Feedback regulation mediated by Bcl-2 and DARPP-32 regulates inositol 1,4,5-trisphosphate receptor phosphorylation and promotes cell survival

Significance

The B-cell lymphoma-2 (Bcl-2) protein binds to the inositol 1,4,5-trisphosphate receptor (InsP₃R), a ubiquitous intracellular Ca²⁺ channel, thereby promoting cell survival by preventing excessive Ca²⁺ elevation. We report here that this mechanism involves Bcl-2 interaction with the Ca²⁺-activated protein phosphatase calcineurin (CaN) and dopamine- and cAMP-regulated phosphoprotein of 32 kDa (DARPP-32), a CaN-regulated inhibitor of protein phosphatase 1. Our findings indicate that this complex constitutes a rapid negative feedback mechanism that is activated by Ca²⁺ elevation and decreases InsP₃R phosphorylation, thus inhibiting InsP₃R-mediated Ca²⁺ release. We posit that Bcl-2-overexpressing cancer cells may exploit this mechanism to inhibit apoptosis.

Abstract

Bcl-2 interacts with the inositol 1,4,5-trisphosphate receptor (InsP₃R) and thus prevents InsP₃-induced Ca²⁺ elevation that induces apoptosis. Here we report that Bcl-2 binds dopamine- and cAMP-regulated phosphoprotein of 32 kDa (DARPP-32), a protein kinase A (PKA)-activated and calcineurin (CaN)-deactivated inhibitor of protein phosphatase 1 (PP1). Bcl-2 docks DARPP-32 and CaN in a complex on the InsP₃R, creating a negative feedback loop that prevents exaggerated Ca²⁺ release by decreasing PKA-mediated InsP₃R phosphorylation. T-cell activation increases PKA activity, phosphorylating both the InsP₃R and DARPP-32. Phosphorylated DARPP-32 inhibits PP1, enhancing InsP₃R phosphorylation and Ca²⁺ release. Elevated Ca²⁺ activates CaN, which dephosphorylates DARPP-32 to dampen Ca²⁺ release by eliminating PP1 inhibition to enable it to dephosphorylate the InsP₃R. Knocking down either Bcl-2 or DARPP-32 abrogates this feedback mechanism, resulting in increased Ca²⁺ elevation and apoptosis. This feedback mechanism appears to be exploited by high levels of Bcl-2 in chronic lymphocytic leukemia cells, repressing B-cell receptor-induced Ca²⁺ elevation and apoptosis.

Periodic elevations of intracellular Ca²⁺ serve as second messengers regulating mitochondrial metabolism, cell cycle entry, and cell survival (1, 2). Ca²⁺ signals are generated when inositol 1,4,5-trisphosphate receptors (InsP₃Rs) open in response to InsP₃, thus transferring Ca²⁺ from the endoplasmic reticulum into the cytoplasm and mitochondria. InsP₃R-mediated Ca²⁺ release is highly regulated, as excessive Ca²⁺ release causes cellular dysfunction and death (3). A number of proteins bind to InsP₃Rs and regulate channel opening, thus preventing excessive Ca²⁺ release (4). Among these is the antiapoptotic protein B-cell lymphoma-2 (Bcl-2) (5).

The Bcl-2 homology domain 4 (BH4) domain of Bcl-2 mediates its interaction with InsP₃Rs (6, 7). This domain also binds the Ca²⁺/calmodulin-activated serine/threonine (Thr) protein phosphatase calcineurin (CaN) (8). The BH4 domain binds within the regulatory and coupling domain of the InsP₃R, located between the InsP₃ binding site near the N terminus and the channel domain near the C terminus (6). Protein kinase A (PKA) phosphorylates serine (Ser) 1589 and Ser1755 within this domain, increasing InsP₃-mediated channel opening and Ca²⁺ release (9). Bcl-2 is reported to regulate InsP₃R phosphorylation at Ser1755 and to thereby govern InsP₃R-mediated Ca²⁺ release in murine embryonic fibroblasts, although the mechanism is not yet identified (10).

