CaMKII autophosphorylation can occur between holoenzymes without subunit exchange

Abstract

The dodecameric protein kinase CaMKII is expressed throughout the body. The alpha isoform is responsible for synaptic plasticity and participates in memory through its phosphorylation of synaptic proteins. Its elaborate subunit organization and propensity for autophosphorylation allow it to preserve neuronal plasticity across space and time. The prevailing hypothesis for the spread of CaMKII activity, involving shuffling of subunits between activated and naive holoenzymes, is broadly termed subunit exchange. In contrast to the expectations of previous work, we found little evidence for subunit exchange upon activation, and no effect of restraining subunits to their parent holoenzymes. Rather, mass photometry, crosslinking mass spectrometry, single molecule TIRF microscopy and biochemical assays identify inter-holoenzyme phosphorylation (IHP) as the mechanism for spreading phosphorylation. The transient, activity-dependent formation of groups of holoenzymes is well suited to the speed of neuronal activity. Our results place fundamental limits on the activation mechanism of this kinase.

Keywords: CaMKII; E. coli; IHP; biochemistry; chemical biology; hek cells; holoenzyme; inter-holoenzyme phosphorylation; kinase signalling; neuroscience; unnatural amino acids.

Figure 1.. CaMKII^WT phosphorylates CaMKII^KD.

(A) Schematic representation of CaMKIIα domain arrangement (top). Yellow circles indicate phosphorylation sites. CaMKIIα holoenzyme structure (bottom, PDB:5u6y). (B) Cartoon representation of activation of CaMKII by calcium:calmodulin (top) and proposed mechanism for spread of kinase activity (bottom). Yellow circles indicate phosphorylation sites. (C) Schematic representation of experiment performed in panel (D). (D) Kinase activity of CaMKII^WT (10 nM) against CaMKII^KD (4 μM). Half-time of maximum phosphorylation (t_1⁄2=1.1 ± 0.2 min) determined by western blot, using an antibody against pT286. These data were fit with a single component, see Figure 1—figure supplement 3 for a fit to a repeated experiment with two components. The shading represents standard deviation (SD) from the mean value of three technical replicates. (E) Kinase activity of EGFP-CaMKII^WT and CaMKII^KD-AviTag in HEK cells and HEK cell lysates. ‘Basal’: activity in cells; ‘activated in vitro’: activity in cell lysates supplemented with 1 μM purified Ca²⁺:CaM and 0.5 mM ATP:Mg²⁺. Lane 1 – autophosphorylation of EGFP-CaMKII^WT in singly transfected cells, lane 2 – autophosphorylation of CaMKII^KD-AviTag in singly transfected cells, lane 3 – autophosphorylation of EGFP-CaMKII^WT and CaMKII^KD-AviTag in co-transfected cells, lane 4 – stimulated kinase activity of EGFP-CaMKII^WT from lysate in lane 1 and CaMKII^KD-AviTag from lysate in lane 2, incubated together at 4 °C for 1 min, lane 5 - stimulated kinase activity of EGFP-CaMKII^WT from lysate in lane 1 and CaMKII^KD-AviTag from lysate in lane 2, incubated together at 37 °C for 1 min, lane 6 - stimulated kinase activity of EGFP-CaMKII^WT from lysate in lane 1 and CaMKII^KD-AviTag from lysate in lane 2, incubated together at 37 °C for 10 min, lane 7 - stimulated kinase activity of EGFP-CaMKII^WT from lysate in lane 1 and 8 μM purified CaMKII^KD-AviTag, incubated together at 37 °C for 10 min, lane 8 - stimulated kinase activity of EGFP-CaMKII^WT from lysate in lane 1 37 °C for 10 min.

Figure 1—figure supplement 1.. Western blot detection of CaMKII^KD phosphorylation by CaMKII^WT.

(A) Blots of pT286 detection on CaMKII^KD (4 μM) after phosphorylation with CaMKII^WT (10 nM) in the presence of ATP:Mg²⁺ and low Ca²⁺:CaM concentrations (100 nM). Each time point is done in triplicate. (B) Blots of pT286 detection on CaMKII^KD (4 μM), after phosphorylation with CaMKII^WT (10 nM) in the presence of ATP:Mg²⁺ and high Ca²⁺:CaM concentrations (2 μM). Each time point is done in triplicate. (C) Western blot showing there is no phosphorylation signal from pT286 antibody on CaMKII^KD (4 μM) when CaMKII^WT is omitted from the kinase reaction. (D) Western blot showing that 10 nM CaMKII^WT (as included on the gels in panels A and B) gives no phosphorylation signal from the pT286 antibody. (E) Raw densitometry data from blots in panels A and B fitted with Langmuir function. Half-maximum times for CaMKII^KD phosphorylation by CaMKII^WT under low Ca²⁺:CaM conditions is t_1/2 = 0.9 ± 0.4 min, and under high Ca²⁺:CaM conditions is t_1/2 = 1.1 ± 0.2 min. The shading represents standard deviation (SD) from the mean value of three technical replicates.

