Region-specific proteolysis differentially regulates type 1 inositol 1,4,5-trisphosphate receptor activity

The inositol 1,4,5 trisphosphate receptor (IP3R) is an intracellular Ca2+ release channel expressed predominately on the membranes of the endoplasmic reticulum. IP3R1 can be cleaved by caspase or calpain into at least two receptor fragments. However, the functional consequences of receptor fragmentation are poorly understood. Our previous work has demonstrated that IP3R1 channels, formed following either enzymatic fragmentation or expression of the corresponding complementary polypeptide chains, retain tetrameric architecture and are still activated by IP3 binding despite the loss of peptide continuity. In this study, we demonstrate that region-specific receptor fragmentation modifies channel regulation. Specifically, the agonist-evoked temporal Ca2+ release profile and protein kinase A modulation of Ca2+ release are markedly altered. Moreover, we also demonstrate that activation of fragmented IP3R1 can result in a distinct functional outcome. Our work suggests that proteolysis of IP3R1 may represent a novel form of modulation of IP3R1 channel function and increases the repertoire of Ca2+ signals achievable through this channel.

The inositol 1,4,5 trisphosphate receptor (IP 3 R) is an intracellular Ca 2؉ release channel expressed predominately on the membranes of the endoplasmic reticulum. IP 3 R1 can be cleaved by caspase or calpain into at least two receptor fragments. However, the functional consequences of receptor fragmentation are poorly understood. Our previous work has demonstrated that IP 3 R1 channels, formed following either enzymatic fragmentation or expression of the corresponding complementary polypeptide chains, retain tetrameric architecture and are still activated by IP 3 binding despite the loss of peptide continuity. In this study, we demonstrate that region-specific receptor fragmentation modifies channel regulation. Specifically, the agonist-evoked temporal Ca 2؉ release profile and protein kinase A modulation of Ca 2؉ release are markedly altered. Moreover, we also demonstrate that activation of fragmented IP 3 R1 can result in a distinct functional outcome. Our work suggests that proteolysis of IP 3 R1 may represent a novel form of modulation of IP 3 R1 channel function and increases the repertoire of Ca 2؉ signals achievable through this channel.
Calcium ions (Ca 2ϩ ) are utilized widely as an intracellular second messenger and interact with effectors to induce a diverse array of cellular activities. These events include gene expression, cell migration, muscle contraction, secretion, autophagy, and cell death (1)(2)(3). To use Ca 2ϩ to finely control cellular activities with high specificity and accuracy, cells have evolved a "calcium signaling toolbox" consisting of Ca 2ϩ channels, pumps, and binding proteins (4). These components function in concert and encode unique information in the forms of amplitude, frequency, and subcellular location of Ca 2ϩ signals. An important component of the toolbox is the inositol 1,4,5trisphosphate receptor (IP 3 R). 2 IP 3 R are intracellular Ca 2ϩ release channels expressed predominately in the membrane of the endoplasmic reticulum (ER) in most of eukaryotic species (5)(6)(7)(8). In response to IP 3 binding, IP 3 R are activated, resulting in Ca 2ϩ release from the intracellular store into the cytosol (9).
A series of complex regulatory events allows IP 3 R to encode Ca 2ϩ signals with distinct temporal and spatial characteristics. First, following binding of IP 3 , the interaction of Ca 2ϩ , nucleotides, and binding proteins with IP 3 R can regulate channel activity (10 -18). Second, posttranslational modifications, including phosphorylation, ubiquitination, and oxidation, can also influence channel activity and shape the IP 3 R Ca 2ϩ release profile (19 -23). As a further level of complexity, our laboratory has recently demonstrated that the particular isoform complement of the IP 3 R heterotetramer can either contribute to or largely determine the specific characteristics of IP 3 R-mediated Ca 2ϩ signals (24,25).
A further potential route of regulation of IP 3 R activity is through proteolytic fragmentation (26 -28). The functional consequence of IP 3 R fragmentation has been a subject of debate for years (29 -35). One of the major controversies centered over the question whether IP 3 R1 is a preferred substrate of caspase (35). Although several reports have demonstrated that caspase and calpain can cleave IP 3 R1 into at least two fragments (26,27), other evidence has suggested that IP 3 R1 is not generally cleaved under apoptotic conditions and is, in fact, only a late substrate of caspase during intense staurosporine-induced apoptosis (35). The latter evidence argued that fragmentation of IP 3 R1 does not play a key role in the process of apoptosis.
A further area of contention surrounds the biophysical properties of fragmented IP 3 R. Proteolytic fragmentation of IP 3 R1 by caspase and calpain results in at least two receptor fragments: the N-terminal fragment consists of the IP 3 binding core and much of the cytosolic peptide chain, whereas the C-terminal fragment contains the Ca 2ϩ permeation pore and the C-terminal cytosolic tail (26,27,33). One proposal suggests that receptor fragmentation physically dissociates the N-terminal region from the ER-associated C-terminal channel domain. Such cleavage and dissociation would be predicted to functionally uncouple the regulation of IP 3 binding from channel gating, leading to leaky C-terminal fragments retained in the ER, disruption of Ca 2ϩ homeostasis, and apoptosis (31,36). However, our laboratory proposed and demonstrated an alternative scenario by showing that IP 3 R1 retains its tetrameric architecture even after proteolytic fragmentation. Moreover, co-expression of complementary IP 3 R1 peptides, designed based on caspaseand calpain-fragmented IP 3 R1, can reconstitute the tetrameric channel structure that is functionally gated by IP 3 binding (33).
These data unambiguously demonstrate that IP 3 R1 is still tightly regulated by its endogenous ligand even after proteolytic fragmentation.
In this study, we have continued to investigate the consequences of proteolytic fragmentation of IP 3 R1. Given that fragmented IP 3 R1 are still functional in response to IP 3 binding, we hypothesized that disruption of peptide continuity by proteolytic cleavage may affect the fine regulation of IP 3 R1 and, subsequently, alter IP 3 R1-mediated Ca 2ϩ signals. This study demonstrates that proteolytic fragmentation has profound effects on IP 3 -mediated Ca 2ϩ signals, resulting in alteration of the signature temporal pattern of Ca 2ϩ release through IP 3 R1. Further, fragmentation can abolish PKA regulation of the receptor in a cleavage region-specific manner. More importantly, we show that the altered Ca 2ϩ signals elicited by fragmented IP 3 R1 can specifically activate distinct downstream effectors compared with IP 3 R WT. Our results therefore strongly suggest that proteolytic fragmentation may represent a novel form of regulation of IP 3 R activity that expands the repertoire of signaling through IP 3 R1 activation.

