Disease-associated mutations in inositol 1,4,5-trisphosphate receptor subunits impair channel function

Abstract

The inositol 1,4,5-trisphosphate (IP₃) receptors (IP₃Rs), which form tetrameric channels, play pivotal roles in regulating the spatiotemporal patterns of intracellular calcium signals. Mutations in IP₃Rs have been increasingly associated with many debilitating human diseases such as ataxia, Gillespie syndrome, and generalized anhidrosis. However, how these mutations affect IP₃R function, and how the perturbation of as-sociated calcium signals contribute to the pathogenesis and severity of these diseases remains largely uncharacterized. Moreover, many of these diseases occur as the result of autosomal dominant inheritance, suggesting that WT and mutant subunits associate in heterotetrameric channels. How the in-corporation of different numbers of mutant subunits within the tetrameric channels affects its activities and results in different disease phenotypes is also unclear. In this report, we investigated representative disease-associated missense mutations to determine their effects on IP₃R channel activity. Additionally, we designed concatenated IP₃R constructs to create tetrameric channels with a predefined subunit composition to explore the functionality of heteromeric channels. Using calcium imaging techniques to assess IP₃R channel function, we observed that all the mutations studied resulted in severely attenuated Ca²⁺ release when expressed as homotetramers. However, some heterotetramers retained varied degrees of function dependent on the composition of the tetramer. Our findings suggest that the effect of mutations depends on the location of the mutation in the IP₃R structure, as well as on the stoichiometry of mutant subunits assembled within the tetrameric channel. These studies provide insight into the pathogenesis and penetrance of these devastating human diseases.

Inositol 1,4,5-trisphosphate (IP₃) receptor (IP₃R)-mediated calcium (Ca²⁺) release regulates many cellular activities including proliferation, secretion, division, contraction, and even apoptosis (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11). There are three IP₃R genes in the human genome (ITPR1, ITPR2, and ITPR3). The resulting monomeric isoforms of IP₃R (IP₃R1, IP₃R2, and IP₃R3) share ∼60–70% sequence homology and form homo- and heterotetrameric Ca²⁺ channels predominantly located in membranes of intracellular Ca²⁺ stores (3, 12, 13, 14, 15, 16, 17, 18). In response to various extracellular stimuli that induce phospholipase C-mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate, IP₃ is generated and diffuses through the cytosol to bind and subsequently activate IP₃R channels resulting in movement of Ca²⁺ down its concentration gradient into the cytosol (19, 20). Each monomeric IP₃R isoform is a ∼300 kDa protein and is generally divided into four functional domains (Fig. 1). The N-terminal suppressor domain (SD; 1-225 aa)1 (21) and ligand-binding domain (LBD; 226-578 aa) (22), the central regulatory and coupling domain (579-2232 aa) (23), and the C-terminal channel domain containing six membrane spanning regions and a cytosolic tail (2233-2710 aa) (24, 25, 26). It is believed that IP₃ binding to the N terminus of IP₃R triggers conformational changes that are transmitted intramolecularly or intermolecularly to open the channel pore located at the C terminus (27, 28, 29, 30); however, the precise mechanism of this gating is uncertain (31).

Figure 1.

Linear schematic of the four domains of the IP₃R and the disease-associated IP₃R mutations found in the LBD, regulatory and coupling, and C-terminal domains.A, disease-associated mutations found in the LBD (49,50, 51, 52,53, 68,69, 70,71). B, disease-associated mutations found in the regulatory and coupling domain (47–49, 52–54, 65, 72,73, 74, 75, 76, 77, 78, 79, 80, 81,82). C, disease-associated mutations found in the C-terminal domain (52,53, 54,55, 65, 83,84, 85, 86, 87, 88,89). Diseases associated with mutations in at least one isoform of IP₃R include SCA29 (blue), Gillespie syndrome (orange), anhidrosis (pink), neuropathy (green), pontocerebellar hypoplasia (dark blue), spinocerebellar ataxia 15/16 (light blue), head and neck squamous cell carcinoma (red), Sézary Syndrome (purple), and familial isolated primary hyperparathyroidism (gray).

IP₃R genes are ubiquitously expressed, although various tissues and cell types often preferentially express one subtype or a combination of subtypes (32, 33, 34, 35). Despite the overlapping expression patterns, these receptors can mediate distinct functions depending on the cell's physiological context. IP₃R1 is highly enriched in the nervous system, whereas IP₃R2 and IP₃R3 are predominately expressed in the periphery including the digestive system and secretory cells of the exocrine system (36, 37, 38). Studies have shown that IP₃R1 is essential for normal nervous system development as mice lacking IP₃R1 often die in utero or suffer from severe ataxia and seizure prior to death very early in postnatal life (39, 40, 41). Mice lacking either IP₃R2 or IP₃R3 appear to develop only subtle phenotypes. However, mice lacking both IP₃R2 and IP₃R3 suffer from severe salivary and pancreatic insufficiencies and die after weaning due to inability to assimilate macronutrients from dry food (42).

Many devastating human diseases have been associated with abnormal Ca²⁺ homeostasis. Indeed, there is a growing body of literature illustrating that diseases, such as Huntington's disease, schizophrenia, ALS, Parkinson's disease, Alzheimer's disease, and spinocerebellar ataxia (SCA), are all accompanied by dysregulated Ca²⁺ signals (43, 44, 45, 46). Although the underlying etiologies of these diseases are heterogeneous, many studies have shown that perturbed IP₃R signaling is among the contributing or causal factors. Mutations in IP₃R genes have been associated with several human diseases, including SCA (47, 48, 49, 50, 51, 52, 53), Gillespie syndrome (GS) (52, 54), and anhidrosis (55). These mutations have generally been identified through whole exome sequencing and linkage analysis in a clinical setting. Nevertheless, the molecular and cellular consequences of many of these mutations, as well as the mechanism whereby individual mutations lead to the clinical manifestations, remain largely unknown. As mutations have been identified in all four functional domains of the IP₃R, it is conceivable that mutations can interfere with various aspects of IP₃R biology including IP₃ binding, allosteric regulation by various cellular factors, ion permeation, protein folding, stability, and subcellular localization. It is likely that in patients, tetrameric IP₃R channels are assembled from a combination of both WT and mutant subunits and given the autosomal dominant inheritance of the majority of cases, the incorporation of mutant subunits likely results in dominant-negative activity. Nevertheless, how the specific stoichiometry of the resultant heterotetrameric channels might affect function and lead to different clinical manifestations and disease penetrance have not been explored.

