Thimerosal stimulates Ca2+ flux through inositol 1,4,5-trisphosphate receptor type 1, but not type 3, via modulation of an isoform-specific Ca2+-dependent intramolecular interaction

Abstract

Thiol-reactive agents such as thimerosal have been shown to modulate the Ca2+-flux properties of IP3 (inositol 1,4,5-trisphosphate) receptor (IP3R) via an as yet unidentified mechanism [Parys, Missiaen, De Smedt, Droogmans and Casteels (1993) Pflügers Arch. 424, 516-522; Kaplin, Ferris, Voglmaier and Snyder (1994) J. Biol. Chem. 269, 28972-28978; Missiaen, Taylor and Berridge (1992) J. Physiol. (Cambridge, U.K.) 455, 623-640; Missiaen, Parys, Sienaert, Maes, Kunzelmann, Takahashi, Tanzawa and De Smedt (1998) J. Biol. Chem. 273, 8983-8986]. In the present study, we show that thimerosal potentiated IICR (IP3-induced Ca2+ release) and IP3-binding activity of IP3R1, expressed in triple IP3R-knockout R23-11 cells derived from DT40 chicken B lymphoma cells, but not of IP3R3 or [D1-225]-IP3R1, which lacks the N-terminal suppressor domain. Using a 45Ca2+-flux technique in permeabilized A7r5 smooth-muscle cells, we have shown that Ca2+ shifted the stimulatory effect of thimerosal on IICR to lower concentrations of thimerosal and thereby increased the extent of Ca2+ release. This suggests that Ca2+ and thimerosal synergetically regulate IP3R1. Glutathione S-transferase pull-down experiments elucidated an interaction between amino acids 1-225 (suppressor domain) and amino acids 226-604 (IP3-binding core) of IP3R1, and this interaction was strengthened by both Ca2+ and thimerosal. In contrast, calmodulin and sCaBP-1 (short Ca2+-binding protein-1), both having binding sites in the 1-225 region, weakened the interaction. This interaction was not found for IP3R3, in agreement with the lack of functional stimulation of this isoform by thimerosal. The interaction between the IP3-binding and transmembrane domains (amino acids 1-604 and 2170-2749 respectively) was not affected by thimerosal and Ca2+, but it was significantly inhibited by IP3 and adenophostin A. Our results demonstrate that thimerosal and Ca2+ induce isoform-specific conformational changes in the N-terminal part of IP3R1, leading to the formation of a highly IP3-sensitive Ca2+-release channel.

Figure 1. Expression of IP₃Rs in R23-11 cells

Microsomes (10 μg) from different R23-11 cell lines were separated on SDS/polyacrylamide gels and analysed by immunoblotting. The blot in (A) was probed with the anti-IP₃R1 polyclonal antibody Rbt03, whose epitope lies within residues 2735–2749, and the blot in (B) was probed with the anti-IP₃R polyclonal antibody Rbt475, whose epitope corresponds to the conserved stretch of amino acids in the N-terminus of all IP₃R isoforms. The arrows indicate the positions of the full-size IP₃Rs.

Figure 2. Ca²⁺-release properties in permeabilized R23-11 cells heterologously expressing IP₃R1 (A) and [Δ1–225]-IP₃R1 (B)

Free [Ca²⁺] was measured using fluo-3 fluorescence. After permeabilization of the plasma membrane with 50 μM digitonin, the cells accumulated Ca²⁺ in their intracellular stores, resulting in a decrease in medium-free [Ca²⁺]. After stabilization of the signal, adenophostin A (concentrations are shown in the Figures) was added. At the end of each experiment, 8 μM A23187 was added to estimate the total releasable amount of Ca²⁺. Results are typical for four independent experiments.

Figure 3. Effects of thimerosal and NEM on IP₃ binding and IICR in R23-11 cells heterologously expressing IP₃R1, [Δ1–225]-IP₃R1 and IP₃R3

(A) [³H]IP₃-binding assay performed on 200 μg of microsomes in a binding buffer at pH 8.3 using 10 nM [³H]IP₃ in the presence of 100 μM thimerosal or 300 μM NEM. Binding in the presence of 1 mM 2-mercaptoethanol was used as the control value (100%). Results are expressed as the means±S.E.M. for three independent experiments each performed in triplicate. (B–D) The effect of 100 μM thimerosal on IICR in permeabilized R23-11 cells heterologously expressing IP₃R1 (B), [Δ1–225]-IP₃R1 (C) and IP₃R3 (D). The experimental procedure was the same as that described in Figures 2(A) and 2(B). In these experiments, a subthreshold concentration of IP₃ (10 nM), which cannot provoke Ca²⁺ release, was used. After incubation with 100 μM thimerosal for 5 min, the same concentration of IP₃ (10 nM) was added. (E) The effect of 100 and 200 μM NEM on IICR in permeabilized R23-11 cells heterologously expressing IP₃R1. Results are typical for four independent experiments.

