Skip to main content
. 2020 Sep 9;9:e57784. doi: 10.7554/eLife.57784

Figure 6. Interfaces at which docking occur are stabilized in an open form, while remaining interfaces close.

(A) Schematic showing the naming scheme used to identify subunits. The interhelix angle between axes of αA helices are used to measure the rotation of each hub domain with respect to the adjacent hub domain. (B) Variation of the angle between the axes of the αA helices of adjacent subunits in the simulation with the regulatory segment. The darker traces are the time-averaged values of the interhelix angles calculated using a moving window of 240 ns, while the lighter shades are the actual distances. The two interfaces where regulatory segments dock between subunits L and B (colored blue), and between subunits F and H (colored pink) open and the remaining interfaces close. In the lower ring the interfaces mirror the behavior of the interfaces in the upper ring. The interhelix angle between the αA helices of adjacent subunits from the crystal structure of an autoinhibited, dodecameric holoenzyme (PDB ID - 3SOA) (Chao et al., 2011), which is perfectly symmetrical, is indicated by the horizontal black line. (C) Variation in the angle between the axes of the αA helices of hub domains that flank the interfaces where the calmodulin-binding elements dock (shown in red), and the distance between a residue from the calmodulin-binding element and the interfacial β-sheet (shown in blue), over the course of the simulation. The darker traces are the time-averaged values, calculated using a moving window of 12 ns, while the lighter shades are the actual distances. The interhelix angle at the start of the simulation is indicated by a horizontal grey line.

Figure 6—source data 1. Numerical data plotted in Figure 6 and accompanying figure supplement.

Figure 6.

Figure 6—figure supplement 1. Close-up of hub interfaces from the molecular dynamics simulation of CaMKII-α.

Figure 6—figure supplement 1.

(A) Close-up of interface at which Docking 1 occurs, showing instantaneous snapshots from the end of the simulation overlaid with the start of the simulation, while aligning on the subunit to the right (subunit L). Subunit B (to the left) rotates away from subunit L over the course of the simulation. For clarity, only the hub domains are shown. (B) Hydrogen bond interactions between subunits B and L (same interface as (A)) at the start (top) and end (bottom) of the simulation. (C) Close-up of interface at which a regulatory segment does not dock, showing instantaneous snapshots from the end of the simulation overlaid with the start of the simulation, while aligning on the subunit to the right (subunit J). Subunit L (to the left) rotates toward from subunit J over the course of the simulation. For clarity, only the hub domains are shown. (D) Hydrogen bond interactions between subunits L and J (same interface as (C)) at the start (top) and end (bottom) of the simulation. The hydrogen bonds are intact in both the interface that opens and the interface that closes.
Figure 6—figure supplement 2. Individual subunits of the dodecameric hub are not very dynamic, while the interfaces between the subunits are intrinsically dynamic.

Figure 6—figure supplement 2.

(A) Mean probability distribution of r. m. s. deviation of the 12 hub domains from the simulation of the dodecamer with the regulatory segments and linkers present. The pale green shaded region indicates the standard deviation of the distribution of the r.m.s. deviation of the 12 hub domains. (B) Variation of the angle between the axes of αA helices of adjacent subunits in the simulation without the regulatory segments. The darker traces are the time-averaged values of the interhelix angles calculated using a moving window of 240 ns, while the lighter shades are the actual distances. In this simulation four interfaces in the lower ring open and the remaining two close (left), whereas in the upper ring two interfaces open and four close (right), indicating that the hub assembly is intrinsically highly dynamic. The angle between the axes of the αA helices of adjacent subunits from the crystal structure of an autoinhibited, dodecameric holoenzyme (PDB ID - 3SOA) (Chao et al., 2011), which is perfectly symmetrical, is indicated by the horizontal black line.
Figure 6—figure supplement 3. Model of regulatory segment docked onto the interface between subunits L and J, which closes over the course of the simulation.

Figure 6—figure supplement 3.

For clarity, only the hub domains on either side of the interface (in white) and the calmodulin-binding element (in pink, with surface representation), are shown. The calmodulin-binding element was taken from subunit L, whose calmodulin-binding element docks at the interface between subunits B and L (Figure 5—figure supplement 2), from a snapshot at the end of the simulation. The model was built by aligning the hub domain of subunit L from a snapshot from the end of the simulation, onto the hub domain of subunit J, also taken from a snapshot at the end of the simulation. This positions the regulatory segment so that it can form backbone interaction with the edge of the inter-facial β-sheet of subunit J, similar to Docking 1. However, there are numerous steric clashes with the αD helix of subunit L on the left, and this interface would have to undergo considerable distortions to accommodate binding by the calmodulin-binding element.