In this study, we have examined the consequences of a representative set of missense mutations on IP₃R channel function that are illustrative of mutations in distinct functional domains of the IP₃R and of differing modes of inheritance. Missense mutations of residues Arg-269 (R269W) (50, 51, 52, 53) and Asn-602 (N602D) (47, 48, 52) both lead to multiple cases of autosomal dominant, early-onset, nonprogressive ataxia, even though they are located in the LBD and regulatory/coupling domain, respectively (Fig. 1, A and B). In contrast, the G2498S mutation, located in the C-terminal selectivity filter of the IP₃R2 channel pore (Fig. 1C), is inherited in an autosomal recessive fashion such that individuals homozygous for the mutation suffered from generalized anhidrosis, whereas their family members heterozygous for the mutation were asymptomatic for the disease (55). We utilized two heterologous expression systems, which are unambiguously IP₃R null (56, 57) to directly investigate how expression of these mutant channels affect channel expression and localization, tetramer formation, IP₃ binding, channel gating, Ca²⁺ store content, and Ca²⁺ release activities. To investigate how incorporation of WT and disease-associated mutant subunits into the assembled receptor alters channel function, we also used a concatenated subunit approach that allows for a defined subunit stoichiometry. Our data indicate that the impact of missense mutations is domain dependent and markedly impacted by the stoichiometry of incorporation into the functional heterotetrameric channel. These studies provide a framework to understand the molecular basis of human IP₃R pathologies and offer novel insight into roles of key residues in IP₃R channel activities.

Results

IP₃R1 is highly expressed in the nervous system (35) and mutations result in a broad range of disease phenotypes: ataxia alone or ataxia with cognitive difficulties, iris hypoplasia and/or cerebellar hypotrophy. Indeed, there is a growing list of missense (47, 48, 49, 50, 51, 52, 53, 54, 55), nonsense, insertion/deletion, and splice mutations (58, 59) scattered throughout the IP₃R1 gene (Fig. 1). Nevertheless, how IP₃R1 mutations affect channel activity and how these mutations lead to pathogenesis remains to be elucidated. Here we investigated in detail, the functional consequences of three representative mutations located in different functional domains of IP₃R following expression of homotetrameric or defined heterotetrameric mutant channels in an IP₃R null background.

Mutation of R269W attenuates IP₃R activity by reducing IP₃ binding

R269W has been identified in patients diagnosed with autosomal dominant, nonprogressive congenital ataxia (NPCA) (50), spinocerebellar ataxia 29 (SCA29) (52), ataxic cerebral palsy (51), and early-onset ataxia (EOA) (53). These patients exhibit nonprogressive early-onset ataxia, hypotonia, and developmental delays (Table S1). Residue Arg-269 is located in the ligand-binding core of the IP₃R1 (Table S2) that also constitutes the IP₃R-binding protein released with inositol 1,4,5-trisphosphate (IRBIT)-binding domain (60, 61). This mutation (R269W) substitutes an evolutionarily conserved (Fig. 2A), positively charged, arginine residue, with a large hydrophobic, aromatic, tryptophan residue (Fig. 2B). This residue has been previously identified as one of the 12 positively charged amino acids absolutely required for IP₃ binding and early studies have demonstrated that mutation of arginine to glutamine at this site (R269Q) resulted in an 80% loss of IP₃ binding (22, 62). We therefore investigated the consequences of the R269W mutation on IP₃ binding. WT, rat IP₃R1 (WT rIP₃R1), and rat cDNA engineered to express the R269W mutation (rIP₃R1^RW) were stably transfected into DT40-3KO cells (56) (Fig. 2C). [³H]IP₃ binding was measured by competitive binding assays. When exogenously expressed in DT40-3KO, rIP₃R1^RW IP₃-binding activity was severely compromised compared with WT rIP₃R1 (Fig. 2D). Total specific binding was reduced by 60%, whereas the apparent EC₅₀ for binding was not markedly altered under these conditions. This latter observation is surprising but may reflect severe alterations in the binding kinetics (either an alteration of on- or off-rate) such that equilibrium binding is not reached for the rIP₃R1^RW mutant under the same conditions as the WT rIP₃R1.

Figure 2.

rIP3R1RW is nonfunctional when expressed in DT40-3KO due to decreased IP3 binding.A, Arg-269 (red) is conserved among all three human IP₃R isoforms and evolutionarily conserved in the IP₃R1 isoform. B, chimera (PDB code 3JAV) was used to visualize WT Arg-269 (yellow, left panel) and R269W mutant Trp-269 (yellow, right panel) interacting with IP₃ in the ligand binding cleft. C, WT rIP₃R1 and mutant rIP₃R1^RW cell lines were generated in the IP₃R-null DT40-3KO cells and Western blotted. D, binding of 2.5 nm [³H]IP₃ to WT rIP₃R1 and rIP₃R1^RW in the presence of increasing concentrations (0 nm, 1 nm, 3 nm, 10 nm, 30 nm, 100 nm, 300 nm, 1 μm, 3 μm, 5 μm, 10 μm, and 50 μm) of cold IP₃ in a competitive radioligand-binding assay. Data are mean ± S.E. of three (n = 3) independent experiments. E, representative traces show Ca²⁺ signals of IP₃R-null DT40-3KO cells (blue), WT rIP₃R1 (green), and rIP₃R1^RW (red) in response to trypsin (500 nm) when loaded with Fura-2/AM. F, scatter plots summarizing change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in E. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent 5th and 95th percentiles and mean is represented by colored circles. G, stacked bar graph summarizing the percentage of amplitudes from F, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in E. Unless otherwise stated, all data above comes from at least n = 3 experiments. ***, p < 0.001 when compared with WT rIP₃R1 cell line and ###, p < 0.001 when compared with DT40-3KO cell line; one-way ANOVA with Tukey's test performed in F (F_19,4699= 753.0, p < 0.0001) and G (F_19,50 = 284.6, p < 0.0001).

Next, we assessed the ability of rIP₃R1^RW to support IP₃-induced Ca²⁺ release when stably expressed in DT40-3KO cells. In single cell imaging experiments, stimulation with trypsin, which activates the Gα_q-coupled protease-activated receptor 2 (PAR2), generates IP₃ and results in Ca²⁺ release in cells expressing rIP₃R1 but not in IP₃R null, DT40-3KO (Fig. 2E) (56, 57). Similarly, whereas stimulation of cells expressing rIP₃R1 receptors mediated robust Ca²⁺ signals, cells expressing rIP₃R1^RW did not exhibit any measurable response to a maximal concentration of trypsin (Fig. 2E). The pooled data illustrates that despite retaining modest residual binding activity, in terms of the magnitude of signal (Fig. 2F), and the number of cells that respond above a defined threshold (Fig. 2G), rIP₃R1^RW is refractory to stimulation in DT40-3KO cells. Additionally, multiple, distinct clonal rIP₃R1^RW cell lines generated in DT40-3KO cells exhibited similar de-creases in amplitude (data not shown). This lack of response may be related to the requirement of each IP₃R monomer to be occupied by IP₃ for activation (57), which given the reduced binding, would be predicted to limit this occurrence.

In support of the idea that assemblies of mutant (binding deficient) IP₃R1 and WT IP₃R1 are generated in cells, we performed co-immunoprecipitation experiments in cells transfected with a mCherry-tagged rIP₃R1 and IP₃R1^RW. The IP₃R1^RW mutant was readily detected in samples immunoprecipitated with a mCherry antibody (Fig. S1A) and thus heterotetrameric channels containing both WT and mutant subunits were formed.