Functional fragmented IP 3 R1 are assembled from complementary IP 3 R1 fragments
We have reported previously that heterologous expression of complementary polypeptide fragments, designed based on putative caspase and calpain proteolytic cleavage sites, can be assembled into IP 3 R1 tetramers (33). This strategy provides an experimental platform to investigate the functional consequences of IP 3 R fragmentation. Our previous work strongly suggested that tetramers assembled from complementary IP 3 R1 fragments are functional in terms of IP 3 -gated Ca 2ϩ release. To further characterize the functional consequence of receptor fragmentation, we designed additional complementary IP 3 R1 fragments according to defined IP 3 R1 trypsin cleavage sites in the rat IP 3 R1 (Fig. 1, A and B). Previous studies have shown that exposure of purified IP 3 R1 protein to a low concentration of trypsin in vitro results in five receptor fragments (37)(38)(39). These data have been interpreted to indicate that IP 3 R1 consists of five compact globular domains connected by four solvent-exposed linker regions. Throughout this work, cleavage sites introduced after the second or third trypsin cleavage sites, respectively, are denoted as IP 3 R1 I-IIϩIII-V (tryp) and IP 3 R1 I-IIIϩIV-V (tryp) (Fig. 1B). Caspase and calpain both cleave IP 3 R1 in the same solvent-exposed region (after tryptic fragment IV) and are denoted as IP 3 R1 I-IVϩV(casp) and IP 3 R1 I-IVϩV (calp) (Fig. 1B). To confirm that all fragmented IP 3 R1 are functional, DT40-3KO cells (null for all IP 3 R) (40) expressing various complementary receptor fragments (Fig. 2, A-D) were loaded with the Ca 2ϩ indicator fura-2 and then challenged with the G␣ q /IP 3 -linked, protease-activated receptor 2 (PAR2) agonist trypsin. Consistent with our previous work, all types of complementary receptor fragments were capable of inducing an elevation in [Ca 2ϩ ] i in response to maximal PAR2 activation (Fig. 2, E-I). Further, DT40-3KO cells stably expressing only the channel fragment (IP 3 R1 V) showed no Ca 2ϩ response when PAR2 was activated (Fig. 2J). Remarkably, functional channels were assembled by co-transfecting the remaining four complementary receptor fragments into these cells (Fig. 2K). To confirm this finding, we performed the co-transfection experiments in HEK cells. We and others have thoroughly characterized and demonstrated that HEK-3KO cells (a HEK cell line null for all IP 3 R) are completely devoid of any functional IP 3 R (41,42). Consistently, trypsin-stimulated PAR2 activation induced Ca 2ϩ release in HEK-3KO cells only when the complementary five tryptic fragments in a monomer were transiently expressed in cells (Fig. 2L). Together, these data confirm that complementary receptor fragments are assembled into functional channels that are gated by IP 3 binding. Next we investigated whether IP 3 R fragmentation results in functionally equivalent IP 3 R or, alternatively, whether the biophysical profile or regulation of the fragmented IP 3 R differs from IP 3 R formed from conventional peptide monomer chains with linear continuity.