Bcl-2 plays an important role in regulating T-cell development and selection, processes that involve Ca²⁺ signaling and apoptosis regulation (11, 12). A 20-amino acid peptide (InsP₃R-derived peptide, or IDP) corresponding to the Bcl-2 binding site on the InsP₃R inhibits Bcl-2–InsP₃R interaction, thus eliminating Bcl-2’s control over InsP₃R-mediated Ca²⁺ elevation (6, 7). This peptide induces marked Ca²⁺ elevation and Ca²⁺-mediated apoptosis in primary chronic lymphocytic leukemia (CLL) cells and in B-cell lymphoma lines, indicating that Bcl-2–InsP₃R interaction contributes to the apoptosis-resistance characteristic of these lymphoid malignancies (13, 14).

Here we report that Bcl-2 interacts with dopamine- and cAMP-regulated phosphoprotein of 32 kDa (DARPP-32), a potent inhibitor of protein phosphatase 1 (PP1) that is activated by PKA-mediated phosphorylation at Thr 34 and deactivated by CaN-mediated dephosphorylation at this site (15). Our findings indicate that Bcl-2 docks DARPP-32 and CaN on the InsP₃R, creating a negative feedback loop that responds to InsP₃R-mediated Ca²⁺ release by inhibiting InsP₃R phosphorylation at Ser1755, thereby preventing excessive Ca²⁺ elevation capable of inducing cell death. Our findings implicate this Bcl-2–CaN–DARPP-32 feedback mechanism in regulating Ser1755 phosphorylation and apoptosis in primary human CLL cells.

Results

Bcl-2 Regulates InsP₃R-1 Ser1755 Phosphorylation.

T-cell lines are used in the present work, as in previous studies (8, 13, 16). Jurkat is a human CD4⁺ T-cell line that expresses readily detectable levels of Bcl-2 (Fig. 1A). The murine WEHI7.2 T-cell line is CD4⁺/CD8⁺-positive and thus corresponds to an early developmental stage in which Bcl-2 levels are low. These cells are referred to as “Lo Bcl-2” cells to distinguish them from Bcl-2 overexpressing WEHI7.2 cells, referred to as “Hi Bcl-2” cells (Fig. 1A). Bcl-2 levels are undetectable in the S49 murine T-cell line (Fig. 1A). This line is useful because of two available mutants deficient in PKA activity: one lacking Gα_s (Cyc⁻, historical naming) and the other lacking functional PKA (PKA⁻) (17).

Fig. 1.

Bcl-2 inhibits anti-CD3–induced elevation of P-Ser1755 InsP₃R-1 and Ca²⁺. (A) Immunoblot showing relative Bcl-2 levels in T-cell lines. (B) Anti-CD3 (20 µg/mL)-induced Ca²⁺ elevation in WEHI7.2 cells; each trace is average cytoplasmic Ca²⁺ level in 85 cells in an individual experiment representative of four experiments. (C) Peak cytoplasmic Ca²⁺ levels before and after adding anti-CD3 to WEHI7.2 cells (mean ± SE, four experiments, 340 cells per experiment). (D) Immunoblot representative of four experiments with similar results, illustrating dynamic changes in P-Ser1755 InsP₃R-1 before (0 min) and at 5 and 10 min after adding 20 µg/mL anti-CD3 to WEHI7.2 cells. (E) P-Ser1755 InsP₃R-1 levels relative to total InsP₃R-1 levels (mean ± SE, n = 4) 5 and 10 min after adding 20 µg/mL anti-CD3 antibody. Ratios representing the 0-min points (before adding anti-CD3) were normalized to a value of one.