A prediction from our previous data (57) is that heterotetramers consisting of rIP₃R1 and rIP₃R1^RW would also fail to respond to stimulation. To test this idea, we generated cDNA encoding concatenated dimers consisting of rIP₃R1 and rIP₃R1^RW expressed as a single polypeptide chain of IP₃R1-IP₃R1^RW (R1/R1^RW) or IP₃R1^RW-IP₃R1 (R1^RW/R1) and stably expressed these constructs in DT40-3KO cells (Fig. 3A). Dimers are assembled by the cell to produce tetrameric channels consisting of two IP₃R1 and two IP₃R1^RW subunits. Cells expressing concatenated dimers encoding only rIP₃R1 (R1/R1) responded to PAR2 stimulation with robust Ca²⁺ signals, as previously reported (38, 57), whereas cells expressing concatenated dimers encoding rIP₃R1^RW (R1^RW/R1^RW) failed to respond (Fig. 3B). Additionally, cells expressing R1/R1^RW or R1^RW/R1 dimers also failed to respond to trypsin stimulation (Fig. 3B). The pooled data illustrates that despite some R1/R1^RW or R1^RW/R1 dimers exhibiting a very small increase in the magnitude of the signal (Fig. 3C), the number of cells that respond above to a defined threshold remained not significantly different from the null DT40-3KO cell line (Fig. 3D). Additional R1/R1^RW, R1^RW/R1, and R1^RW/R1^RW dimeric clonal cell lines generated in DT40-3KO cells exhibited similar decreases in amplitude and percentage of cells responding above a specific threshold (data not shown). Consistent with these observations, IP₃ also failed to evoke Ca²⁺ release in permeabilized DT40-3KO cells stably expressing channels containing two or more rIP₃R1^RW subunits (Fig. 3E). These data indicate that tetramers containing mutant subunits likely are poorly functional. Thus, the implication is that patients with heterozygous expression of the mutant protein, only tetramers consisting of nonmutant IP₃R1 subunits would support significant function.

Figure 3.

Heterotetramers of WT rIP₃R1 and rIP₃R1^RW are predominantly nonfunctional when expressed in DT40-3KO.A, monomeric WT rIP₃R1 and mutant rIP₃R1^RW cell lines, as well as dimeric R1/R1, R1/R1^RW, R1^RW/R1, and R1^RW/R1^RW cell lines were generated in the IP₃R-null DT40-3KO cells and Western blotted. B, representative traces show Ca²⁺ signals of IP₃R-null DT40-3KO cells (blue), WT rIP₃R1 (green), and rIP₃R1^RW (red), R1/R1 (dark green), R1/R1^RW (orange), R1^RW/R1 (purple), and R1^RW/R1^RW (dark red) in response to trypsin (500 nm) when loaded with Fura-2/AM. C, scatter plots summarizing change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in B. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent 5th and 95th percentiles and mean is represented by colored circles. D, stacked bar graph summarizing the percentage of amplitudes from C, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in B. E, traces show Ca²⁺ signals of β-escin permeabilized WT rIP₃R1 (green), and rIP₃R1^RW (red), R1/R1 (dark green), R1/R1^RW (orange), R1^RW/R1 (purple), and R1^RW/R1^RW (dark red) cell lines in response to IP₃ (30 μm) when loaded with Mag-Fura-2/AM. Data are mean ± S.E. of three (n = 3) independent experiments. Data for DT40-3KO, rIP₃R1, and rIP₃R1^RW in B–D were from Fig. 2. Unless otherwise stated, all data above comes from at least n = 3 experiments. ***, p < 0.001 when compared with WT rIP₃R1 cell line and ###, p < 0.001 when compared with DT40-3KO cell line; one-way ANOVA with Tukey's test was performed in C (F_19,4699 = 753.0, p < 0.0001) and D (F_19,50 = 284.6, p < 0.0001).

Although the DT40-3KO chicken lymphocyte cell line has been used for last two decades to probe IP₃R biology, we have recently used CRISPR/Cas9 technology to disrupt all three IP₃R genes in human embryonic kidney (HEK293) cells to generate IP₃R null HEK293 cells (HEK-3KO) (57). Tangible advantages of these cells include their mammalian origin, with presumably more relevant cellular milieu in terms of potential regulatory factors, and that HEK293 cells can be easily maintained and transfected. Therefore, we next asked if the R269W mutant behaves similarly in HEK-3KO compared with DT40 cells. Stable HEK-3KO cell lines were generated expressing WT rIP₃R1, rIP₃R1^RW, or hIP₃R1^RW (Fig. 4A). Stable expression resulted in an obvious reticular pattern of expression, not obviously different from the localization of WT IP₃R1 (Fig. 4B). Expression of rIP₃R1^RW did not result in alteration of the apparent content of the ER Ca²⁺ store, as assessed by the Ca²⁺ released following exposure to a SERCA pump inhibitor in a Ca²⁺-free extracellular buffer (Fig. S2, A and B). See supplemental experimental procedures. In single cell assays, stimulation of HEK-3KO expressing WT rIP₃R1 with 5, 50, or 500 nm trypsin generated robust Ca²⁺ release (Fig. 4, C and D) with nearly 100% of cells responding (Fig. 4E). Notably, stimulation of cells expressing rIP₃R1^RW also resulted in measurable Ca²⁺ increases (Fig. 4, C and D), albeit, at much lower maximal amplitudes and with a lower percentage of responding cells responding (Fig. 4E). A small increase in amplitude was likewise seen in other clonal cell lines expressing rIP₃R1^RW and hIP₃R^RW in HEK-3KO cells (data not shown). These findings were recapitulated in populations of cells where [Ca²⁺] changes were monitored in response to varying [trypsin] using a Flexstation3 plate reader (Fig. 4F). Additionally, the corresponding mutation in the human IP₃R1 (hIP₃R1^RW) stably expressed in HEK-3KO cells displayed a similar marked loss of function when compared with trypsin-induced release evoked in HEK-3KO cells stably expressing hIP₃R1 (Fig. 4F). The seemingly disparate findings obtained upon expression of R269W mutations in DT40-3KO and HEK-3KO could reflect differences in stable expression level of the mutant IP₃R in HEK-3KO cells compared with DT40-3KOs (Fig. S3A). In total, these findings show in two expression systems that R269W exhibits compromised IP₃ binding that results in severely reduced channel activity.