Proteolytic fragmentation alters the temporal pattern of IP 3 R1-mediated Ca 2؉ release in a region-specific manner
IP 3 R exhibit subtype-specific Ca 2ϩ release profiles when continuously exposed to IP 3 . This is best exemplified following activation of B cell receptors (BCR) on DT40-3KO cells evoked by cross-linking with ␣-IgM (43). For example, stimulation of DT40-3KO cells stably expressing IP 3 R1 WT with ␣-IgM characteristically evokes only a few transient increases in [Ca 2ϩ ] i ( Fig. 3A and Refs. 41,43,44), whereas stimulation of cells expressing mouse IP 3 R2 WT elicits robust Ca 2ϩ oscillations (13,41,43). A higher concentration of ␣-IgM (2 g/ml) had no impact on the profile of Ca 2ϩ release evoked in cells expressing IP 3 R1 WT (Fig. 3G). In addition, this general pattern was observed in various clones of DT40-3KO cells with different IP 3 R1 WT expression levels (data not shown), indicating that the temporal pattern of [Ca 2ϩ ] i signal likely reflects an intrinsic property of the particular IP 3 R. Next we examined the Ca 2ϩ release profile of cells expressing various complementary receptor fragments. When the cleavage site was located closer to the N terminus of the receptor, such as with IP 3 R1 I-IIϩIII-V (tryp), similar to IP 3 R1 WT, a low number of Ca 2ϩ transients were evoked upon ␣-IgM stimulation (Fig. 3, B and C). Remarkably, when cleavage sites were introduced further toward the C terminus of the receptor (for example, IP 3 R1 I-IIIϩIV-V (tryp), IP 3 R1 I-IVϩV (casp), and R1 I-IVϩV (calp), a significant increase in the ability of BCR stimulation to did not respond to trypsin (500 nM) stimulation. K, transient co-expression of IP 3 R1 I (tryp), IP 3 R1 II (tryp), IP 3 R1 III (tryp), and IP 3 R1 IV (tryp) into DT40-3KO cells stably expressing IP 3 R1 V (tryp) constituted functional Ca 2ϩ release channels. L, three representative HEK-3KO cells demonstrating that functional IP 3 R1 can be assembled from 20 IP 3 R1 polypeptide fragments corresponding to those produced by trypsinization of each monomer into five fragments. Ca 2ϩ imaging assays were repeated five times for each set of complementary receptor fragments. evoke oscillatory activity was observed (Fig. 3, D-G). These data provide evidence that IP 3 R fragmentation at sites corresponding to cleavage by calpain and caspase can markedly alter the activity of the receptor and subsequently alter the temporal profile of Ca 2ϩ release. We reported previously that a significant increase in the ability of IP 3 R1 to support Ca 2ϩ oscillations occurred following PKA-mediated phosphorylation of the receptor (44). To investigate whether PKA phosphorylation plays a role in the increased oscillatory activity of particular fragmented IP 3 R1, BCR-stimulated [Ca 2ϩ ] i signals were studied in cells expressing PKA non-phosphorylatable (S1589A, S1755A) and phosphomimetic (S1589E, S1755E) IP 3 R1 I-IVϩV (calp). These mutations had no impact on the Ca 2ϩ release profile mediated by IP 3 R1 I-IVϩV (calp) (Fig. 4, A-D), indicating that PKA phosphorylation does not underlie the gain of oscillatory activity observed in the specific fragmented IP 3 R1. In addition, Ca 2ϩ oscillations were retained for an extended period of time in the absence of extracellular Ca 2ϩ (Fig. 5, A and B), suggesting that Ca 2ϩ influx is not necessary to promote Ca 2ϩ oscillations from these fragmented IP 3 R1.
We next investigated the profile of Ca 2ϩ signals mediated by fragmented IP 3 R1 more directly by photorelease of a cell-permeable, poorly degradable, caged IP 3 (45,46). DT40-3KO cells expressing IP 3 R1 WT showed a sustained monophasic [Ca 2ϩ ] i signal in response to photorelease of caged IP 3 (Fig. 6A). In contrast, cells expressing IP 3 R2 WT exhibited robust oscillatory Ca 2ϩ signals (Fig. 6B). Notably, IP 3 R1 I-IVϩV (calp) showed both sustained monophasic and oscillatory responses (Fig. 6, C and D). A distribution of the frequency of Ca 2ϩ oscillations shows that cells expressing IP 3 R1 WT mainly evoked a limited number of Ca 2ϩ transients in response to IP 3 exposure (Fig. 6E), whereas cells expressing IP 3 R1 I-IVϩV (calp) displayed an increase in the population of cells that exhibit robust oscillatory Ca 2ϩ release (Fig. 6F). This observation was consistent with the statistics showing that cells expressing IP 3 R1 I-IVϩV (calp) induced significantly more Ca 2ϩ transients Fura-2/AM, followed by cross-linking the B cell receptor using ␣-IgM (500 ng/ml). Two representative Ca 2ϩ traces are shown for each pair of complementary receptor fragments. The numbers of Ca 2ϩ transients in 15 min of experiments were calculated. C and G, scatterplots indicate that a cleavage site in solvent-exposed region II has no effect on the temporal Ca 2ϩ release profile (C, Student's t test), whereas fragmentation sites more toward to the C terminus significantly increase the number of Ca 2ϩ transients mediated by complementary receptor fragments (G, one-way ANOVA followed by Dunnett post-test). *, statistical significance determined by Dunnett post-test; n.s., not significant. Ca 2ϩ imaging assays were repeated seven times, with more than 40 cells in each experimental run for each set of complementary receptor fragments.

Region-specific proteolytic regulation of IP 3 R1
compared with IP 3 R1 WT in response to photorelease of caged IP 3 (Fig. 6G).