In WEHI7.2 cells, Bcl-2 overexpression inhibits anti-CD3–induced Ca²⁺ elevation, as previously reported (6, 7, 18, 19) (Fig. 1 B and C). Before adding anti-CD3, basal InsP₃R-1 phosphorylation at Ser1755 in Hi Bcl-2 WEHI7.2 cells is equal to or greater than phosphorylation at this site in Lo Bcl-2 WEHI7.2 cells. On average, the ratio of phosphorylated (P)-Ser1755 InsP₃R-1 levels in Hi Bcl-2 to Lo Bcl-2 cells, normalized to actin, is 1.3 ± 0.3 (n = 7; P = 0.16). InsP₃R-1 P-Ser1755 levels increase after anti-CD3 treatment in Lo Bcl-2 WEHI7.2 cells but decrease in Hi Bcl-2 WEHI7.2 cells (Fig. 1 D and E). The increase in Ser1755 phosphorylation in Lo Bcl-2 WEHI7.2 cells is inhibited by the PKA inhibitor H89 (N-[2-[[3-(4-Bromophenyl)-2-propenyl]amino]ethyl]-5-isoquinolinesulfonamide), as is phosphorylation of cAMP response element-binding protein (CREB), a known PKA target (Fig. S1A). In addition, anti-CD3–mediated T cell receptor (TCR) activation increases InsP₃R-1 Ser1755 phosphorylation in S49 cells, but not in Cyc⁻ and PKA⁻ S49 cells (Fig. S1C). These findings are consistent with evidence that TCR activation increases the activity of PKA (20) and indicate that anti-CD3 induced phosphorylation of InsP₃R-1 at Ser1755 is PKA-mediated. Moreover, the findings indicate that InsP₃R Ser1755 phosphorylation contributes to the increase in cytoplasmic Ca²⁺ concentration induced by anti-CD3 treatment, as H89 treatment or PKA deficiency are associated with a marked decrease in anti-CD3–induced Ca²⁺ elevation (Fig. S1 B and D).

In summary, anti-CD3–induced TCR activation increases PKA-mediated phosphorylation of Ser1755 in InsP₃R-1, and this phosphorylation enhances anti-CD3–induced Ca²⁺ elevation. Bcl-2, in contrast, is associated with a decline in Ser1755 phosphorylation induced by anti-CD3 antibody treatment, and this decline correlates with Bcl-2-mediated inhibition of anti-CD3–induced Ca²⁺ elevation.

Bcl-2-Mediated Inhibition of Ser1755 Phosphorylation Involves Both CaN and PP1.

Because the BH4 domain of Bcl-2 binds CaN, we investigated a possible role of CaN in mediating the Bcl-2-associated decline in Ser1755 phosphorylation. Intracellular Ca²⁺ chelation by BAPTA-AM [1,2-Bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester)] reverses the inhibitory effect of Bcl-2 on anti-CD3–induced elevation of InsP₃R-1 P-Ser1755 (Fig. 2A). In addition, the CaN inhibitor tacrolimus (FK506) (21) increases InsP₃R-1 P-Ser1755 levels (Fig. 2 B and C) and reverses the inhibitory effect of Bcl-2 on anti-CD3–induced Ca²⁺ elevation (Fig. 2 D and E). These findings implicate CaN in the mechanism by which Bcl-2 decreases InsP₃R-1 P-Ser1755 after TCR activation. However, PKA-mediated protein phosphorylation is typically regulated by PP1 (22), and the effects of CaN on the InsP₃R are predicted to be indirect and secondary to the action of PP1 (23).

Fig. 2.