Figure 4.

rIP₃R1^RW is poorly functional when expressed in HEK-3KO.A, multiple WT rIP₃R1, WT hIP₃R1, mutant rIP₃R1^RW, mutant rIP₃R1^ND, and mutant hIP₃R1^RW cell lines were generated in the IP₃R-null HEK-3KO cells and Western blotted. B, immunocytochemistry for HEK-3KO cell lines expressing either WT rIP₃R1 (left) or mutant rIP₃R1^RW (right). Top, IP₃R1 detection (green); middle, DAPI detection (blue); bottom, merged images of IP₃R1 and DAPI. Scale bars, 10 μm. C, representative traces show Ca²⁺ signals of IP₃R-null HEK-3KO cells (blue), WT rIP₃R1 (green), and rIP₃R1^RW (red) in response to trypsin (500 nm) when loaded with Fura-2/AM. D, scatter plots summarizing change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in C when treated with 5, 50, and 500 nm of trypsin. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent 5th and 95th percentiles and mean is represented by colored circles. E, stacked bar graph summarizing the percentage of amplitudes from D, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in C. F, dose-response curve showing Ca²⁺ response of Fura-2/AM-loaded WT rIP₃R1, WT hIP₃R1, rIP₃R1^RW, and hIP₃R1^RW cells when treated with increasing concentrations (1 nm, 10 nm, 30 nm, 100 nm, 300 nm, 1 μm, and 3 μm) of trypsin using a Flexstation3 96-well–plate reader. Data are mean ± S.E. of three (n = 3) independent experiments. ***, p < 0.001 when compared with WT rIP₃R1 cell line and ###, p < 0.001 when compared with HEK-3KO cell line; one-way ANOVA with Tukey's test was performed in D (F_15,1352 = 407.1, p < 0.0001) and E (F_15,34 = 108.5, p < 0.0001). Unless otherwise stated, all data above comes from at least n = 3 experiments.

Mutation of N602D attenuates IP₃R1 activity without altering IP₃ binding

The N602D mutation in IP₃R1 has been identified in two unrelated families and is associated with diagnoses of ataxic cerebral palsy without cerebral atrophy (48) and autosomal dominant congenital nonprogressive spinocerebellar ataxia (47, 52) (Table S1). Asn-602 is conserved among the three IP₃R subtypes and across many species (Fig. 5A). The residue is located just outside of the LBD, but formally within the IRBIT-binding domain (Table S3) and serves to connect two α-helical regions of the receptor (Fig. 5B). The mutation introduces a negatively charged residue at this site, replacing an asparagine residue with a polar neutral side chain with an aspartate residue harboring a polar acidic side chain. DT40-3KO cells stably expressing WT rIP₃R1 or rIP₃R1 expressing N602D (rIP₃R1^ND) mutations were generated (Fig. 5C). To establish the effects of this mutation on channel function, we first assessed IP₃ binding to the mutant proteins. In a competitive binding assay, rIP₃R1^ND exhibited binding comparable with that of WT rIP₃R1, suggesting that the mutation does not interfere with IP₃ binding despite the proximity of the mutation to the LBD (Fig. 5D). To examine the consequences of the mutation on Ca²⁺ release activities, cells expressing WT rIP₃R1 or rIP₃R1^ND were stimulated with the PAR agonist, trypsin. Trypsin did not elicit any measurable [Ca²⁺] increase in cells expressing rIP₃R1^ND (Fig. 5E). The pooled data demonstrates that in terms of the magnitude of signal (Fig. 5F), and the number of cells that respond above a defined threshold (Fig. 5G), rIP₃R1^ND is refractory to stimulation in DT40-3KO cells.

Figure 5.

rIP₃R1^ND is nonfunctional when expressed in DT40-3KO.A, Asn-602 (red) and Thr-594 (blue) are conserved among all three human IP₃R isoforms and evolutionarily conserved in the IP₃R1 isoform. B, chimera (PDB 6MU1) was used to visualize WT Asn-602 (yellow) between two α-helical regions in the ARM1 domain (red) and adjacent to the β-TF1 and β-TF2 domains in the N terminus (blue). C, WT rIP₃R1 and several mutant rIP₃R1^ND cell lines generated in the IP₃R-null DT40-3KO cells were Western blotted. D, binding of 2.5 nm [³H]IP₃ to WT rIP₃R1 (green) and rIP₃R1^ND (red) in the presence of increasing concentrations (0 nm, 1 nm, 3 nm, 10 nm, 30 nm, 100 nm, 300 nm, 1 μm, 3 μm, 5 μm, 10 μm, and 50 μm) of cold IP₃ in a competitive radioligand binding assay. Data are mean ± S.E. of three (n = 3) independent experiments. E, representative traces show Ca²⁺ signals of IP3R-null DT40-3KO cells (blue), WT rIP₃R1 (green), and rIP₃R1^ND (red) in response to trypsin (500 nm) when loaded with Fura-2/AM. F, scatter plots summarizing change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in E. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent 5th and 95th percentiles and mean is represented by colored circles. G, stacked bar graph summarizing the percentage of amplitudes from F, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in E. Unless otherwise stated, all data above comes from at least n = 3 experiments. ***, p < 0.001 when compared with WT rIP3R1 cell line and ###, p < 0.001 when compared with DT40-3KO cell line; one-way ANOVA with Tukey's test was performed in F (F_7,3070 = 525.9, p < 0.0001) and G (F_9,19 = 177.2, p < 0.0001).

Given, that the rIP₃R1^RW mutation exhibited more activity when stably expressed in HEK-3KO cells when compared with DT40-3KOs, we next investigated whether rIP₃R1^ND was similarly more active when stably expressed in this mammalian cell line (Fig. 6A). Expression of rIP₃R1^ND in HEK-3KO cells did not result in alteration of the apparent content of the ER Ca²⁺ store (Fig. S2C) and resulted in an obvious reticular pattern of expression (Fig. S4). Consistent, with data obtained in cells expressing rIP₃R1^RW, stimulation of single HEK-3KO stably expressing rIP₃R1^ND with various [trypsin] resulted in modest Ca²⁺ release (Fig. 6B) manifested as reduced maximal initial release (Fig. 6C) and a lower percentage of cells responding at all [trypsin] (Fig. 6D). Once again, differences observed when the mutant is stably expressed in HEK-3KO cells compared with DT40-3KOs may result from differences in mutant protein expression in the two cell lines (Fig. S3B). A mutation in a neighboring residue in the same region, Thr-594 (T594I) (Fig. S5A), which is associated with infantile onset SCA (49), exhibited a similar decreased Ca²⁺ release (Fig. S5D) and percentage of cells responding (Fig. S5E) when stably expressed in HEK-3KO cells (Fig. S5B) and stimulated with a maximal concentration of trypsin (Fig. S5C). Reflecting the behavior at the single cell level, in population assays, the concentration versus response relationship for trypsin-induced [Ca²⁺] release reflected a decrease in both efficacy and sensitivity of stimulation (Fig. 6E and Fig. S5F). Thus, although competent to bind IP₃, homomeric IP₃R1 harboring mutations in this region are largely nonfunctional.

