Molecular and cellular correlates in Kv channel clustering: entropy-based regulation of cluster ion channel density

Figures

• 

  Figure 1. A ‘ball and chain’ mechanism for Kv channel clustering? (a) Schematic representation of the ‘ball and chain’ mechanism for Kv channel binding to the PSD-95 scaffold protein (for clarity, only two channel subunits are presented). In this mechanism, the intrinsically disordered 'chain' at the Kv channel C-terminal binds PSD-95 upon interaction of the ‘ball’ PDZ binding motif to PSD-95 PDZ domain(s). This mechanism is reminiscent of the mechanism underlying fast channel inactivation. In this former mechanism, the length of the Kv channel C terminal 'chain' governs its interaction with PSD-95 in a manner which is entropy-controlled. Given the ability of PSD-95 to aggregate and the stoichiometry of the interaction, channel clustering at unique site results (b). Whether or not Kv channel 'chain' length regulates channel density within clustering sites was addressed in the current study (c). The rectangular shape corresponds to the assembly T1 domain of the channel while the membrane-embedded portion corresponds to the channel voltage-sensor and pore domains. The crescent, box and rectangular shapes represent the PDZ, SH3 and guanylate kinase-like domains of the PSD-95 protein,…

• 

  Figure 2. Cell expression and clustering patterns of the native A and B Shaker Kv channel variants. (a) High-resolution confocal microscopy images of SH-SY5Y neuroblastoma cells expressing either the native Shaker A or B Kv channel variants alone or the PSD-95-GFP scaffold protein partner. Typical images of cells co-expressing either the A or B native channel variants and PSD-95-GFP are shown in the upper and lower sections of (b). Three images are shown for each cell, with the red channel-associated fluorescent signal shown in the left column, the green PSD-95-associated signal presented in the middle columns, and the merged shown in the right column. (c) Typical images of SH-SY5Y cells co-expressing PSD-95-GFP and either the Shaker A or B native Kv channel variants lacking the terminal 'ball' PDZ binding motif (∆'ball'). Scale bars in panels (a) to (c) correspond to 10 μm. Numbers next to each channel notation indicate C-terminal amino acid ‘chain’ length.

• 

  Figure 3. The native ‘chain’ length channel variants exhibit differences in clustering attributes. (a) Comparison of mean expression levels (reflected in the integrated cell fluorescence signal and normalized to cell area) of the A and B Kv channel variants in the absence (left) or presence (middle) of PSD-95. Expression levels of A and B mutants lacking the 'ball’ motif, along with PSD-95, are shown in the right panel. (b) Comparison of the number of clusters per cell (normalized to cell area) of the A and B Kv channel variants. (c) Comparison of the mean fraction of A and B Kv channel variants expressed in clustering sites. (d) Cluster Kv channel density distributions of the A and B variants (evaluated by cluster signal intensity), as supported by PSD-95. Distributions are in steps of 2 (× 105) IFSc/μm2 are presented (n = 4,000–6,000 clusters). (e) Comparison of the mean value of PSD-95-mediated A or B cluster signal intensity.

• 

  Figure 4. Cell expression and clustering of all Shaker Kv channel 'chain' length variants. Typical high-resolution confocal microscopy images of cells co-expressing PSD-95-GFP and either native or modified Kv channel 'chain'-length variants presenting a length series of 9, 24, 39, 50, 61, 88, 104, 106 and 144 amino acids. For each cell, only the merged yellow co-localization image is shown (See Supplementary Fig. S2 for all typical red (channel-associated), green (PSD-95-associated) and yellow signal (merged) images). Scale bars correspond to 10 μm. Numbers next to each channel notation indicate C-terminal 'chain' length.

• 

  Figure 5. ‘Chain’-length modulation of cluster Kv channel density. (a) Cluster Kv channel density distribution for variants presenting chain lengths of 24, 61 and 144 amino acids, as evaluated by cluster signal intensity and supported by PSD-95. Distributions in steps of 2 (× 105) IFSc/μm2 are presented (n = 350–7,000 clusters). The mean values of cluster ion channel density for the different variants are compared in the panel to the right of the distributions panel. All differences were found to be statistically significant, based on ANOVA analysis (the p-values for all three comparisons are smaller than 0.00001). (b) Comparison of cluster Kv channel density distribution of the wild type or reversed C terminal ‘chain’ Shaker B channels exhibiting identical (144 amino acid long) C-terminal chain lengths and composition (n = 3,000–4,000). The mean values of cluster ion channel density for the two variants are compared in the panel to the right. No significant difference was observed between the two variants (p-value = 0.09). (c) Dependence of the mean value of PSD-95-mediated cluster Kv channel density on C-terminal ‘chain’ length. Differences in cluster ion channel density were all found…

• 

  Figure 6. Entropy-based regulation of cluster Kv channel density. (a) Correlation plot relating the mean cluster ion channel density and PSD-95 binding affinity of the different Kv channel ‘chain’-length variants. The solid curve represents a linear regression between the compared quantities (R2 = 0.96) observed for Kv channel variants presenting chain lengths of 61 amino acids or higher. (b) Dependence of the mean channel expression levels of the different Kv channel variants on Kv channel-PSD-95 affinity (expressed as binding energy). The number next to each data point corresponds to C-terminal ‘chain’ length.