CaN and PP1 mediate P-Ser1755 InsP₃R-1 regulation by Bcl-2. (A) Immunoblot representative of three experiments indicating that Ca²⁺ chelation reverses the inhibitory effect of Bcl-2 on anti-CD3 (20 µg/mL)-induced P-Ser1755 InsP₃R-1. (B) Pretreatment with FK506 (0.5 µM for 2 h) reverses the inhibitory effect of Bcl-2 on anti-CD3-induced P-Ser1755 InsP₃R-1. (C) P-Ser1755 InsP₃R-1 levels relative to total InsP₃R-1 levels (mean ± SE, n = 3) 5 min after anti-CD3 antibody, indicating that pretreatment with FK506 (0.5 µM for 2 h) reverses the inhibitory effect of Bcl-2 on anti-CD3-induced P-Ser1755 InsP₃R-1. (D) Ca²⁺ traces representing the average cytoplasmic Ca²⁺ level in 85 cells in individual experiments (n = 3), indicating that pretreatment with FK506 (0.5 µM for 2 h) reverses the inhibitory effect of Bcl-2 on anti-CD3-induced Ca²⁺ elevation. (E) Summary of cytoplasmic Ca²⁺ concentrations before and after anti-CD3 addition to Hi Bcl-2 WEHI7.2 cells pretreated for 2 h with DMSO vehicle or FK506 (mean ± SE, n = 3 experiments). (F) In vitro phosphatase assays indicating that PP1, but not PP2A, dephosphorylates P-Ser1755 InsP₃R-1 (Upper, representative immunoblots; Lower, P-Ser1755 InsP₃R-1 phosphorylation levels relative to total InsP₃R-1 levels determined by densitometry of immunoblots; mean ± SD). (G) Immunoblots from a representative in vitro phosphatase assay indicating that CaN does not directly dephosphorylate InsP₃R-1 P-Ser1755; Ca²⁺-dependent dephosphorylation of NFAT-c2 serves as a positive control.

Therefore, in vitro phosphatase assays were used to clarify the respective roles of CaN and PP1 in regulating Ser1755 phosphorylation on InsP₃R-1. WEHI7.2 cells were pretreated with forskolin to increase InsP₃R-1 P-Ser1755 levels, and then InsP₃R-1s were immunoprecipitated from cell lysates. Immune-captured InsP₃R-1 was incubated with purified PP1 or PP2A enzymes in a phosphatase assay buffer (24). Under these conditions, recombinant PP1 and PP2A retain their substrate specificities (24). PP1, but not PP2A, causes a dose-dependent decrease of InsP₃R-1 P-Ser1755 (Fig. 2F). Consistent with this conclusion, 2 µM okadaic acid, a concentration that inhibits PP1 (22, 25, 26), reverses the inhibitory effect of Bcl-2 on anti-CD3–induced elevation of P-Ser1755 InsP₃R-1 and Ca²⁺ (Fig. S2). Moreover, an in vitro phosphatase assay was used to assess the effect of CaN on InsP₃R-1 P-Ser1755 (21). Incubation with purified CaN does not alter the phosphorylation of immune-captured InsP₃R-1, regardless of the presence of Ca²⁺ (Fig. 2G). CaN shifts the mobility of a typical CaN target, NFATc2, ensuring that CaN is active in the assay system (Fig. 2G).

These findings indicate a direct role for PP1 and an indirect role for CaN in Bcl-2-mediated regulation of InsP₃R-1 phosphorylation at Ser1755, suggesting potential involvement of DARPP-32 in regulating InsP₃R phosphorylation at Ser1755.

DARPP-32 and CaN Interact with Bcl-2 on the InsP₃R-1 and Regulate Ca²⁺ Elevation and Cell Survival.

DARPP-32 is detected at equal levels in both Lo Bcl-2 and Hi Bcl-2 WEHI7.2 cells, along with InsP₃R-1, PKA, PP1, and CaN, the levels of which are similar regardless of differing Bcl-2 levels (Fig. 3A). Bcl-2 immunoprecipitation from Lo Bcl-2 and Hi Bcl-2 cells pulls down both DARPP-32 and CaN, whereas these proteins are not pulled down by control IgG antibody (Fig. 3B). Bcl-2–DARPP-32 interaction is confirmed in the reciprocal experiment, where DARPP-32 is immunoprecipitated, pulling down both InsP₃R-1 and Bcl-2 (Fig. 3C). DARPP-32 also coimmunoprecipitates with ΔBH4-Bcl-2, a previously described Bcl-2 mutant that lacks the BH4 domain and neither interacts with the InsP₃R-1 or CaN nor regulates TCR-induced Ca²⁺ elevation (6, 16) (Fig. 3D).