Figure 6.

rIP₃R1^ND is poorly functional when expressed in HEK-3KO.A, multiple WT rIP3R1 and mutant rIP₃R1^ND cell lines were generated in the IP₃R-null HEK-3KO cells and Western blotted. B, representative traces show Ca²⁺ signals of IP3R-null HEK-3KO cells (blue), WT rIP₃R1 (green), and rIP₃R1^ND (red) in response to trypsin (500 nm) when loaded with Fura-2/AM. C, scatter plots summarizing change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in B when treated with 0.5, 1, 10, and 500 nm trypsin. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent 5th and 95th percentiles and mean is represented by colored circles. D, stacked bar graph summarizing the percentage of amplitudes from C, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in B. E, dose-response curve showing Ca²⁺ response of Fura-2/AM-loaded WT rIP₃R1 and rIP₃R1^ND cells when treated with increasing concentrations (0.5 nm, 1 nm, 3 nm, 10 nm, 30 nm, 100 nm, 300 nm, 500 nm, 1 μm, 2.5 μm, and 5 μm) of trypsin using a Flexstation3 96-well–plate reader. Data are mean ± S.E. of three (n = 3) independent experiments. **, p < 0.01 and ***, p < 0.001 when compared with WT rIP₃R1 cell line and ###, p < 0.001 when compared with HEK-3KO cell line; one-way ANOVA with Tukey's test was performed in C (F_10,791 = 532.1, p < 0.0001) and D (F_10,22 = 108.6, p < 0.0001). Unless otherwise stated, all data above comes from at least n = 3 experiments.

We next investigated the impact of the presence of variable numbers of mutant rIP₃R1^ND within a heterotetrametric assembly of subunits, which likely represents the situation in patients heterozygous for the mutation. Once again, the IP₃R1 protein was readily detected in samples immunoprecipitated with a mCherry antibody suggesting heterotetrameric channels containing both WT and mutant channels are formed (Fig. S1A). Concatenated dimers were generated containing one WT rIP₃R1 and one rIP₃R1^ND (rIP3R1-rIP₃R1^ND: R1/R1^ND or rIP₃R1^ND-rIP₃R1: R1/R1^ND) and stably expressed in DT40-3KO cells (Fig. 7A). As shown previously, tetramers expressed from either monomeric rIP₃R1 (WT rIP₃R1) or dimeric rIP₃R1 (R1/R1) were capable of supporting robust PAR2-mediated Ca²⁺ signaling, however, tetramers assembled from R1^ND/R1 or R1/R1^ND were completely refractory to stimulation at maximal [trypsin] (Fig. 7, B–D). These data indicate that tetramers harboring two mutant IP₃R1^ND subunits are likely severely compromised even given the modest activity of this mutant in HEK cells. To investigate if tetramers harboring only a single mutant rIP₃R1^ND retain any activity, concatenated tetramers were generated with three WT rIP₃R1 and one rIP₃R1^ND (rIP₃R1-rIP₃R1-rIP₃R1-rIP₃R1^ND: R1/R1/R1/R1^ND) and stably expressed in DT40-3KO cells (Fig. 7E). Notably, some cells expressing R1/R1/R1/R1^ND tetramers retained a comparable ability to support Ca²⁺ release to a concatenated tetramer assembled from WT rIP₃R1 in terms of both trypsin (Fig. 7, F–H) and IP₃-evoked Ca²⁺ release (Fig. 7I). In total these data demonstrate that rIP₃R^ND retains modest activity and that tetramers harboring this gating mutant are not necessarily completely disabled.

Figure 7.

Heterotetramers of WT rIP₃R1 and rIP₃R1^ND are predominantly nonfunctional when expressed in DT40-3KO.A, monomeric WT rIP₃R1 and mutant rIP₃R1^ND cell lines, as well as dimeric R1/R1, R1/R1^ND, and R1^ND/R1 cell lines were generated in the IP₃R-null DT40-3KO cells and Western blotted. B, representative traces show Ca²⁺ signals of IP₃R-null DT40-3KO cells (blue), WT rIP₃R1 (green), and rIP₃R1^ND (red), R1/R1 (dark green), R1/R1^ND (orange), and R1^ND/R1 (purple) in response to trypsin (500 nm) when loaded with Fura-2/AM. C, scatter plots summarizing the change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in B. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent the 5th and 95th percentiles and the mean is represented by colored circles. D, stacked bar graph summarizing the percentage of amplitudes from C, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in B. E, monomeric WT rIP3R1, dimeric R1/R1, trimeric R1/R1/R1, and tetrameric R1/R1/R1/ND cell lines generated in the IP3R-null DT40-3KO cells were Western blotted. F, multiple representative traces show Ca²⁺ signals of IP3R-null DT40-3KO cells expressing R1/R1/R1/R1 (green) and R1/R1/R1/ND (red) tetramers in response to trypsin (500 nm) when loaded with Fura-2/AM. G, scatter plots summarizing change in amplitude for experiments similar to those shown in F when treated with 500 nm trypsin. H, stacked bar graph summarizing the percentage of amplitudes from G, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in F. I, traces show Ca²⁺ signals of β-escin permeabilized IP3R-null DT40-3KO cells (blue), R1/R1/R1/R1 (green), and R1/R1/R1/ND (red) cell lines in response to IP₃ (30 μm) when loaded with Mag-Fura-2/AM. Data are mean ± S.E. of three (n = 3) independent experiments. Data for rIP3R1 and rIP₃R1^ND in B–D and DT40-3KO in B–H came from Fig. 5. Unless otherwise stated, all data above comes from at least n = 3 experiments. ***, p < 0.001 when compared with WT rIP3R1 cell line and ###, p < 0.001 when compared with DT40-3KO cell line; one-way ANOVA with Tukey's test was performed in C (F_7,3070 = 525.9, p < 0.0001), D (F_9,19 = 177.2, p < 0.0001), G (F_7,3070 = 525.9, p < 0.0001), and H (F_9,19 = 177.2, p < 0.0001).

Mutation of G2498S in the channel pore attenuates IP₃R2 activity

We next investigated a missense mutation (G2498S) in IP₃R2, identified in a family with generalized anhidrosis (55). IP₃R2 is the predominant IP₃R isoform expressed in eccrine sweat glands. Homozygous individuals were reported to have morphologically normal eccrine sweat glands but suffer from severe hyperthermia resulting from an absence of perspiration in response to physiological or pharmacological stimuli (Table S1). Notably, however, heterozygous siblings did not exhibit any symptoms of disease and thus this mutation represents a rare example of autosomal recessive inheritance of disease symptoms associated with IP₃R mutations. The Gly-2498 residue in IP₃R2 is located in a motif (GGGXG) conserved in all IP₃R subtypes (Fig. 8A) as well as the superfamily of cation channels and forms the selectivity filter of the ion conducting pathway (Fig. 8B). This ion permeability pathway is a hot spot of other disease-associated mutations in multiple IP₃R types (Table S4). This mutation (G2498S; designated mIP₃R2^GS) substitutes an evolutionarily conserved, nonpolar, hydrophobic, glycine residue, with a polar, hydrophillic serine residue and thus might reasonably be expected to disrupt the pore of the channel. Consistent with previous studies, when stably expressed in DT40-3KO cells (Fig. 8C), IP₃ binding to mIP₃R2^GS was comparable with WT mIP₃R2 (Fig. 8D). Nevertheless, trypsin stimulation failed to elicit any Ca²⁺ release in these cells (Fig. 8, E–G). Again, stable expression of mIP₃R2^GS in HEK-3KO cells did not alter the apparent store content of the ER (Fig. S2D) and resulted in an obvious reticular pattern of expression (Fig. S4B). However, notably, in contrast to the previous mutations investigated, when expressed in HEK-3KO cells (Fig. 9A) mIP₃R2^GS was unable to support any measurable Ca²⁺ release (Fig. 9, B–E). Similar results were seen in other clonal lines expressing the G2498S mutations (data not shown), as well as HEK-3KO cells stably expressing a mutation in hIP₃R1 G2506R (Fig. S6, A–F), corresponding to the neighboring glycine in the GGGXG motif. This mutation is associated with spinocerebellar ataxia and Gillespie syndrome. Again, differences in expression in HEK-3KO cells versus DT40-3KO cells may in part explain these results (Fig. S3C). Overall, these data confirm that homozygous expression of mIP₃R2^GS results in a complete abrogation of IP₃R activity and is consistent with the severe phenotype in individuals homozygous for the mutation.