Proteolytic fragmentation alters the single-channel open probability of IP 3 R1
One caveat of the single-cell imaging assay above is that fragmented IP 3 R1 is not generated from the proteolytic cleavage of full-length IP 3 R1 WT but from the assembly of complementary receptor fragments. Therefore, we next performed patchclamp recording in the "on nucleus" configuration to fragment IP 3 R1 WT in situ and directly investigate the biophysical consequences of IP 3 R1 fragmentation at the single-channel level (18). A submaximal concentration of IP 3 (1 M) resulted in an increase in the steady-state open probability (P o ) of IP 3 R1 to ϳ20% (Fig. 7, A and G). Addition of active caspase-3 (3 ng/ml) in the patch pipette in the presence of IP 3 (1 M) significantly augmented the channel P o to ϳ70% (Fig. 7, B and G). Higher concentrations of active caspase-3 (10 ng/ml or 30 ng/ml) either diminished the conductance of the channel or completely inactivated the receptor, likely because of nonspecific digestion followed by destruction of the receptor (Fig. 7, C, D, and G). Addition of the caspase inhibitor Z-VAD (20 M) completely blocked the effect of active caspase on channel P o (Fig. 7, E and G). In addition, in cells expressing constructs where the caspase cleavage site (DEVD at amino acids 1888 -1891) was mutated to be noncleavable (IEVA) (Fig. 7, F and G) (30), no increase in P o was observed in the presence of active caspase-3. These data strongly suggest that specific receptor fragmentation by caspase-3 at Asp-1891 enhances IP 3 R1 channel activity (Fig.  7G), and this likely reflects the single-channel correlate of the increase in oscillatory activity observed in intact cells expressing fragmented IP 3 R1 following stimulation with IP 3 .

PKA regulation is abolished in calpain-fragmented IP 3 R1
Although fragmented IP 3 R is still gated by IP 3 binding, how important individual modes of regulation of IP 3 R1 are impacted by receptor fragmentation remains unclear. Phosphorylation of IP 3 R1 at Ser-1589 and Ser-1755 by PKA significantly increases channel P o at the single-channel level (21,22). This biophysical alteration is manifested as an increase in Ca 2ϩ release at the single-cell level determined in Ca 2ϩ imaging assays. Interestingly, the calpain fragmentation site is located between the PKA phosphorylation sites and the receptor channel domain. As a result, proteolytic receptor fragmentation by calpain separates PKA phosphorylation sites and the Ca 2ϩ permeation pore into two different peptide fragments. Given the location of the calpain fragmentation site in IP 3 R1, we next investigated the effects of fragmentation on PKA regulation of IP 3 R1. Activation of PKA resulted in IP 3 R1 phosphorylation and, subsequently, a significant increase in Ca 2ϩ release in cells expressing IP 3 R1 compared with DMSO-treated con-

. Altered temporal Ca 2؉ release profile mediated by IP 3 R1 I-IV؉V (calp) is not due to PKA-mediated receptor phosphorylation.
A-C, DT40-3KO cells stably expressing IP 3 R1 I-IVϩV (calp) (A), IP 3 R1 I-IVϩV (calp) (S1589A, S1755A) (B), or IP 3 R1 I-IVϩV (calp) (S1589E, S1755E) (C) were loaded with 2 M Fura-2/AM, followed by cross-linking the B cell receptor using ␣-IgM (500 ng/ml). Two representative Ca 2ϩ traces are shown for each set of complementary receptor fragments. D, scatterplot indicating that there is no significant difference among three types of receptor fragments with respect to the number of Ca 2ϩ transients in 30 min of recording (one-way ANOVA). Ca 2ϩ imaging assays were repeated three times, with more than 40 cells in each experimental run for each complementary receptor fragments. n.s., not significant.

Figure 5. Ca 2؉ influx is not necessary for complementary receptor fragments to induce robust Ca 2؉ oscillation.
A, DT40-3KO cells stably expressing IP 3 R1 I-IVϩV (calp) were loaded with 2 M Fura-2/AM, followed by crosslinking the cell surface B cell receptor using ␣-IgM (500 ng/ml). Perfusion buffers with or without extracellular Ca 2ϩ were alternated every 10 min during recording as indicated. B, column statistics suggest that there is no significant difference between conditions with or without extracellular Ca 2ϩ with respect to the number of Ca 2ϩ transients (Student's t test). Experiments were repeated five times, with more than 40 cells in each experimental run. n.s., not significant.
trols in response to all PAR2 agonist concentrations (Fig. 8,  A, B, and F). The increase in phosphorylation and regulation of Ca 2ϩ signals was completely abolished in cells expressing non-phosphorylatable IP 3 R1 mutants at both PKA phosphorylation sites (Fig. 8, C and F). Notably, when a disruption of peptide continuity was introduced at the third trypsin cleavage site, IP 3 R1 I-IIIϩIV-V (tryp), PKA phosphoregulation was maintained (Fig. 8, D and F). In contrast, although forskolin pretreatment increased the level of phosphorylated IP 3 R1 I-IVϩV (calp), no significant difference in terms of the amplitude of Ca 2ϩ response was observed. (Fig. 8, E and G). These data demonstrate that PKA regulation of IP 3 R activity requires the phosphorylated residues to be proximal or, possibly, in the same fragment as the channel domain and further suggest that other particular modes of regulation of IP 3 R1 activity may be altered depending on the specific site of fragmentation.