Fig. 3.

Bcl-2 recruits DARPP-32 and CaN to InsP₃R-1, regulating DARPP-32 phosphorylation at Thr34 after anti-CD3-mediated TCR activation. (A) Immunoblots comparing levels of InsP₃R-1 and other proteins in Lo Bcl-2 and Hi Bcl-2 WEHI7.2 cells and in WEHI7.2 cells expressing Bcl-2 6,7mut, which does not interact with InsP₃R-1. (B) Coimmunoprecipitation (coip) of DARPP-32 and CaN with Bcl-2 (confirmed in four experiments). CaN input is best seen on prolonged exposure, whereas CaN coip is best compared in Lo vs. Hi Bcl-2 cells on short exposure. (C) coip of InsP₃R-1 and Bcl-2 with DARPP-32 (confirmed in four experiments). (D) coip of DARPP-32 with either Bcl-2 or Bcl-2 6,7mut and with ΔBH4Bcl-2, which lacks the BH4 domain (confirmed in three experiments). (E) Immunoblots demonstrating an increase in P-Thr34 DARPP-32 levels induced by anti-CD3 in Lo Bcl-2 WEHI7.2 cells and in WEHI7.2 cells expressing Bcl-2 6,7mut. In contrast, P-Thr34 DARPP-32 levels decline in Hi Bcl-2 WEHI7.2 cells after anti-CD3 addition (confirmed in three experiments).

These findings indicate that DARPP-32 interaction with Bcl-2 by itself is insufficient to regulate Ca²⁺ and that CaN interaction with the BH4 domain of Bcl-2 and Bcl-2 interaction with the InsP₃R are required to form a functional Bcl-2–CaN–DARPP-32 complex. This conclusion is substantiated by experiments in WEHI7.2 cells expressing Bcl-2 6,7mut, a previously described dicodon mutant in which arginine at position 6 and serine at position 7 in the BH4 domain of Bcl-2 are replaced by glycine (6). Immunoprecipitation of this mutant form of Bcl-2 pulls down DARPP-32 (Fig. 3D). Because this Bcl-2 mutant neither interacts with InsP₃Rs nor inhibits anti-CD3–induced Ca²⁺ elevation (6), the coimmunoprecipitation of DARPP-32 with Bcl-2 6,7mut confirms that the Bcl-2–DARPP-32 complex must be localized to the InsP₃R to function in regulating InsP₃R-mediated Ca²⁺ release. Consistent with these observations, TCR activation induces PKA-mediated DARPP-32 phosphorylation at Thr34 in cells expressing low levels of Bcl-2 or Bcl-2 6,7mut, whereas in cells expressing high levels of wild-type Bcl-2, basal P-Thr34 levels are already elevated and decline after TCR activation (Fig. 3E).

These findings suggest that the decline in InsP₃R Ser1755 phosphorylation observed in Bcl-2-positive WEHI7.2 cells is mediated by a Ca²⁺/CaN-induced dephosphorylation of DARPP-32 at Thr34, thus eliminating its inhibition of PP1 and resulting in PP1-mediated InsP₃R Ser1755 dephosphorylation.

Reversal of Bcl-2’s Action by Knocking Down Bcl-2 or DARPP-32.

The Bcl-2 level in Jurkat T-cells is sufficient to attenuate anti-CD3–induced Ca²⁺ elevation (6). Here we find that siRNA-mediated Bcl-2 knockdown in Jurkat cells reverses the inhibitory effect of Bcl-2 on InsP₃R Ser1755 phosphorylation (Fig. 4 A and B). Thus, an anti-CD3–induced elevation of InsP₃R Ser1755 phosphorylation is observed in cells in which Bcl-2 is knocked down, but not in control cells in which Bcl-2 is present and increased after anti-CD3 treatment. As reported previously (6), knocking down Bcl-2 results in increased Ca²⁺ elevation in response to anti-CD3 (Fig. 4C). Knocking down DARPP-32 (Fig. 4D) also increases InsP₃R Ser1755 phosphorylation (Fig. 4 E and F). Moreover, DARPP-32 knockdown increases anti-CD3–induced Ca²⁺ elevation (Fig. 4G) and apoptosis (Fig. 4H) in Jurkat cells, suggesting a role for this feedback mechanism in preserving cell survival after TCR activation.