Figure 8.

mIP₃R2^GS is nonfunctional when expressed in DT40-3KO.A, Gly-2498 (red) is conserved among all three human IP3R isoforms and evolutionarily conserved in the IP₃R2 isoform. B, chimera (PDB 6MU1) was used to visualize WT Gly-2498 (yellow) in the selectivity filter (blue) of two monomers, just prior to the 6th transmembrane (TM) domain (red). C, WT mIP₃R2 and mutant mIP₃R2^GS cell lines were generated in the IP₃R-null DT40-3KO cells and Western blotted. D, binding of 2.5 nm [³H]IP₃ to WT mIP₃R2 and mIP₃R2^GS in the presence of a maximal concentration of 50 μm cold IP₃ in a competitive radioligand binding assay (p = 0.9077). Data are mean ± S.E. of three (n = 3) independent experiments. E, representative traces show Ca²⁺ signals of IP3R-null DT40-3KO cells (blue), WT mIP₃R2 (green), and mIP₃R2^GS (red) in response to trypsin (500 nm) when loaded with Fura-2/AM. F, scatter plots summarizing change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in E. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent 5th and 95th percentiles and mean is represented by colored circles. G, stacked bar graph summarizing the percentage of amplitudes from F, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in E. Unless otherwise stated, all data above comes from at least n = 3 experiments. ***, p < 0.001 when compared with WT rIP3R1 cell line and ###, p < 0.001 when compared with HEK-3KO cell line; unpaired t test was performed in D and one-way ANOVA with Tukey's test was performed in F (F_9,1568 = 437.7, p < 0.0001) and G (F_9,27 = 61.27, p < 0.0001).

Figure 9.

mIP₃R2^GS is predominantly nonfunctional when expressed in HEK-3KO.A, multiple WT mIP₃R2 and mutant mIP₃R2^GS cell lines were generated in the IP₃R-null HEK-3KO cells and Western blotted. B, representative traces show Ca²⁺ signals of IP₃R-null HEK-3KO cells (blue), WT mIP₃R2 (green), and mIP₃R2^GS (red) in response to trypsin (500 nm) when loaded with Fura-2/AM. C, scatter plots summarizing change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in B when treated with 500 nm trypsin. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent 5th and 95th percentiles and mean is represented by colored circles. D, stacked bar graph summarizing the percentage of amplitudes from C, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in B. E, dose-response curve showing Ca²⁺ response of Fura-2/AM-loaded WT mIP₃R2 and mIP₃R2^GS cells when treated with increasing concentrations (1 nm, 10 nm, 30 nm, 100 nm, 300 nm, 1 μm, and 3 μm) of trypsin using a Flexstation3 96-well–plate reader. Data are mean ± S.E. of three (n = 3) independent experiments. ***, p < 0.001 when compared with WT rIP₃R1 cell line and ###, p < 0.001 when compared with HEK-3KO cell line; one-way ANOVA with Tukey's test was performed in C (F_6,809 = 707.1, p < 0.0001) and D (F_6,17 = 335.6, p < 0.0001). Unless otherwise stated, all data above comes from at least n = 3 experiments.

As the heterozygous siblings of the affected individuals were seemingly not markedly impacted by harboring the mutation, we next investigated the behavior of tetrameric channels assembled from WT and mIP₃R2^GS. Immunoprecipitation once again confirmed that mutated mIP₃R2^GS monomers could interact with other monomeric subunits (Fig. S1, B and C). The current structural model based on the cryo-EM structure of IP₃R1 (PDB codes 6MU1 and 6MU2) indicates that the Ser substitution for Gly-2547 (Gly-2498 in IP₃R2) may form H bonds with the neighboring Arg-2544 residue (Fig. 10A), potentially resulting in constrained movement and restricted opening of the pore in a homotetramer harboring this mutant. Concatenated cDNA encoding either homodimeric mIP₃R2-mIP₃R2 (R2/R2) and mIP₃R2^GS-mIP₃R2^GS (R2^GS/R2^GS) or heterodimeric IP₃R2^GS-IP₃R2 (R2^GS/R2 and R2/R2^GS) were generated and stably expressed in DT40-3KO cells (Fig. 10B). As shown previously (38), trypsin stimulation of cells expressing both mIP₃R2 and R2/R2 elicited vigorous Ca²⁺ release (Fig. 10C). Although cells expressing R2^GS/R2^GS were refractory to stimulation, surprisingly, stimulation of cells expressing R2^GS/R2 or R2/R2^GS exhibited robust Ca²⁺ release (Fig. 10C and pooled data in Fig. 10, D and F). This partial channel functionality was confirmed through stimulation of additional cell lines expressing R2^GS/R2 or R2/R2^GS (data not shown). Similarly, exposure to IP₃ in permeabilized cells also evoked robust Ca²⁺ release in WT homotetramers, intermediate Ca²⁺ release in heterotetramers, and no Ca²⁺ release in mutant homotetramers (Fig. 10F). These studies therefore reveal that heterotetrameric channels composed of two WT and two G2498S mutant subunits retain significant function as IP₃-sensitive channels. Our data suggest that substitution of two or fewer Gly-Ser results in less restriction of pore movement required for gating, such that significant Ca²⁺ flux can occur that likely supports activity sufficient to render heterozygous individuals asymptomatic.

Figure 10.

