicate cross-peak shifts. Cross-peaks are labeled with residue assignments based on spectra acquired on methionine mutants (Figure 6—figure supplement 2); the ? symbols after the residue numbers indicate the tentative nature of the assignments. Green boxes indicate regions of the spectra that were plotted at lower levels to show weak cross-peaks. New cross-peaks caused by the D326K mutation are indicated by **.

DOI: http://dx.doi.org/10.7554/eLife.24278.014

Figure 6—figure supplement 1. Structure of squid Munc18-1.

Ribbon diagram of the crystal structure of squid Munc18-1 (PDB code 1EPU) (Bracher et al., 2000) showing the unfurled structure of the loop connecting two helices of domain 3a.

DOI: http://dx.doi.org/10.7554/eLife.24278.015

Figure 6—figure supplement 2. Assignment of methionine cross-peaks in ¹H-¹³C HMQC spectra of 50%-²H-ILMV-¹³CH₃-Munc18-1.

(A–F) Superpositions of the methionine methyl region of ¹H-¹³C HMQC spectra of WT (black contours) and M316L (A,B), M330E (C,D) or M334L (E,F) mutants (red contours) 50%-²H-ILMV-¹³CH₃-Munc18-1 free (A,C,E) or bound to syntaxin-1 (2–253) (B,D,F). The spectra were plotted at different contour levels to help with visualizing the most pronounced perturbations caused by the mutations. Arrows show cross-peak shifts. Tentative assignments are indicated for cross-peaks that shift or disappear due to the mutations. All assignments must be considered tentative (hence the ? symbols) because of the complexity of the perturbations caused by the mutations, but we feel more confident with the assignments of the syntaxin-1 (2–253)-bound form. Blue boxes indicate regions of the spectra that were plotted at lower levels to show weak cross-peaks. New cross-peaks caused by the mutations that do not arise from an obvious cross-peak shift are labeled with **.

DOI: http://dx.doi.org/10.7554/eLife.24278.016

Figure 6—figure supplement 3. The D326K mutation causes selective decreases in the intensities of the ¹H-¹³C HMQC cross-peaks of methionines from the domain 3a loop of Munc18-1.

The plots show the ratios between the intensities of methionine cross-peaks observed for D326K 50%-²H-ILMV-¹³CH₃-Munc18-1 and those observed for WT 50%-²H-ILMV-¹³CH₃-Munc18-1 in the absence (left panel) and presence (right panel) of syntaxin-1 (2–253). The data for the cross-peaks assigned to the four methionines from the domain 3a loop are grouped on the left of each plot, while those for other well-resolved (unassigned) cross-peaks are grouped on the right. To account for small differences in protein concentrations, all intensities in each spectrum were normalized with the average intensity of the unassigned cross-peaks.

DOI: http://dx.doi.org/10.7554/eLife.24278.017