Figure 1—figure supplement 2.. Radioactivity detection of CaMKII^KD phosphorylation by CaMKII^WT.

(A) Gels of CaMKII^KD phosphorylation detection in the presence of ATP:Mg²⁺ and low Ca²⁺:CaM concentrations (100 nM). Each time point is done in triplicates. (B) Gels of CaMKII^KD phosphorylation detection in the presence of ATP: Mg²⁺ and high Ca²⁺:CaM concentrations (2 μM). Each time point is done in triplicates. (C) Raw densitometry data from gels in A and B fitted with Langmuir function. Half-maximum times for CaMKII^KD phosphorylation by CaMKII^WT under low Ca²⁺:CaM conditions is t_1/2 = 1.16 ± 0.6 min, and under high Ca²⁺:CaM conditions is t_1/2 = 4.3 ± 1.13 min. The shading represents standard deviation (SD) from the mean value of three technical replicates.

Figure 1—figure supplement 3.. Phosphorylation of CaMKII^KD by CaMKII^WT is concentration-dependent.

(A) Graph showing dependence of CaMKII^KD phosphorylation on CaMKII^WT concentration on a linear timescale. Each fit used the two-component exponential function (see Materials and methods for details). In these ‘free fits’. the rates r₁ and k₂ were allowed to vary between the four fits (10 free parameters in all, including the amplitudes). Note that we repeated the experiments at 10 nM CaMKII^WT (essentially the same as Figure 1D). Because the phosphorylation reaction was so fast at 100 nM CamKII^WT, we fitted only a single component (phosphorylation rate was 2.5 min^–1). Fit parameters (±1 SD) were as follows: 10 nM fit A₁=0.27 ± 0.06; k₂=0.043 ± 0.006 nM^–1 min^–1; r₁ = 0.035 ± 0.009 min^–12 nM fit A₁=0.6 ± 0.1; k₂=0.007 ± 0.003 nM^–1 min^–1; r₁ = 0.44 ± 0.13 min^–10.5 nM fit A₁=0.5 ± 0.3; k₂=0.02 ± 0.03 nM^–1 min^–1; r₁ = 0.17 ± 0.11 min^–1 Same free fit as panel (B) on a logarithmic timescale to reveal the two components. (C) Global fit (5 free parameters) using common values of rates r₁ and k₂ for each curve. r₁=0.41 min^–1 and k₂=0.017 nM^–1 min^–1. For the 100 nM fit, the amplitude A₁ was set to zero. Note that at 100 nM, the first order rate k₂ * 100 nM is about 3 x faster (1.7 min^–1) than the rate estimated for concentration-independent phosphorylation. (D) Representative western blots of the time course of phosphorylation on T286 of CaMKII^KD by CaMKII^WT at different concentrations (concentrations of proteins used is indicated on the left side of each panel). Lanes which are marked with ‘kinase-dead (1 hr)' show that no pT286 signal can be detected on CaMKII^KD (4 µM) after 1 hr of incubation without CaMKII^WT. Lanes labeled with concentration of CaMKII^WT demonstrate that the wild-type protein is invisible for anti-pT286 antibody at the reaction amounts loaded on these gels for all except for 100 nM WT lane, where a phosphorylation signal is detected after 1 h of incubation of CaMKII^WT with activation stimuli, but without CaMKII^KD. However, the signal is about 20% of the signal detected on CaMKII^KD after 1 hr. The bottom blot shows pT286 detection of triplicate reactions at 1, 2, 5 and a single reaction at 10 min. The other blots used to construct the curves in (A–C) are in Figure 1—figure supplements 4–7.

Figure 1—figure supplement 4.. Phosphorylation of CaMKII^KD by 0.5 nM CaMKII^WT.