Heterotetramers of WT mIP₃R2 and mIP₃R2^GS retain partial functionality when expressed in DT40-3KO.A, the current IP₃R structural model based on the cryo-EM structure of IP₃R1 (PDB codes 6MU1 and 6MU2) indicates that the G2498S substitution may form H bonds with neighboring Arg-2544. B, monomeric WT mIP₃R2 and mutant mIP₃R2^GS cell lines, as well as dimeric R2/R2, R2/R2^GS, R2^GS/R2, and R2^GS/R2^GS cell lines generated in the IP₃R-null DT40-3KO cells were Western blotted. C, representative traces show Ca²⁺ signals of IP₃R-null DT40-3KO cells (blue), WT mIP₃R2 (green), and mIP₃R2^GS (red), R2/R2 (dark green), R2/R2^GS (orange), R2^GS/R2 (purple), and R2^GS/R2^GS (dark red) in response to trypsin (500 nm) when loaded with Fura-2/AM. D, scatter plots summarizing change in amplitude (peak ratio – basal ratio: average of initial 5 ratio points) for experiments similar to those shown in C. Boxes represent the 25th, 50th, and 75th percentiles, whereas whiskers represent the 5th and 95th percentiles and mean is represented by colored circles. E, stacked bar graph summarizing the percentage of amplitudes from D, which fall into pre-determined ranges such that only those cells with an amplitude change greater than 0.1 ratio units (black portion of bars) are considered to be responding to the trypsin stimulus shown in C. F, traces show Ca²⁺ signals of β-escin permeabilized WT mIP₃R2 (green), and mIP₃R2^GS (red), R2/R2 (dark green), R2/R2^GS (orange), R2^GS/R2 (purple), and R2^GS/R2^GS (dark red) cell lines in response to IP₃ (30 μm) when loaded with Mag-Fura-2/AM. Data are mean ± S.E. of three (n = 3) independent experiments. Data for DT40-3KO, mIP₃R2, and mIP₃R2^GS in D–-F came from Fig. 8. *, p < 0.05 and ***, p < 0.001 when compared with corresponding WT cell line (WT mIP₃R2 for monomers and R2/R2 cell line for dimers) and ###, p < 0.001 when compared with DT40-3KO cell line; one-way ANOVA with Tukey's test was performed in D (F_9,1568 = 437.7, p < 0.0001) and E (F_9,27 = 61.27, p < 0.0001). Unless otherwise stated, all data above comes from at least n = 3 experiments. ns, non-significant.

Discussion

Mutations have been identified throughout human IP₃R genes and are associated with numerous diseases (58, 59). How these mutations affect channel function and contribute to the pathogenesis of the disease remain, in most cases, largely unexplored. Here we showed that three different mutations in three distinct domains of the IP₃R result in decreased channel function. The extent and penetrance of disease varies with these mutations, and our data suggest that this depends on the mechanism of channel disruption and upon the pattern of inheritance. Both factors combine to alter the stoichiometry of functional subunits in the IP₃R tetramer and ultimately the extent of Ca²⁺ release supported by the assembled channel.

Expression of each mutation in DT40-3KO cells resulted in nonfunctional IP₃R Ca²⁺ channels, consistent with previous studies (22, 87). In each case, the complete loss in function was not due to a difference in protein localization and expression or ER store Ca²⁺ content. These data indicate that it is unlikely that these mutations result in mislocalization, altered degradation, or a constitutively leaky channel pore. Instead, the altered functionality occurred by mutation specific mechanisms. Patients harboring a mutation of Arg-269 to both a larger tryptophan and smaller glycine present with early-onset, nonprogressive SCA. Our data show that the R269W mutation resulted in a severe disruption of IP₃ binding. Mechanistically this is consistent with the substitution of the positively charged arginine residue with the neutral tryptophan residue impeding the coordination of negatively charged phosphate moieties of IP₃ in the binding pocket. In addition, it is likely that the bulky rings of the tryptophan residue may also sterically hinder IP₃ binding to other critical residues in the binding pocket core. Both factors likely severely compromise the conformational change in the binding pocket necessary for communication of the gating information required for channel opening.

The N602D mutant, in contrast to the R269W mutant, did not alter IP₃ binding. These results correlate with data showing that this region of the receptor is outside of the IP₃-binding core (22). This mutation, together with the T594I mutation is located between two α-helices in the ARM1 domain, adjacent to the LBD. Although it is possible that these mutations result in the alteration of IP₃R activity by channel modulators, it is more likely that these mutations result in an alteration of the native IP₃R conformation, which abrogates the communication of information between the binding core and the adjacent ARM1 domain. Thus, these mutations are likely best termed “gating mutants.”

Binding of IP₃ was also not altered by mutation of G2498S. This residue is integral to the selectivity filter in the conduction pathway of the channel pore. It is likely that the mutation from a small glycine residue to a larger serine residue results in a “dead pore.” Consistent with this observation, mutation of a neighboring selectivity filter residue in IP₃R1, Gly-2506, to both arginine (Fig. S6), and alanine also produces a similarly nonfunctional, pore mutant (26).

In contrast to expression of the mutant constructs in DT40-3KO, stable expression the R269W and N602D mutations in HEK-3KO cells resulted in channels with varying degrees of limited functionality in terms of the ability of Gα_q-coupled agonists to elicit Ca²⁺ release in a subset of cells. Increased channel activity of these mutants observed in the HEK-3KO cells could be due in part to increased IP₃R expression in HEK-3KO cells compared with DT40-3KO cells (Fig. S3). Conceivably, this higher expression level could facilitate receptor clustering, which would be expected to amplify small, unitary Ca²⁺ signals. Additionally, it is possible that HEK cells express so-called “licensing” factors, which are reported to render refractory IP₃R, activatable, thus revealing a degree of functionality to mutant IP₃R modulated in this fashion (63). Other factors could also contribute to increased functionality in HEK-3KOs including differences in accessory proteins, regulation, and the intracellular environment present in avian versus mammalian cells. In contrast to R269W and N602D rIP₃R1, stable expression of G2498S mIP₃R2 or G2506R hIP₃R1 in HEK-3KO cells failed to yield functional channels, reflecting the severe disruption of the channel pore.

Experiments documenting the activity in both DT40-3KO and HEK-3KO of the mutants stably expressed as homotetramers provides some insight into the mechanisms underlying the altered activity. However, both the R269W and N602D are inherited in an autosomal dominant manner, as are the vast majority of documented diseases associated with IP₃R mutations (Tables S2–S4). Given that IP₃Rs are well documented to form heterotetramers and this is the dominant IP₃R form when multiple species are expressed (64), it would be predicted that these patients express a variety of tetramers comprised of different ratios of WT IP₃R1 and mutant monomers. Indeed, our data confirms that heterotetramers of WT and WT channels are readily assembled. How incorporation of mutant monomers with altered activity influences the function of the assembled channel is therefore an important question. We address this issue using concatenated IP₃R constructs as a powerful tool to precisely control the stoichiometry of WT versus mutant subunits in the tetramer.

For example, heterotetramers consisting of two WT subunits and two R269W subunits are predominantly nonfunctional. Our previous studies have indicated that because of the requirement of each monomer to be liganded by IP₃, binding deficient mutants effectively act as dominant-negative modulators if incorporated into the heterotetramer (57). These results are also consistent with previous work from our laboratory showing other IP₃R1 mutations inherited in an autosomal dominant manner that result in Gillespie syndrome are dominant-negative (65). Moreover, these results suggest that in patients with heterozygous expression of these mutant proteins, only WT homotetramers will be able to contribute a significant amount of Ca²⁺ release. An extension of this idea, if it is assumed that tetramers form without bias, is that given a normal distribution of mutant and WT subunits in the tetramer, only a small percentage (∼9%) would be predicted to be homotetramers of WT subunits, likely severely attenuating Ca²⁺ release of the compliment of IP₃R1 expressed.