Fragmented IP 3 R1 can activate distinct downstream effectors
Based on the observation that specific fragmented IP 3 R1 can induce different Ca 2ϩ signals, we next investigated whether fragmented IP 3 R1 can specifically activate distinct downstream effectors. Oscillatory Ca 2ϩ signals, but not single Ca 2ϩ transients, have been shown to specifically activate the transcription factor nuclear factor of activated T cells (NFAT). Ca 2ϩ oscillations are thought to deliver signals with the appropriate spatial and temporal properties necessary to activate the phosphatase calcineurin, which dephosphorylates NFAT and facilitates its translocation to the nucleus (45,47,48). We hypothesized that caspase-or calpain-fragmented IP 3 R1, but not IP 3 R1 WT, might provide the necessary Ca 2ϩ signal to activate NFAT translocation. GFP-tagged NFAT (NFAT-GFP) was transfected into cells expressing either IP 3 R1WT, IP 3 R1 I-IVϩV (casp), or IP 3 R1 I-IVϩV (calp). In the quiescent state, NFAT-

Region-specific proteolytic regulation of IP 3 R1
GFP was mainly located in the cytosol (Fig. 9, B, E, and H). Upon BCR stimulation, NFAT-GFP efficiently translocated from the cytosol into the nucleus in cells expressing either IP 3 R1 I-IVϩV (casp) or IP 3 R1 I-IVϩV (calp) (Fig. 9, A-F, J, and K). Translocation of NFAT-GFP was not observed in cells expressing IP 3 R1 WT (Fig. 9, G-K). These data strikingly illustrate that the distinct patterns of Ca 2ϩ signal evoked through particular fragmented IP 3 R have the ability to activate distinct downstream events compared with the intact IP 3 R1.

Discussion
The data presented here expand on our earlier findings exploring the functional consequences of fragmentation of IP 3 R1 by intracellular proteases. We now demonstrate that expression of complementary polypeptides that correspond to any pair of fragments representing the five globular domains generated in vitro by limited tryptic exposure similarly result in functional channels (37). Indeed, co-expression of individual cDNA encoding the five domains separately, remarkably, leads to the assembly of a functional channel gated by IP 3 . Although these data firmly establish that peptide continuity is not required for channel opening per se, the major finding of this study is that fragmentation of IP 3 R1 markedly alters allosteric regulation of the channel by key modulators to alter Ca 2ϩ release activity. Specifically, as an example, we show that a prominent mode of regulation of IP 3 R1, augmented Ca 2ϩ release following PKA phosphorylation, is lost when fragments are expressed that mimic calpain cleavage to generate fragmented IP 3 R1 I-IVϩV. However, regulation by PKA is clearly evident when peptide continuity is lost more proximal to the N terminus to yield IP 3 R1 I-IIIϩIV-V. Notably, in this case, the key phosphorylation sites at Ser-1589 and Ser-1755 are present in the same fragment as the channel pore in fragment V, suggesting that peptide continuity is required to communicate the conformation change imparted by phosphorylation to modulation of gating of the pore. Further, these data are consistent with the observation that PKA phosphorylation does not alter IP 3 binding to the binding core in the N terminus, which is present in the complementary fragment (49). These data clearly demonstrate that alterations in allosteric modulation of IP 3 R activity are dependent on the site of cleavage.
A further demonstration that the activity of IP 3 R1 is dramatically altered by cleavage is the observation that the temporal profile of Ca 2ϩ release following agonist stimulation is temporally transformed when the channel is assembled from particular complementary polypeptide chains. We and others have consistently reported that sustained stimulation of cells expressing individual IP 3 R subtypes in isolation supports specific patterns of Ca 2ϩ release that can be considered a "signature" for that subtype (13,41,43,44). As the signal is largely independent of extracellular Ca 2ϩ , these signatures are the result of the inte-  regulatory input received by the particular IP 3 R expressed. Notably, in this study, we show that expression of pairs of complementary fragments, in particular those corresponding to products derived from caspase or calpain proteolytic fragmentation, results in a transformation of the Ca 2ϩ signal from a largely transient increase into an oscillatory profile consisting of numerous organized transients present throughout stimulation. In addition, single-channel recordings indicate that in situ cleavage of IP 3 R1 by caspase significantly increased the channel open probability. This provides a potential underlying mechanism for altering IP 3 R1 activity through proteolytic fragmentation. Consistent with the widely held view that the spatial and temporal properties of Ca 2ϩ signals are important for the activation of downstream effectors, these oscillatory Ca 2ϩ signals were capable of driving the nuclear localization of NFAT-GFP whereas signals through the IP 3 R1 WT could not. Taken together, these data provide support for the hypothesis that cleavage by proteases can potentially have significant consequences for IP 3 R1 activity and might be considered a novel mode of regulation influencing multiple modulatory inputs.
The cryo-EM structure of IP 3 R1 was recently solved to near atomic resolution and provides structural details consistent with our findings (50). The suppressor domain (TF␤1) and the IP 3 binding core (TF␤2 and the N-terminal region of ARM1) at the N terminus of the receptor physically interact with the C-terminal domain of the adjacent subunit. This interaction

. The increase in Ca 2؉ release in IP 3 R1 by PKA phosphorylation is regulated by receptor fragmentation in a cleavage region-specific manner.
A, cells were loaded with 1 M Fluo-2/AM for 1 h, followed by a FlexStation assay to monitor the change in [Ca 2ϩ ] i . Preincubation of cells stably expressing IP 3 R1 WT with forskolin significantly increased the amplitude of Ca 2ϩ release in response to PAR2 activation. B and D, this increase was observed at all agonist concentrations tested for IP 3 R1 WT (B) and at high agonist concentrations tested for IP 3 R1 I-IIIϩIV-V (tryp) (D). C and E, forskolin had no effect on the amplitude of Ca 2ϩ release in cells expressing IP 3 R1 (S1589A, S1755A) (C) or IP 3 R1 I-IVϩV (calp) (E). F and G, pretreatment of forskolin results in IP 3 R1 phosphorylation at residue Ser-1755 in both full-length and fragmented receptors. Statistics were performed using one-way ANOVA followed by Dunnett post-test. Experiments were repeated three times for each set of IP 3 R1 or complementary receptor fragments. WB, Western blot.