Figure 4. Effect of thimerosal on IICR in the presence or absence of 1 μM free Ca²⁺ in permeabilized A7r5 cells

Data are plotted as fractional loss (the amount of ⁴⁵Ca²⁺ leaving the stores in 2 min divided by the total store Ca²⁺ content at that time) as a function of time in (A, B). (A) Effect of the presence (•) or absence (○) of 1 μM thimerosal on Ca²⁺ release provoked by 200 nM IP₃ in the absence of Ca²⁺. (B) Effect of the presence (•) or absence (○) of 1 μM thimerosal on Ca²⁺ release provoked by 200 nM IP₃ in the presence of 1 μM free Ca²⁺. (C) Bell-shaped dependence of the effect of thimerosal on IICR. Data are plotted as the difference between the fractional loss after the addition of IP₃ and the fractional loss before the addition of IP₃ as a function of the thimerosal concentration: ○, 200 nM IP₃, 1 mM EGTA; □, 200 nM IP₃, 1 μM free Ca²⁺; Δ, 750 nM IP₃, 1 mM EGTA. (D) Effect of NEM (○), diamide (□) and SIN-1 (Δ) on IICR. Data are plotted as the difference between the fractional loss after the addition of IP₃ and the fractional loss before the addition of IP₃ as a function of the NEM, diamide or SIN-1 concentration respectively. The experiment was performed using 750 nM IP₃ and in the presence of 1 mM EGTA. Each curve represents the means±S.E.M for three wells.

Figure 5. Interaction between amino acids 1–225 and 226–604 of IP₃R1

(A) Pull-down experiment applied for GST–226–604 with 1–225 wild-type and with 1–225 C56A/C61A in the presence of 1 mM 2-mercaptoethanol or 100 μM thimerosal. C56 and C61, two conserved cysteine residues in the suppressor domain, were mutated into alanine residues. The recombinant suppressor domains were incubated with GST or GST–226–604 immobilized on glutathione–Sepharose 4B. After washing, the retained proteins were eluted. After SDS/PAGE, the proteins were analysed by immunoblotting with the anti-IP₃R polyclonal antibody Rbt475. Input (1–225 protein) was always 50 ng, whereas the lanes representing samples obtained after GST-pull down contained 2.5-fold more material. (B) Pull-down experiment applied for GST–226–604 with 1–225 in the presence of 100 μM thimerosal. The experimental procedure was the same as that described in (A). During the incubation reaction, 25 μM IP₃, 25 μM adenophostin A, 10 μM CaM or 10 μM sCaBP-1 was added. CaM and sCaBP-1 significantly inhibited the interaction to 44±13 and 37±12% of the control respectively. (C) Pull-down experiment applied for GST–226–604 with 1–225 in the presence or absence of Ca²⁺. Thimerosal (100 μM) was added during the incubation step. The experimental procedure was the same as that described in (A). The absence of Ca²⁺ inhibited the interaction to 26±18% of the control. Results are the means±S.D. for at least three independent experiments.

Figure 6. Effect of thimerosal and IP₃ on the interaction between the IP₃-binding and transmembrane domains of IP₃R1

(A) Pull-down experiment applied for GST–1–604 with FLAG(2170–2749) in the presence of 1 mM 2-mercaptoethanol or 100 μM thimerosal. EGTA (1 mM) or 25 μM Ca²⁺ was added as indicated in the Figure. The recombinant FLAG(2170–2749) was incubated with GST or GST–1–604 immobilized on glutathione–Sepharose 4B. After washing, the retained proteins were eluted. After SDS/PAGE, the proteins were analysed by immunoblotting with ANTI-FLAG® M2 monoclonal antibody. Lane ‘microsomes’ contains 0.5 μg of protein and lane ‘soluble fraction’ contains 2 μg of protein, whereas the lanes representing samples obtained after GST-pull down contain 15-fold more material than the soluble fraction. (B) Pull-down experiment applied for GST–1–604 with FLAG(2170–2749) in the presence of 1 mM 2-mercaptoethanol. During the incubation reaction, 20 μM IP₃ or adenophostin A was added. The experimental procedure was the same as that described in (A). IP₃ and adenophostin A significantly inhibited the interaction to 45±5 and 32±8% of the control respectively. Results are the means±S.D. for at least three independent experiments.