Oxidation Driven Reversal of PIP2-dependent Gating in GIRK2 Channels

Figures

• 

  Figure 1.. Strategy to test redox effect on GIRK2 activity in vitro. (A) Protein purification and in vitro flux assay scheme to tightly control the oxidation state of the purified proteins with details in “Method” section. Proteins purified without 2 mM TCEP in the oxidizing condition were subjected to experiments to test different redox conditions while protein purified with 2 mM TCEP in the reducing condition was kept in a strictly reducing condition with additional 2 mM TCEP during 3-hour RT incubation. Inset diagram describing the steps for proteoliposome formation. (B) ACMA flux assay strategies to test K+ specific activity in the absence or presence of Na+ by replacing external salt from NMDG Cl to NaCl or to test Na+ specific activity by replacing internal salt from KCl to NaCl while keeping the external salt with NMDG Cl. Lipids forming the proteoliposomes shown in yellow, reconstituted proteins in gray blobs, proton ionophore CCCP in yellow blob. Curved arrows indicate ion or proton movements down the electrochemical gradients, and black straight arrows show the movement of ACMA across the liposome membrane with the red X mark indicating the inability of ACMA to cross the membrane…

• 

  Figure 2.. Canonical GIRK2 activity in reducing conditions. ACMA flux assay of reduced GIRK2 in the presence of different combinations of agonists: no agonist (cyan), PIP2 (blue), PIP2 + Na+ (green), PIP2 + Gβγ (pink), and PIP2 + Na++Gβγ (purple). The details are in the “Method” section. (A) Representative traces of normalized fluorescence intensity for empty liposomes without proteins (dashed lines) or proteoliposomes (solid lines with marks) in the presence of different combinations of agonists. CCCP was added at 0 min, and potassium ionophore valinomycin (Val) was added at 15 min. The channel activity was inferred (C) from the differential ACMA signal decay (ΔF) between the empty liposomes and the proteoliposomes (B) at 10 min after the addition of proton ionophore (CCCP). Data are presented as the mean ± SEM n = 5, and One way ANOVA followed by Tukey’s multiple comparison was used for the statistical test. No agonist vs. PIP2 + Na, P = .0052; no agonist vs. PIP2 + Na++Gβγ, P = .0164; PIP2 vs. PIP2 + Na, P = .0109; PIP2 vs. PIP2 + Na++Gβγ, P = .0081; PIP2 + Gβγ vs. PIP2 + Na++Gβγ, P = .0057. Representative differential ACMA signal decay (ΔF) with increasing amount of the proteins in the…

• 

  Figure 3.. Anomalous GIRK2 behavior in oxidizing conditions. (A) Differential ACMA signal decay (ΔF) by GIRK2 purified without reducing reagents and incubated in reducing (+TCEP; left two columns) or oxidizing (-TCEP; right two columns) conditions for 30 min (Top) or 3 h (Bottom) at RT. Assays performed with external NMDG or Na+ as indicated. (B) Activities either in reducing (black) or oxidizing (red) conditions are overlaid for 30-min (Left) or 3-hour (Right) incubated samples. Data are presented as the mean ± SEM n = 5 in each case. Two-way ANOVA with Sidak’s multiple comparisons test was used for the statistical test. Asterisks at the top of the figure indicate statistical significance between conditions as indicated. The total of 30 min + 2TCEP: no agonist vs. PIP2 + Na+, P = .0004; Na+ vs. PIP2 + Na+, P < .0001. 3 h + 2TCEP: no agonist vs. PIP2 + Na+, P < .0001; Na+ vs. PIP2 + Na+, P < .0001. The total of 3 h + 0TCEP: no agonist vs. PIP2, P < .001, no agonist vs. PIP2 + Na+, P < .0001; Na+ vs. PIP2 + Na+, P < .0001. The vertically aligned asterisks indicate statistical significance for the given condition between reducing and oxidizing conditions. The total of 30 min: no agonist, P <…

• 

  Figure 4.. Nonspecific inhibition of the anomalous activity by PIPs. Activities in the presence of 1% (A) or 3% (B) PI(4)P, substituting equal amounts of PIP2 after 3-hour RT incubation are overlaid on the Control with PIP2, in reducing (2 mM TCEP) or oxidizing conditions, as indicated. The details are in the “Method” section. The same plotting scheme used in Figure 3C was applied. Data are presented as the mean ± SEM n = 3 experiments in each condition. Two-way ANOVA with Sidak’s multiple comparisons test was used for the statistical test. Asterisks at the top of the figure indicate statistical significance between conditions as indicated. 1PIP2 with 2 TCEP: no agonist vs. PIP2 + Na+, P < .0001; Na+ vs. PIP2 + Na+, P < .0001. 1PIP2 with 0 TCEP: no agonist vs. PIP2, P < .0001, no agonist vs. PIP2 + Na+, P < .0001; Na+ vs. PIP2 + Na+, P = .0007. 1PI(4)P with 0 TCEP: no agonist vs. PIP2, P = .0009, no agonist vs. PIP2 + Na+, P = .0001; Na+ vs. PIP2 + Na+, P = .0050. 3PIP2 with 2 TCEP: no agonist vs. PIP2 + Na+, P < .0001; Na+ vs. PIP2 + Na+, P < .0001. 3PIP2 with 0 TCEP: no agonist vs. PIP2, P < .0001, no agonist vs. PIP2 + Na+, P < .0001; Na+ vs. PIP2 + Na+, P < .0001. 3PI(4)P with 0 TCEP:…