Region-specific proteolytic regulation of IP 3 R1
provides a physical basis for a direct communication between IP 3 binding at the N terminus and the distant channel opening at the C terminus without requiring signaling propagation along the whole peptide sequence. These data are consistent with our conclusion that peptide continuity in the receptor coupling domain, which comprises the helical domain, armadillo solenoid folds (ARM), and the intervening later domain, is not required for channel opening following IP 3 binding. However, the coupling domain is crucial for integrating intracellular input and, accordingly, imposing regulations on the channel. (casp), IP 3 R1 I-IVϩV (calp), and IP 3 R1 WT were transiently transfected with NFAT-GFP, followed by 12-h recovery in a 39°C, 5% CO 2 incubator. Cells were then mounted in a perfusion chamber, the location of NFAT-GFP was monitored following excitation with 488 nm, and emitted fluorescence recorded above 510 nm. Prior to stimulation, NFAT-GFP was mainly located in the cytosol (B, E, and H). A, D, and G, 3D heat plots, with the x and y axes indicating spatial coordinates the and z axis indicating the amplitude of the GFP signal. A, C, D, F, G, I, and J, addition of ␣-IgM (500 ng/ml) activated and translocated NFAT-GFP from the cytosol into the nucleus in cells expressing IP 3 R1 I-IVϩV (casp) (A, C, and J) and IP 3 R1 I-IVϩV (calp) (D, F, and J). This translocation was not observed in cells expressing IP 3 R1 WT (G, I, and J). C, F, and I, nucleus NFAT-GFP signals are shown in red and cytosolic NFAT-GFP signals in black and were used to calculate the nucleus/cytosol ratio (N/C, blue). J, the percentage of cells where translocation could be observed in individual trials in which more than 20 cells were imaged per trial. K, the magnitude of the N/C ratio in individual cells from each trial that showed translocation. These data demonstrate a significant increase in nucleus/cytosol ratio for fragmented IP 3 R1 compared with IP 3 R1 WT in response to ␣-IgM stimulation. Experiments were repeated four times for IP 3 R1 and three times for each type of complementary receptor fragments. *, P Ͻ 0.01. This concept was again strongly supported by the recent structure that shows that flexible structures of the ARM domains were amenable to generating interfaces for recognition and binding of various regulators. Noticeably, both caspase and calpain cleavage sites as well as the fragmentation site for IP 3 R1 I-IIIϩIV-V are located in ARM regions. Given the critical role of the coupling domain in terms of receptor regulation, it is conceivable that disruption of peptide continuity in the ARM regions (IP 3 R1 I-IIIϩIV-V and IP 3 R1 I-IVϩV) are likely to either interfere with the intrinsic regulation of ARM on the channel domain or affect other mediators that regulate channel activity through binding to or modifying ARM domains and thus alter IP 3 R1-induced Ca 2ϩ signals.
What is the mechanism underlying the ability of particular fragmented IP 3 R1 to support robust Ca 2ϩ oscillations? A canonical model for class I Ca 2ϩ oscillations provides a possible explanation for this remarkable alteration (51). This model suggests that cells can control the Ca 2ϩ oscillation period by modulating the rate of Ca 2ϩ activation of IP 3 R in response to the change in [Ca 2ϩ ] i . Of note, although there was no EF-hand Ca 2ϩ binding motif found in IP 3 R1, the ARM3 region in the coupling domain contains a putative Ca 2ϩ binding region with a highly conservative Glu-2100 critical for Ca 2ϩ binding and regulation of IP 3 R1 (52,53). Based on the model and the location of the Ca 2ϩ sensor in the ARM regions, we speculate that fragmentation of IP 3 R1 at particular ARM regions impacts the Ca 2ϩ modulation of channel gating and, consequently, alters the Ca 2ϩ activation rate for IP 3 R1. This hypothesis is supported by mathematical simulations showing that, by solely changing the rate of IP 3 R1 activation by Ca 2ϩ , the normal monotonic Ca 2ϩ release pattern through wild-type IP 3 R1 Ca 2ϩ can be converted into robust oscillatory Ca 2ϩ signals observed with IP 3 R1 I-IIIϩIV-V and IP 3 R1 I-IVϩV (51).
Although IP 3 R1 has been reported to be a substrate of caspase and calpain for more than a decade, any cellular role of cleavage has yet to be firmly established. We and other laboratories report that, even when caspase and calpain are massively activated during apoptosis, only a small fraction of IP 3 R1 is actually fragmented. Therefore, we envision that, when caspase or calpain are mildly activated under non-apoptotic conditions, a small proportion of a particular fragmented IP 3 R1 is generated in cells that might then locally activate specific downstream effectors and fulfill unique roles. When might lower levels of protease activity occur? Recent reports have suggested that caspase and calpain activity is also essential for processes other than initiating cell death. These include protein maturation, cell proliferation and differentiation, and myoblast fusion (54 -61). More importantly, there is burgeoning evidence suggesting that, beyond negating protein activity, caspase can activate its substrates by proteolysis (54,56,60,62). Consistent with this idea, this study provides evidence that enzymatic cleavage per se might uniquely regulate IP 3 R1 activity and thus might potentially contribute to the regulation of different cellular activities rather than cell death.
In summary, our work provides the first evidence showing that proteolytic fragmentation may serve as a novel regulatory event for IP 3 R1 by altering its Ca 2ϩ release profile. Future work is necessary to specifically investigate the pathophysiological conditions under which IP 3 R1 is cleaved and the corresponding significance of receptor fragmentation for those processes. We envision that spatially confined protease activity may cleave the IP 3 R1 to obtain a transient alteration of receptor-mediated Ca 2ϩ signals, which can rapidly switch on alternative signaling pathways for unique cellular activities. In addition, there is evidence showing that both IP 3 R2 and IP 3 R3 can also be fragmented by proteases (63). We are currently investigating the functional consequence of proteolytic fragmentation for IP 3 R2 and IP 3 R3. Our study will answer the important question of whether proteolytic fragmentation may be a general regulatory event for all isoforms of IP 3 R.