(A) Replicate 1 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 0.5 nM CaMKII^WT. Left panel is showing detection of CaMKII by anti-pan CaMKII antibody. Middle panel is showing detection of pT286 signal on CaMKII by anti-pT286 antibody. Right panel is showing merged signal of pan-CaMKII and pT286 signal. Lane labeled with ‘kinase-dead’ is showing that although CaMKII^KD can be detected by pan antibody, the signal on T286 (middle panel) is absent after 60 min of the kinase reaction, when CaMKII^WT is excluded from the reaction. Lane labeled with ‘0.5 nM WT’ demonstrates that the wild-type protein is invisible for anti-pan and anti-pT286 antibodies after 60 min of kinase reaction without CaMKII^KD at the protein amount loaded on this gel. (B) Replicate 2 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 0.5 nM CaMKII^WT. The labeling of the blots is the same like in (A). (C) Replicate 3 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 0.5 nM CaMKII^WT. The labeling of the blots is the same like in (A) and (B).

Figure 1—figure supplement 5.. Phosphorylation of CaMKII^KD by 2 nM CaMKII^WT.

(A) Replicate 1 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 2 nM CaMKII^WT. Left panel shows detection of CaMKII by anti-pan CaMKII antibody. Middle panel shows detection of pT286 signal on CaMKII by anti-pT286 antibody. Right panel shows merged signal of pan-CaMKII and pT286 signal. Lane labeled with ‘kinase-dead’ shows that although CaMKII^KD can be detected by pan antibody, the signal on T286 (middle panel) is absent after 60 min of the kinase reaction, when CaMKII^WT is excluded from the reaction. Lane labeled with ‘2 nM WT’ demonstrates that the wild-type protein is invisible for anti-pan and anti-pT286 antibodies after 60 min of kinase reaction without CaMKII^KD at the protein amount loaded on this gel. Lane labeled with ‘100 nM WT’ demonstrates that the wild-type protein is somewhat visible for anti-pan and anti-pT286 antibodies after 60 min of kinase reaction without CaMKII^KD at the protein amount loaded on this gel, but this signal represents about 20% of the pT286 signal coming from reactions incubated with CaMKII^KD. (B) Replicate 2 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 2 nM CaMKII^WT. The labeling of the blots is the same as in (A). (C) Replicate 3 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 2 nM CaMKII^WT. The labeling of the blots is the same as in (A) and (B).

Figure 1—figure supplement 6.. Phosphorylation of CaMKII^KD by 10 nM CaMKII^WT.

(A) Replicate 1 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 10 nM CaMKII^WT. Left panel shows detection of CaMKII by anti-pan CaMKII antibody. Middle panel shows detection of pT286 signal on CaMKII by anti-pT286 antibody. Right panel shows merged signal of pan-CaMKII and pT286 signal. Lane labeled with ‘kinase-dead’ is showing that although CaMKII^KD can be detected by pan antibody, the signal on T286 (middle panel) is absent after 60 min of the kinase reaction, when CaMKII^WT is excluded from the reaction. Lane labeled with ‘10 nM WT’ demonstrates that the wild-type protein is invisible for anti-pan and anti-pT286 antibodies after 60 min of kinase reaction without CaMKII^KD at the protein amount loaded on this gel. (B) Replicate 2 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 10 nM CaMKII^WT. The labeling of the blots is the same as in (A). (C) Replicate 3 showing western blot detection of phosphorylation development on T286 of CaMKII^KD over the course of 60 min catalyzed by 10 nM CaMKII^WT. The labeling of the blots is the same as in (A) and (B).

Figure 1—figure supplement 7.. Phosphorylation of CaMKII^KD by 100 nM CaMKII^WT.

(A) Western blot detection of phosphorylation development on T286 over time when CaMKII^KD is incubated with 100 nM CaMKII^WT. Reaction at each time point is done in triplicate. The triplicates for 1, 2, and 5 min were loaded one after another on the same gel. Lane labeled with ‘10’ is a single reaction at 10 min time point. Lane labeled with ‘kinase-dead’ is showing that although CaMKII^KD can be detected by pan antibody, the signal on T286 (middle panel) is absent after 60 min of the kinase reaction, when CaMKII^WT is excluded from the reaction. (B) Western blot detection of phosphorylation development on T286 over time when CaMKII^KD is incubated with 100 nM CaMKII^WT. Reaction at each time point is done in triplicate. The duplicate for 10 min time point and then triplicates for 15, 30, and 60 min were loaded one after another on the same gel.

Figure 2.. Crosslinking CaMKII subunits in the hub domain does not change the rates of trans-autophosphorylation.