Fig. 4.

Knockdown of Bcl-2 or DARPP-32 reverses Bcl-2’s inhibition of InsP₃R-1 phosphorylation, Ca²⁺ elevation, and apoptosis. (A) Immunoblots from two experiments in which siRNA Bcl-2 knockdown reversed Bcl-2’s inhibition of anti-CD3-induced P-Ser1755 InsP₃R-1 elevation. (B) P-Ser1755 InsP₃R-1 levels relative to total InsP₃R-1 levels at 5 and 10 min after anti-CD3 addition; ratios representing the 0-min point (before anti-CD3) were normalized to a value of one (two experiments). (C) Symbols represent peak Ca²⁺ elevation (mean ± SEM, n = 3 experiments, 85 cells per treatment per experiment) induced by anti-CD3 in Jurkat cells and measured by digital imaging. Cells were pretreated with control siRNA or Bcl-2 siRNA to knockdown Bcl-2. (D) Immunoblot documenting DARPP-32 knockdown. (E) Immunoblot representative of two experiments with the same result, showing the effect of DARPP-32 knockdown on P-Ser1755 InsP₃R levels before and 5 min after adding anti-CD3 to Jurkat cells. (F) P-Ser1755 InsP₃R-1 levels relative to total InsP₃R-1 levels 5 min after anti-CD3 addition (same experiments as in E; ratios representing the 0-min point before anti-CD3 addition were normalized to one). (G) Ca²⁺ elevation induced by anti-CD3 addition (arrow) in Jurkat cells pretreated with control or DARPP-32 siRNA. Traces represent average Ca²⁺ levels in 85 cells in a single experiment representative of three experiments. (H) Percentage of cells undergoing apoptosis 8–18 h after anti-CD3 addition (background apoptosis, before anti-CD3 addition, subtracted; mean ± SEM, six experiments).

Regulation of P-Ser1755 InsP₃R-1 Involving Bcl-2 and DARPP-32 in CLL Cells.

To investigate the potential relevance of these findings in pathological conditions, we investigated InsP₃R phosphorylation in CLL, a common form of leukemia invariably associated with elevated Bcl-2 levels, combined with tonic B-cell receptor signaling expected to generate InsP₃R-mediated Ca²⁺ elevation (27). In recent studies, we found that treating primary human CLL cells with the peptide TAT-IDP_DD/AA, which inhibits Bcl-2–InsP₃R interaction, induces Ca²⁺ elevation and, consequently, apoptosis (13). Here we detect the expression of DARPP-32 in primary human CLL cells (Fig. S3A) and find that treating these cells with TAT-IDP_DD/AA increases P-Ser1755 InsP₃R-1 levels (Fig. S3 B and C), followed by Ca²⁺ elevation (Fig. S2D). These findings indicate that peptide-mediated disruption of Bcl-2–InsP₃R interaction overcomes Bcl-2-imposed repression of P-Ser1755 InsP₃R-1 levels and InsP₃R-1-mediated Ca²⁺ release. These findings suggest that CLL cells exploit the Bcl-2–CaN–DARPP-32-mediated negative feedback mechanism to prevent Ca²⁺ elevation and cell death.