Reagents
All restriction enzymes and T4-DNA ligase were from New England Biolabs. RPMI 1640 medium, penicillin/streptomycin, G418 sulfate, ␤-mercaptoethanol, and chicken serum were purchased from Invitrogen. Fetal bovine serum was from Gemini Bioproducts. Iso-Ins(1,4,5)P 3 /PM (caged), Z-VAD-fmk, and active caspase-3 were from Enzo Life Science. Fura-2 was from Teflabs. All reagents for SDS-PAGE were from Bio-Rad. The N-terminal antibody for IP 3 R1 and phospho-IP 3 R1 were from Cell Signaling Technology. DyLight TM 800CW secondary antibody was from Thermo Scientific. Forskolin was from Sigma-Aldrich. Mouse ␣-chicken IgM was from Southern Biotech. The antibody against the C-terminal 19 amino acids of IP 3 R1 was generated by Pocono Rabbit Farms and Laboratories.

Constructs
The method for generation of IP 3 R1 I-IVϩV (casp) and IP 3 R1 I-IVϩV (calp) was first described elsewhere (33). In brief, to create IP 3 R1 I-IVϩV (casp), cDNA encoding rat IP 3 R1 flanked by the NheI and NotI sites at the 5Ј and 3Ј ends in pcDNA3.1 was used as the template. All PCR modifications described here were conducted using Pfu Ultra II Hotstart 2X Master Mix (Agilent), and only forward primers are shown here. To generate the construct coding for N-terminal and C-terminal fragments predicted to result from caspase cleavage of IP 3 R1 at the DEVD1891 consensus site, IP 3 R1 cDNA was modified by PCR (forward, 5Ј-GAAAGATGATGAAGTGGA-CTAGAATTCGCGGCCGCGCTAGCATGCGGGATGCCC-CATCCCGAA-3Ј). This modification introduced a stop codon after residue Asp-1891 and also a Kozak sequence and an initiation methionine in-frame with the sequence coding for the membrane fragment, designed to ensure efficient expression. To obtain a two-promoter vector (TPV) encoding both N-and C-terminal fragments, two-step ligation was performed (33). First, NheI-NotI IP 3 R1 I-IV (casp) was inserted into the TPV digested with NheI and PspOMI (NotI and PspOMI have compatible cohesive ends). Second, the NotI-NotI fragment coding IP 3 R1 V (casp) was inserted into the TPV that was digested with NotI. The TPV encoding IP 3 R1 I-IIϩIII-V (tryp), IP 3 R1 I-IIIϩIV-V (tryp), and IP 3 R1 I-VϩV (calp) were constructed in an identical manner using corresponding primers: forward 5Ј-GGCAGCAACGTGATGAGATAGGCGGCCGCGCTAG-CATGTCTATCCATGGAGTTGG-3Ј) for IP 3 R1 I-IIϩIII-V (tryp), forward 5Ј-CTGGCGGTTATCAGCCCGCTAGG-

Western blot analysis
Cells were harvested by centrifugation, washed once with ice-cold PBS, and solubilized in cell lysis buffer containing 10 mM Tris-HCl, 10 mM NaCl, 1 mM EGTA, 1 mM EDTA, 1 mM NaF, 20 mM Na 4 P 2 O 7 , 2 mM Na 3 VO 4 , 1% Triton X-100 (v/v), and 10% glycerol with a mixture of protease inhibitors. After 30 min on ice, cell lysates were precleared by centrifugation at 16,000 ϫ g for 10 min at 4°C. Cleared lysates were transferred into fresh tubes, and protein concentrations were measured using a D c protein assay kit (Bio-Rad). Protein were resolved on 5-7.5% SDS-PAGE gels, transferred to nitrocellulose, and processed for immunoblotting with the indicated primary antibodies and corresponding secondary antibodies. Proteins were detected using an Odyssey infrared imaging system (LI-COR Biosciences).