(A) Position of F394 residue (purple) in the hub domain of CaMKIIα (PDB: 5u6y), showing orientation towards the interface between adjacent hub domains within one hub ring. (B) Coomassie stained gel of UV-induced crosslinking of CaMKII^F394BzF (right) to higher order oligomers, and CaMKII^WT (left) as a control. Purple ramp indicates increasing UV exposure. (C) Kinase activity of UV treated (+UV) and untreated (-UV) CaMKII^F394BzF against CaMKII^KD - AviTag. Phosphorylation was measured as incorporation of radioactive AT³²P at phosphorylation sites on CaMKII^KD -AviTag (pCaMKII). Half-times of maximum phosphorylation are determined to t1⁄2=2.6 ± 0.3 min for untreated and t1⁄2=2 ± 0.3 min for UV treated CaMKII^F394BzF. The shading represents standard deviation (SD) from the mean value of three technical replicates.

Figure 2—figure supplement 1.. Properties of CaMKII^F394BzF: kinase activity and size exclusion chromatography.

(A) Gels of CaMKII^KD phosphorylation detection by UV treated CaMKII^F394BzF. Each time point is done in triplicates. These data are used for Figure 2C. (B) Gels of CaMKII^KD phosphorylation detection by untreated CaMKII^F394BzF. Each time point is done in triplicates. These data are used for Figure 2C. (C) Size exclusion chromatography (Superose 6 10/300 increase column) of CaMKII wild-type and mutants used in this study. Dashed lines indicate the peak elution volume for wild-type (V_e = 13.8 mL).

Figure 2—figure supplement 2.. Hub domain mutant CaMKII^H418BzF phosphorylates CaMKII^KD irrespective of crosslinking.

(A) Coomassie stained gel showing UV-dependent oligomerization of CaMKII^H418BzF. (B) Phosphorylation of CaMKII^KD by UV-treated or -untreated CaMKII^H418BzF. Langmuir fit determined half-maximum times of phosphorylation t_1⁄2 = 7.2 ± 0.8 min for untreated CaMKII^H418BzF and t_1⁄2 = 6.5 ± 1.5 min for UV-treated CaMKII^H418BzF.

Figure 2—figure supplement 3.. Radioactivity detection of substrate phosphorylation by CaMKII^F394BzF.

(A) Gels of CaMKII substrate (GST-Syn) phosphorylation by UV-treated CaMKII^F394BzF. Each time point is done in triplicate. (B) Gels of CaMKII substrate (GST-Syn) phosphorylation by control CaMKII^F394BzF (no UV treatment). Each time point is done in triplicate. (C) Phosphorylation of GST-Syn by UV-treated or -untreated CaMKII^F394BzF. Langmuir fit determined half-maximum times of GST-Syn phosphorylation t_1⁄2 = 8.7 ± 2.4 min for untreated CaMKII^F394BzF and t_1⁄2 = 8.8 ± 1.2 min for UV-treated CaMKIIF394BzF. The shading represents standard deviation (SD) from the mean value of three technical replicates.

Figure 3.. CaMKII holoenzymes do not mix during activation.

(A) Schematic representation of the experiment performed in panel (B) and possible outcomes. (B) Western blot detection of potential CaMKII^KD-AviTag incorporation in CaMKII^F394BzF holoenzymes. Lane 1 – CaMKII^F394BzF, lane 2 - CaMKII^F394BzF treated with UV, lane 3 - CaMKII^KD-AviTag, lane 4 - CaMKII^KD-AviTag treated with UV, lane 5 - CaMKII^F394BzF incubated with CaMKII^KD-AviTag, lane 6 - CaMKII^F394BzF incubated with CaMKII^KD-AviTag, then UV treated, lane 7 - CaMKII^F394BzF incubated with CaMKII^KD-AviTag and activation stimuli (Ca²⁺:CaM and Mg²⁺:ATP), lane 8 - CaMKII^F394BzF incubated with CaMKII^KD-AviTag and activation stimuli (Ca²⁺:CaM and Mg²⁺:ATP), then UV treated, lane 9 - CaMKII^KD-AviTag incubated with activation stimuli (Ca²⁺:CaM and Mg²⁺:ATP), lane 10 - CaMKII^KD-AviTag incubated with activation stimuli (Ca²⁺:CaM and Mg²⁺:ATP), then UV treated. Blue arrowheads indicate nonspecific, UV-induced, dimerization of CaMKII^KD-AviTag, independent of BzF.