• 

  Figure 5.. Cysteine knockout mutants mimicking anomalous activities. (A) Multiple Kir channel sequence alignment of the regions encompassing the first two cysteine residues of GIRK2. The conserved C65 and C190 equivalent residues are shown in orange and blue, respectively. The residues interacting with PIP2 are shown in green. (B) Ribbon diagram of two opposing subunits of GIRK2 with zoomed in images showing the first two cysteine residues; C65 in the orange box and C190 in the cyan box. The side chains and PIP2 are shown in sticks, Na+ ion is in green sphere. (C) Cys knockout mutant protein activities in the presence of different combinations of agonists, in reducing and oxidizing conditions. The same plotting scheme used in Figure 3B was applied. The graph for the wild type is the copy of Figure 3B 3-hour condition. Data are presented as the mean ± SEM n = 3 for all mutants. Two-way ANOVA with Sidak’s multiple comparisons test was used for the statistical test. Asterisks at the top of the figure indicate statistical significance between conditions as indicated. C221TC321V with 2 TCEP: no agonist vs. PIP2 + Na+, P = .0356; Na+ vs. PIP2 + Na+, P = .0214. C65V with 0 TCEP: no agonist vs.…

• 

  Figure 6.. Oxidation dependent changes in cKir2.2 activity. (A) Differential ACMA signal decay (ΔF) due to activity of cKir2.2 purified without reducing reagents and incubated in reducing (+TCEP; Left two columns) or oxidizing (−TCEP; Right two columns) conditions for 30 min (Top) or 3 h (Bottom) at RT. (B) The activities in either reducing (black) or oxidizing (red) conditions are overlaid for 30-min (Left) or 3-hour (Right) incubated samples. Data are presented as the mean ± SEM n = 3 for all mutants. Two-way ANOVA with Sidak’s multiple comparisons test was used for the statistical test. Asterisks at the top of the figure indicate statistical significance between conditions as indicated. The total of 30 min with 2 TCEP: no agonist vs. PIP2, P = .0005; no agonist vs. PIP2 + Na+, P = .0003; Na+ vs. PIP2 + Na+, P = .0002. The total of 3 h with 2 TCEP: no agonist vs. PIP2, P < .0001; no agonist vs. PIP2 + Na+, P < .0001; Na+ vs. PIP2 + Na+, P < .0001. The total of 3 h with 0 TCEP: no agonist vs. PIP2 + Na+, P = .0081. The vertically aligned asterisks indicate statistical significance for the given condition reducing and oxidizing conditions. The total of 30 min: no agonist, P = .0082; PIP2, P…

• 

  Figure 7.. Local GIRK2 structures near C65 and C190. Three crystal and a cryo EM structures of GIRK2 wildtype are shown in ribbon diagrams and colored in blue for 3SYO in complex with Na+, green for 3SYA in complex with PIP2 + Na+, and yellow for 6XIT in complex with PIP2, and orange for 4KFM in complex with PIP2 + Na+ + Gβγ. (A) The region near C65 residue is zoomed in, Na+ ions are shown in spheres, and the secondary structures of neighboring parts are shown for comparison. (B) The region near C190 residue is zoomed in, and C190 and the residues close to C190 side chain are shown in sticks. The distance between the thiol group of C190 side chain and the nitrogen atom of V87 amine group is measured.

• 

  Figure 8.. Proposed GIRK2 gating model. (A) CTD conformational changes induced in the presence of different ligands. (Left) Overlayed ribbon diagrams (viewed through the membrane looking into the cytoplasm) of two opposing CTDs of GIRK2 cryo EM structure in complex with PIP2 (PIP2-bound: 6XIT) in green and of Kir6.2 C166S/G334D cryo EM structure in the absence of PIP2 (PIP2-unbound:7S5T). (Right) Overlayed ribbon diagrams of the same two opposing CTDs of GIRK2 in complex with PIP2 (PIP2-bound: 6XIT) in green or in complex with PIP2, Na+, and Gβγ (Holo:4KFM) in orange. The arrows indicate rotational/twisting motions of the CTD induce by binding of either PIP2 alone (Left) or PIP2 plus Na+ and Gβγ (Right); clockwise motion narrows but counterclockwise motion widens the G-loop gate. (B) GIRK2 gating models. Top: canonical GIRK2 gating in the reduced state. In the Apo state, the HBC gate is closed while the G-loop gate is open. PIP2 opens the HBC gate but the induced clockwise rotational/twisting motions of the CTD close the G-loop gate. The additional binding of positive cofactors, Na+ and/or Gβγ, makes the channel conductive by triggering counterclockwise rotational/twisting motions in the…