Fluorescence imaging assay
DT40 cells expressing defined IP 3 R constructs were loaded with 2 M Fura-2/AM on a glass coverslip mounted onto a Warner chamber at room temperature for 20 -30 min. Loaded cells were perfused with HEPES imaging buffer (137 mM NaCl, 4.7 mM KCl, 1.26 mM CaCl 2 , 1 mM Na 2 HPO 4 , 0.56 mM MgCl 2 ,10 mM HEPES, and 5.5 mM glucose (pH 7.4)) and stimulated with the indicated agonist. Ca 2ϩ imaging was performed using an inverted epifluorescence Nikon microscope with a ϫ40 oil immersion objective (NA ϭ 1.3). Cells were alternately excited at 340 and 380 nm, and emission was monitored at 505 nm. Images were captured every second with an exposure of 10 ms and 4 ϫ 4 binning using a digital camera (Cooke Sensicam QE) driven by TILL Photonics software.

Cell culture and plasmid transfection
DT40-3KO cells, a chicken B lymphocyte line with targeted deletion of the three endogenous IP 3 R isoforms, were grown in RPMI 1640 medium supplemented with 1% chicken serum, 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin at 39°C with 5% CO 2 . DT40-3KO cell transfection and generation of stable cell lines was performed as described previously using the Amaxa nucleofector (Lonza Laboratories). HEK-3KO cells were maintained in DMEM supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin at 37°C with 5% CO 2 . HEK-3KO cells were transfected using Lipofectamine 2000 (Invitrogen) following the protocol of the manufacturer.

FlexStation assay
DT40-3KO cells (5 ϫ 10 5 cells/well) expressing the receptor of interest were washed with PBS once, followed by incubation in imaging buffer containing 1% BSA and 5 M Fluo-2/AM for 1 h. Cells were then spun down, washed with imaging buffer containing 1% BSA, and seeded onto a 96-well plate that was precoated with 0.1% poly-L-lysine. The plate was spun at 500 ϫ g for 2 min and rested at room temperature for 15 min. During FlexStation recording, cells were treated with 20 M forskolin (final concentration) or the same volume of DMSO for 3 min, followed by stimulation with different concentrations of PAR2 agonist. For each type of IP 3 R or receptor fragments, a control response was defined as the amplitude of Ca 2ϩ response stimulated by 500 M PAR2 agonist following DMSO preincubation. The amplitudes of Ca 2ϩ responses for all other conditions were calculated and displayed as the percentage of the control response.

NFAT-GFP translocation assay
DT40-3KO cells expressing the desired receptor were transiently transfected with NFAT-GFP. 12 h after recovery, cells were transferred onto coverslips mounted in a Warner chamber. Cells were perfused with HEPES imaging buffer (137 mM NaCl, 4.7 mM KCl, 1.26 mM CaCl 2 ,1 mM Na 2 HPO 4 , 0.56 mM MgCl 2 ,10 mM HEPES, and 5.5 mM glucose (pH 7.4)) and stimulated with mouse ␣-chicken IgM. GFP imaging was recorded using an inverted epifluorescence Nikon microscope with a ϫ40 oil immersion objective. Cells were excited at 488 nm, and emission was monitored at 509 nm. Images were captured every second with an exposure of 10 ms and 2 ϫ 2 binning using a digital camera driven by TILL Photonics software.

Preparation of DT40 cell nuclei
Isolated DT40 nuclei were prepared by homogenization as described previously (64). The homogenization buffer contained 250 mM sucrose, 150 mM KCl, and 10 mM Tris (pH 7.5). Cells were washed and resuspended in homogenization buffer prior to nuclear isolation using a RZR 2021 homogenizer (Heidolph Instruments) with 25 strokes at 1200 rpm. A 3-l aliquot of nuclear suspension was placed in 3 ml of bath solution that contained 140 mM KCl, 10 mM HEPES, 500 M BAPTA and 246 nM free Ca 2ϩ , pH 7.1. Nuclei were allowed to adhere to a plastic culture dish for 10min prior to patching.

On-nuclei patch clamp experiments
Single InsP 3 R channel currents using Potassium ions as the charge carrier (i k ) were measured in the on nucleus patch clamp configuration using pCLAMP 9 and an Axopatch 200B amplifier (Molecular Devices, Sunnydale, CA, USA) as previously described (64). Pipette solution contained 140 mM KCl, 10 mM HEPES, 1 M InsP 3 , 5 mM ATP, and 200 nM free Ca 2ϩ (pH 7.1). Free [Ca 2ϩ ] was calculated using Max Chelator freeware and verified fluorometrically. Active caspase-3 and/or Z-VAD-fmk were included in the pipette solution for the corresponding experiments. Traces were consecutive 3-s sweeps recorded at Ϫ100 mV, sampled at 20 kHz, and filtered at 5 kHz. A minimum of 15 s of recordings was considered for data analyses. Pipette resistances were typically 20 megohms, and seal resistances were Ͼ5 gigaohms.

Data analysis
Single-channel openings were detected by half-threshold crossing criteria using the event detection protocol in Clampfit 9. We assumed that the number of channels in any particular nuclear patch is represented by the maximum number of discrete stacked events observed during the experiment. Even at low P o , stacking events were evident (data not shown). Only patches with one apparent channel were considered for analyses. P o , unitary current (i k ), and open and closed times were calculated using Clampfit 9 and Origin 6 software (Origin Lab, Northampton, MA). All-points current amplitude histograms were generated from the current records and fitted with a normal Gaussian probability distribution function. The coefficient of determination (R2) for every fit was Ͼ0.95. The