Investigations of the contribution of a putative glycine hinge to ryanodine receptor channel gating

Abstract

Ryanodine receptor channels (RyR) are key components of striated muscle excitation-contraction coupling, and alterations in their function underlie both inherited and acquired disease. A full understanding of the disease process will require a detailed knowledge of the mechanisms and structures involved in RyR function. Unfortunately, high-resolution structural data, such as exist for K(+)-selective channels, are not available for RyR. In the absence of these data, we have used modeling to identify similarities in the structural elements of K(+) channel pore-forming regions and postulated equivalent regions of RyR. This has identified a sequence of residues in the cytosolic cavity-lining transmembrane helix of RyR (G(4864)LIIDA(4869) in RyR2) analogous to the glycine hinge motif present in many K(+) channels. Gating in these K(+) channels can be disrupted by substitution of residues for the hinge glycine. We investigated the involvement of glycine 4864 in RyR2 gating by monitoring properties of recombinant human RyR2 channels in which this glycine is replaced by residues that alter gating in K(+) channels. Our data demonstrate that introducing alanine at position 4864 produces no significant change in RyR2 function. In contrast, function is altered when glycine 4864 is replaced by either valine or proline, the former preventing channel opening and the latter modifying both ion translocation and gating. Our studies reveal novel information on the structural basis of RyR gating, identifying both similarities with, and differences from, K(+) channels. Glycine 4864 is not absolutely required for channel gating, but some flexibility at this point in the cavity-lining transmembrane helix is necessary for normal RyR function.

Keywords: Calcium Channels; Calcium Intracellular Release; Glycine Hinge; Ion Channels; Ryanodine Receptor; Sarcoplasmic Reticulum (SR); Single Channel Gating.

FIGURE 1.

Comparison of KcsA and the RyR2 analogy model and sequence alignment of the inner helices of a range of K⁺ channels with the human RyR2 sequence. A, PyMOL (34) view of the RyR2 PFR. The panel depicts a comparison of the structural elements of KcsA (panel i) and our RyR2 pore analogy model (panel ii), constructed using KcsA as a template. Two monomers are shown for clarity. The main chain polypeptide backbone is colored in cyan, whereas the selectivity filter is highlighted in magenta. The luminal side of the pore appears at the top. B, the glycine residue of the conserved hinge motif is highlighted, in all cases, in bold red with the other motif residues in red. Additional residues and motifs involved in gating are highlighted in bold blue.

FIGURE 2.

Expression of hRyR2 proteins in HEK293 cells and Western blot analysis. A, the endogenous fluorescence of eGFP-tagged hRyR2 (right panels b, d, f, and h) was used to determine the efficiency of plasmid transfection by comparison with the total number of cells seen in the phase image (left panels a ,c, e, and g). Panels a and b, WT; panels c and d, G4864A; panels e and f, G4864V; panels g and h, G4864P. B, this panel shows the expression of eGFP-tagged proteins in mixed membrane preparations from HEK293 cells transfected with WT, G4864A, G4864P, and G4864V hRyR2 cDNA. 100 μg of protein was loaded on each lane. Relative levels of expression were quantified by densitometry. Note that the G4864A band was obtained from a different blot.

FIGURE 3.

Caffeine-induced release of Ca²⁺ from intracellular stores in HEK cells transfected with WT and mutant RyR2 channels. Fluorescence of fluo-3-AM loaded cells was monitored continuously before and after the addition of 10 mm caffeine. Traces shown are from representative experiments that have been repeated in at least three different coverslips for three different transfections. Similar responses were measured in all experiments. These graphs depict the normalized change in fluorescence (corrected from the background) ΔF/F₀ over time. Responses from five different cells are shown for each channel type. a.u., absorbance units.

FIGURE 4.

[³H]Ryanodine binding to mixed membranes. Experiments were performed in binding buffer containing 100 μm total Ca²⁺. Reactions contained 100 μg of WT protein and equivalent amounts of mutant RyR2 protein (normalized for differing levels of expression as explained under ”Results“). Data are plotted as mean ± S.E. of nine assays (three replicates from each of three membrane preparations). n.s., not significant.

FIGURE 5.

Current-voltage relationship for WT and G4864A RyR2. Single channel conductance was measured in 210 mm KCl, 20 mm HEPES at pH 7.2. Data are plotted as mean ± S.E. For WT, n = 5 or 6 individual channels. For G4864A, n = between 5 and 9 individual channels.

FIGURE 6.

Variation in WT and G4864A RyR2 open probability with changing cytosolic calcium. Gating activity of representative WT and G4864A RyR2 channels was recorded in 210 mm KCl at a holding potential of +40 mV. Upper traces show gating transitions in 10 μm Ca²⁺ (cytosol and luminal). Open probability is reduced markedly when cytosolic Ca²⁺ is lowered to 0.7 nm following the addition of chelating ligands (middle traces). Subsequent elevation of cytosolic Ca²⁺ to 100 μm increases the open probability of both channels (lower traces). Luminal Ca²⁺ was buffered at [Ca²⁺] = 50 nm in the middle and lower traces.

FIGURE 7.

Variation in mean WT and G4864A RyR2 open probability with changing cytosolic calcium. Experimental conditions were as described in the legend for Fig. 6. Data are displayed as mean ± S.E. for 5–10 individual channels. An unpaired t test was performed and statistical significance is indicated by asterisks. Upper, P_o, open probability, Middle, T_o: mean open time. Lower, T_c, mean closed time. n.s., not significant.

FIGURE 8.

Representative single channel traces of the G4864P RyR2 channel incorporated into planar lipid bilayers. A, current fluctuations from four different channels recorded at a holding potential (HP) of + 30 mV are displayed in the top panel (downward deflections represent opening events). Ai shows a rare example of long closing events within a period of high P_o. Aii and Aiii depict the oscillations between various subconductance (SC) states without return to the closed state. Aiv is the only G4864P channel we found having a very low P_o. B, the single channel activity of Aii and Aiii was further observed at different voltages in Bi and Bii, respectively. Bi shows additional subconductance states observed at −40 mV (upward deflections represent opening events). We cannot establish whether these additional states result from the activity of the same channel or the opening of a second channel. Bii shows a rare full closing event. Note the different time base in panels A and B of this figure.