Discussion

The findings reported here indicate that Bcl-2 provides a molecular platform bringing CaN and DARPP-32 together on InsP₃R-1, producing a rapid feedback mechanism that senses InsP₃R-1-mediated Ca²⁺ release and responds by inhibiting PKA-mediated InsP₃R-1 phosphorylation at Ser1755 (Fig. 5). In a two-step process, CaN initiates the feedback process by sensing Ca²⁺ elevation and responds by dephosphorylating DARPP-32, thereby eliminating its inhibition of PP1. The Ca²⁺ elevation that triggers CaN activation in this step, we posit, is localized at the mouth of the InsP₃R, where CaN is positioned in close proximity to Bcl-2. Thus, this Ca²⁺ signal is not detected by the global measurement of intracellular Ca²⁺ used in this study. In the second step, PP1 dephosphorylates P-Ser1755 InsP₃R-1, decreasing the InsP₃R-mediated Ca²⁺ release that is readily detected by global Ca²⁺ measurements (e.g., Fig. 1B). In creating a negative feedback loop, Bcl-2 exploits the properties of CaN and DARPP-32 by targeting them to InsP₃R-1, where they respond rapidly to the Ca²⁺ exiting the InsP₃R-1 channel (Fig. 5). The requirement for Bcl-2 to assemble CaN and DARPP-32 in close proximity to InsP₃R-1, whether all four proteins are in the same Bcl-2 complex or in close proximity in distinct pools of Bcl-2, is supported experimentally by evidence that the feedback mechanism is not operational when Bcl-2 is substituted with mutant forms of Bcl-2 that do not interact with InsP₃R-1. Moreover, as PKA phosphorylates Ser1588 with similar effects on InsP₃R-1-mediated Ca²⁺ release (9), it is likely that the Bcl-2-mediated feedback mechanism would also regulate phosphorylation at this site, contributing to the effects of Bcl-2 on InsP₃R-mediated Ca²⁺ release.

Fig. 5.

Diagram of how the Bcl-2–CaN–DARPP-32 signaling complex functions as a negative feedback loop controlling InsP₃R-1-mediated Ca²⁺ release. (A) Bcl-2 serves as a platform docking CaN and DARPP-32 to InsP₃R-1. For simplicity, DARPP-32-Thr34 and InsP₃R-1-Ser1755 are depicted as being in a dephosphorylated state before TCR activation, although levels of Ser1755-InsP₃R and Thr34-DARPP-32 phosphorylation are variably increased in Bcl-2 overexpressing cells (see Discussion). (B) TCR stimulation by anti-CD3 increases activities of both phospholipase C (PLC), to increase InsP₃ synthesis, and adenylyl cyclase, to increase cAMP. The latter activates PKA, which phosphorylates Thr34-DARPP-32. PKA also phosphorylates Ser1755-InsP₃R-1, and this phosphorylation step is enhanced by the PP1 inhibitory activity of P-Thr34-DARPP-32. (C) InsP₃ activates InsP₃R-1 channel opening, which is enhanced by its phosphorylation at Ser1755. Ca²⁺ released from the endoplasmic reticulum activates CaN, which dephosphorylates P-Thr34-DARPP-32, alleviating inhibition of PP1 and initiating the feedback process. (D) Continuing the feedback process, PP1 dephosphorylates P-Ser1755, decreasing InsP₃R-1 channel opening and thus reducing Ca²⁺ release, accompanied by the inactivation of InsP₃ either through hydrolysis to InsP₂ or phosphorylation to InsP₄.

In summarizing the findings in Fig. 5, we emphasize the dynamic changes in P-Ser1755-InsP₃R-1 and P-Thr34-DARPP-32 that constitute the rapid feedback mechanism that takes place after TCR activation. For simplicity, this diagram does not show an increased basal phosphorylation of Ser1755-InsP3R-1 and Thr34-DARPP-32, although it is detected in certain experiments (Figs. 1D and 3E). Because PKA associates with InsP₃R-1 (4), we hypothesize that an elevation of basal P-Ser1755 InsP₃R-1 in cells with elevated Bcl-2, when observed, is secondary to constitutive TCR activity mediating basal PKA activity that increases P-Thr34-DARPP-32 levels, in turn inhibiting PP1 and thus raising the basal level of P-Ser1755 InsP₃R-1.