Figure 3—figure supplement 1.. CaMKII holoenzymes do not mix during activation.

(A) Coomassie stained gel of CaMKII^F394BzF and CaMKII^KD crosslinking used in Figure 3B. (B) Western blot detection of potential CaMKII^WT-AviTag incorporation in CaMKII^F394BzF holoenzymes. Blue arrowhead indicates a UV-induced, BzF-independent dimeric band of CaMKII^WT in the presence of 50 mM TCEP. (C) Crosslinking of CaMKII^F394BzF and CaMKII^WT-AviTag in the presence of 1 mM TCEP. Magenta arrowheads in lanes 3 and 4 indicate UV-induced, BzF-independent dimeric bands of CaMKII^WT, which are more abundant here than under conditions in panel B.

Figure 4.. Crosslinking mass spectrometry reveals inter-holoenzyme kinase domain contacts during activation.

(A) Schematic representation of crosslinking experiments and expected outcomes. In the case of subunit exchange, a flat profile of intersubunit crosslinks is expected, whereas for inter-holoenzyme phosphorylation, a bias towards kinase domain crosslinks is expected. (B) Heterotypic interactions plotted as R values of two independent replicates in basal and activating conditions. Incubation time was 30 min, prior to addition of DSS crosslinker. (C) As for panel (B) but with incubation time of 150 min prior to addition of DSS crosslinker. (D) Heat map indicating the number of homotypic (upper) and heterotypic (lower) crosslinks by domain under basal and activating conditions, after 30 and 150 min of incubation. Only peptides that were identified in both samples for each condition were counted in the heat map.

Figure 4—figure supplement 1.. XL-MS identifies interactions of CaMKII holoenzymes.

(A) Zoomed-in exemplary MS1 spectrum showing uni-isotopic (homotypic) and mixed-isotopic (heterotypic) crosslinks. The ratio of mixed-isotopic and uni-isotopic crosslinks is calculated using the intensities of the highest intense peak in each isotopic distribution. (B) The MS1 intensity distribution of the most intense peak of all four isotopic distributions (i.e. ¹⁴N¹⁴N, ¹⁴N¹⁵N, ¹⁵N¹⁴N, ¹⁵N¹⁵N) showing good reproducibility over replicates.

Figure 4—figure supplement 2.. Crosslinking sites involving pT286 peptides from mixed isotypes.

Kinase domain (from the 5u6y PDB structure) with regulatory domain (orange) docked. 6 Lysine residues that gave heterotypic crosslinked peptides including P-Thr286 (DSS link to Lys291) are indicated as spheres. The detection ratio (heterotypic to homotypic mass spectra intensity) for each crosslinked interaction is indicated.

Figure 5.. Mapping crosslinks onto holoenzyme structure.

(A) Holoenzyme structure with two neighboring subunits (green and purple) indicated (PDB: 5u6y). Hub domain is in light blue, regulatory segment (docked) in orange. (B) Basal crosslinks between homo-isotypes (30 min incubation, 136 crosslinks as dashed yellow lines) plotted onto the holoenzyme as intersubunit interactions. Inset shows the 5 homo-isotopic crosslinks found for interactions between neighboring hub domains. (C) Crosslinks between homo-isotypes in activated conditions (30 min, 85 crosslinks as pink dashed lines) showed a similar pattern to the basal condition. (D) Two holoenzymes arranged to allow kinase-kinase contacts to form (between green and purple subunits). (E) Sparse heterotypic crosslinks (8 in total, blue dashed lines) in the basal condition after 150 min. All heterotypic crosslinks involve the kinase domain. (F) In activating conditions, 15 heterotypic kinase-kinase interactions were detected after 30 min (out of 20 total). Over 150 min activation, 32 heterotypic kinase-kinase interactions were found (out of 43 total).

Figure 5—figure supplement 1.. Hypothetical close mode of holoenzyme interaction from MS crosslinks.

Heterotypic DSS crosslinks between kinase domains have plausible lengths (less than 20 Å) when one kinase domain from a holoenzyme is interdigitated into the structure of another holoenzyme. Hub domains are in light blue, regulatory domains (docked) in orange. Docking was done by hand to minimize clashes (see text for details).

Figure 6.. Reversible activity-dependent colocalization of CaMKII holoenzymes.