DARPP-32 has been investigated extensively in the brain, where it localizes to regions enriched in dopaminergic nerve terminals (15). The possibility that DARPP-32 may mediate effects of CaN on InsP₃R-1-induced Ca²⁺ release had been suggested previously (22, 23). Moreover, Tang and colleagues (22) discovered a direct association between PP1 and InsP₃R-1 and established that the association with PP1 facilitates dephosphorylation of PKA-phosphorylated InsP₃R-1. These investigators established the role of AKAP9, a multifunctional PKA anchoring protein, in docking PKA and PP1 to InsP₃R-1 (28) and in experiments with medium spiny neurons from DARPP-32 knock-out mice, demonstrated a regulatory role of DARPP-32 in dopamine-induced Ca²⁺ oscillations (29). Although these results advance our understanding of crosstalk between cAMP and InsP₃-mediated Ca²⁺ signaling pathways in the brain, much less is known about the role of DARPP-32 in peripheral tissues, including lymphocytes, although DARPP-32 has been shown to increase the phosphorylation and activity of various ion channels (30). The findings reported here are an indication of an interaction between Bcl-2 and DARPP-32 and of a role of this interaction in regulating Ca²⁺ signaling and cell survival.

Our findings indicate that the negative feedback mechanism conferred through Bcl-2 interaction with DARPP-32 and CaN contributes to the antiapoptotic function of the Bcl-2 protein. By regulating PKA-mediated InsP₃R phosphorylation at Ser1755, the Bcl-2–CaN–DARPP-32 complex sets a threshold level that Ca²⁺ cannot exceed, thereby preventing Ca²⁺-induced cell death. Knocking down either Bcl-2 or DARPP-32 abrogates this protective mechanism, elevating Ser1755 phosphorylation and thus increasing Ca²⁺ elevation after TCR activation, leading to loss of cell viability. Moreover, IDP_DD/AA-mediated inhibition of Bcl-2–InsP₃R interaction also abrogates this feedback mechanism, leading to cell death in primary human CLL cells. Thus, the Bcl-2–CaN–DARPP-32-mediated feedback mechanism may be essential to block proapoptotic Ca²⁺ signals and thereby prolong survival of CLL cells and possibly other Bcl-2-positive malignancies.

Although the present report focuses on Ca²⁺ signaling in lymphocytes, one can speculate that the Bcl-2–CaN–DARPP-32 feedback mechanism may also function in neuronal cells and that defects in this pathway may contribute to neuropsychiatric illnesses. Bcl-2 plays an important role in the development and survival of neuronal cells (31). Moreover, the neuroprotective effect of Bcl-2 in primary neuronal cells involves shuttling CaN to InsP₃Rs to regulate Ca²⁺ elevation (32). Exaggerated Ca²⁺ signals contribute to neuronal dysfunction in patients with bipolar disorder, schizophrenia, and Alzheimer disease (reviewed in ref. 33). In bipolar disorder, exaggerated Ca²⁺ signals correlate with single nucleotide polymorphisms associated with decreased Bcl-2 expression levels (34, 35). Moreover, the mechanism of action of mood stabilizers effective in treating this disorder involves, at least in part, elevation of Bcl-2 and stabilization of Ca²⁺ signaling (36). DARPP-32, in contrast, regulates many pathways in the central nervous system coupled with long-term plasticity and control of behavior (15). Altered DARPP-32 function has been implicated in the pathogenesis of schizophrenia (15), and DARPP-32 expression is deficient in leukocytes from patients with schizophrenia and bipolar disorder (37). Therefore, future studies are needed to investigate a potential functional link between unregulated Ca²⁺ signaling in these disorders and potential alterations in a Bcl-2–CaN–DARPP-32 feedback pathway required to prevent excessive Ca²⁺ signaling.

Materials and Methods

Reagents, primary cells, and cell lines have been described previously (6, 13). Also described are methods of peptide synthesis, immunoblotting and immunoprecipitation (6), T-cell stimulation by anti-CD3 antibodies and Ca²⁺ imaging (6), RNA interference (6), and in vitro phosphatase assays (24). Details of these methods and statistical analysis are in SI Materials and Methods.