(A) Schematic representation of CaMKII in vitro enzymatic labeling with maleimide dyes and biotin, and TIRF experimental set-up. (B) Representative TIRF images of unactivated (basal) CaMKII^WT sample (left), activated CaMKII^WT sample (middle), and CaMKII^WT sample first activated and then quenched with EDTA (right). Raw, unprocessed, images are shown. (C) Summary of colocalization analysis (Pearson coefficient) for TIRF images for CaMKII under different conditions. Statistical significance was calculated using a multi-comparison test (Dunnett test) with α=0.05. Mean and standard deviation of the mean are indicated. Each condition was done at least two times on different days, with different protein preps. Five ROIs coming from different slides were selected for analysis for each condition. Scale bar: 10 µm.

Figure 6—figure supplement 1.. Activation stimuli are necessary for CaMKII^WT holoenzyme colocalization.

(A) Absence of colocalization between differently labeled CaMKII^WT holoenzymes in the absence of Mg²⁺:ATP. (B) Absence of colocalization between differently labeled CaMKII^WT holoenzymes upon chelation of Ca²⁺ by BAPTA. (C) Absence of colocalization between differently labeled CaMKII^WT holoenzymes in the absence of CaM. (D) Absence of colocalization between differently labeled CaMKII^WT holoenzymes incubated first in activating conditions, and following later chelation of Mg²⁺ by EDTA. Scale bar: 10 µm.

Figure 6—figure supplement 2.. CaMKII^KD holoenzymes cluster only in the presence of activated CaMKII^WT holoenzymes.

(A) Absence of colocalization between CaMKII^KD and CaMKII^WT in basal conditions. (B) Colocalization of CaMKII^KD and CaMKII^WT upon activation. (C) Absence of colocalization between differently labeled CaMKII^KD in basal conditions. (D) Absence of colocalization between differently labeled CaMKII^KD in activating conditions. (E) Summary of colocalization analysis. Probability of no difference from Dunnett’s multiple comparison test. Scale bar: 10 µm.

Figure 7.. Mass photometry detects clusters of holoenzymes forming upon activation.

(A) Mass distribution of 400 nM CaMKII^WT under basal conditions (red curve). Blue curve is multi- Gaussian fit (shown separately on lower graph). Red curve in upper graph is the fit residual. (B) Mass distribution of 400 nM CaMKII^WT under activating conditions (green curve), with multi- Gaussian fit in blue (peaks numbered in lower graph). (C) Mass distribution of 100 nM CaMKII^WT under basal conditions (red curve). (D) Mass distribution of 100 nM CaMKII^WT under activating conditions (green curve). (E) Mass distribution of 10 nM CaMKII^WT under basal conditions (red curve). (F) Mass distribution of 10 nM CaMKII^WT under activating conditions (green curve). Inset shows fit on expanded scale.

Figure 7—figure supplement 1.. CaMKII^WT forms higher order clusters during activation.

(A) Two replicates of the particle mass distribution of 400 nM CaMKII^WT under basal conditions (red curve). Blue curve is multi-Gaussian fit (shown separately on lower graph). Red curve in upper graph is the fit residual. (B) Two replicates of the particle mass distribution of 400 nM CaMKII^WT under activating conditions (green curve). (C) Two replicates of the particle mass distribution of 100 nM CaMKII^WT under basal conditions (red curve). (D) Two replicates of the particle mass distribution of 100 nM CaMKII^WT under activating conditions (green curve).

Figure 7—figure supplement 2.. CaMKII^KD fails to form higher order clusters during activation.

(A) Particle mass distribution of 400 nM CaMKII^KD under basal conditions. (B) Particle mass distribution of 400 nM CaMKII^KD under activating conditions. (C) Western blot detection of pT286 on CaMKII^KD-AviTag. Left blot – detection of total CaMKII, middle blot – detection of phosphorylation on T286, right blot – merged signals of total CaMKII and pT286. Lane 1–4 µM CaMKII^KD-AviTag, 10 nM CaMKII^WT, 2 µM Ca²⁺CaM, 100 µM ATP:Mg²⁺; Lane 2–4 µM CaMKII^KD-AviTag, 10 nM CaMKII^WT, 100 µM ATP:Mg²⁺; Lane 3–4 µM CaMKII^KD-AviTag, 10 nM CaMKII^WT, 2 µM Ca²⁺CaM; Lane 4–4 µM CaMKII^KD-AviTag, 10 nM CaMKII^WT; Lane 5–4 µM CaMKII^KD-AviTag, 2 µM Ca²⁺CaM, 100 µM ATP:Mg²⁺.