Tetramers assembled from dimers consisting of a WT and a N602D mutant, thus expressing two mutant and two WT sub-units, respectively, were similarly largely nonfunctional. Notably, however, expression of a concatenated tetramer with three WT subunits and one mutated N602D subunit resulted in a significant increase in channel function compared with cells expressing equal numbers of mutant and WT subtypes, indicating that the N602D subunit does not function as a true “dominant-negative.” Although the channel activity of this mutant tetramer is still significantly decreased compared with tetramers with four WT IP₃R1 subunits (WT homotetramers), these data would predict that heterozygous individuals would express a greater percentage of IP₃R tetramers, which retain function when compared with individuals expressing the R269W mutation.

In contrast to the R269W and N602D mutants, anhidrosis caused by the G2498S mutation is one of the few documented IP₃R disorders that is inherited in an autosomal recessive manner (55). Although individuals homozygous for the mutation exhibit a profound heat intolerance through the inability to sweat, those heterozygous for the mutation are asymptomatic. Symptomatic individuals are therefore homozygous for the expression of the mutant proteins, whereas heterozygous individuals would be predicted to express a combination of mutant and WT subunits. Our data using concatenated dimers of mutant and WT subunits provides insight into why heterozygous individuals are asymptomatic. We demonstrate that heterotetramers comprised of two WT IP₃R2 subunits and two G2498S mutant subunits retain a significant amount of function, such that Gα_q-coupled stimulation results in Ca²⁺ release activity intermediate between that of the WT homotetramers and the G2498S mutant homotetramers. Thus, these data indicate that in individuals heterozygous for the mutation, potentially a larger fraction of IP₃R2 channels are functional and thus presumably provide a sufficient Ca²⁺ signal to sustain sweat secretion. Notably, numerous cell types in the periphery, including hepatocytes and exocrine cells, express IP₃R2, usually in combination with other isoforms. These data likely also explain the highly specific phenotype observed with this mutation; whereas human sweat glands almost exclusively express IP₃R2, there appeared to be no obvious consequences of expression of the mutant in other tissues likely because the G2498S mutation does not act in a dominant-negative fashion. Thus, heterotetramers of mutant and other WT IP₃R subtypes would not be predicted to be substantially disabled and thus impact cellular function. In total, studies presented here using concatenated subunits provide unprecedented insight into how mutant IP₃R may function in heterozygous individuals and help demystify these challenging channelopathies.

Experimental procedures

Reagents

Antibodies generated by Pocono Rabbit Farms and Laboratories included rabbit polyclonal antibodies for IP₃R2 against amino acids 320-338 (NT2) and amino acids 2686-2701 in mouse IP₃R2 (CT2), as well as a rabbit polyclonal antibody against the C-terminal 19 aa of IP₃R1 (CT1) (66). Reagents used for SDS-PAGE were obtained from Bio-Rad. Protein A/G-agarose beads were from Santa Cruz. Alexa Fluor 488 and Dylight™ 800CW secondary antibodies were from Thermo Scientific. Restriction enzymes and DNA T4 ligase were purchased from New England Biolabs. β-Mercaptoethanol, chicken serum, penicillin/streptomycin, RPMI 1640 media, and G418 sulfate were obtained from Life Technologies. Fetal bovine serum and goat serum were from Gibco. Fura2-AM was from TEFLABS. All other chemicals were obtained from Sigma unless otherwise indicated.

Plasmid construction

Mutagenesis and all DNA modifications were carried out using Pfu Ultra II Hotstart 2× Master Mix (Agilent). Primers used in this study were synthesized by Integrated DNA Technologies (IDT) (Table S5). A QuikChange mutagenesis protocol was used to introduce R269W in cDNAs encoding the rat IP₃R1 (rIP₃R1; NM_001270596) in pDNA3.1 using mutagenic primers (primers 1 and 2). These primers encode R269W, in addition to silently introducing the indicated restriction site (Table S5). Similarly, two mutagenesis primers were used to introduce N602D into rIP₃R1 (primers 5 and 6), whereas mutagenesis primers were used to introduce R269W (primers 3 and 4), T494I (primers 7 and 8), and G2506R (primers 9 and 10 for G > A and primers 11 and 12 for G > C) into human IP₃R1 (hIP₃R1; NM_001099952) in pCDNA3.1 (obtained from Annetta Wronska at Columbia University). Finally, primers 13 and 14 introduced G2498S in mouse IP₃R2 (mIP₃R2; NM_019923) in pcDNA3.1. The coding regions for all constructs were confirmed by sequencing.

IP₃R1 and IP₃R2 concatenated constructs were generated as previously described in detail (38, 64, 67). Briefly, the cDNAs of the corresponding IP₃R subunits, designated as head and tail, in pcDNA3.1 vector was modified and subcloned in a tandem fashion, between the two arms of the pJAZZ mamm linear vector (Lucigen). The resultant construct encodes one ORF consisting of two IP₃R subunits connected with a 14-amino acid linker. Similarly, tetrameric IP₃R1 constructs were generated using four modified IP₃R1 subunits (numbered I-IV). The resulting cDNA forms one reading frame encoding four IP₃R1 subunits with 14 amino acid linkers separating the subunits. To create concatenated constructs harboring one or more mutant subunits, the corresponding mutations were introduced in the individual subunits before assembling the concatenated constructs.

Cell culture and transfection

DT40-3KO, chicken B lymphocyte cells engineered through homologous recombination for the deletion of the three endogenous IP₃R isoforms (56), were grown at 39 °C with 5% CO₂ in RPMI 1640 media supplemented with 1% chicken serum, 10% fetal bovine serum, 100 units/ml of penicillin, 100 µg/ml of streptomycin. DT40-3KO transfection was performed as described before (66). In brief, 5 million cells were washed once with PBS and electroporated with 5-10 µg of DNA using Amaxa cell nucleofector kit T (Lonza laboratories). Cells were allowed to recover for 24 h at 39˚C with 5% CO₂, and then passed into 96-well–plates containing media supplemented with 2 mg/ml of G418. Clones expressing the desired constructs were identified by immunoblots 10-14 days after transfection. HEK-3KO, HEK293 cells engineered in our laboratory though CRISPR/Cas9 for the deletion of the three endogenous IP₃R isoforms were grown at 37 °C with 5% CO₂ in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 units/ml of penicillin, 100 µg/ml of streptomycin.

HEK-3KO transfection was also performed similarly to as previously described (57). In brief, 1 million cells were washed once with PBS and electroporated with 5-10 µg of DNA using Amaxa cell nucleofector kit T (Lonza laboratories). Cells recovered for 48 h at 37 °C with 5% CO₂, and then were passed into new 10-cm plates containing media supplemented with 2 mg/ml of G418. After 7 days of selection, immunoblot was performed to assess expression of the desired constructs in these stable HEK cell pools. Following confirmation of expression, cells were either picked and transferred to new 24-well– plates or diluted at various concentrations into 96-well–plates or containing media supplemented with 2 mg/ml of G418. Monoclonal lines were identified 7 days after dilution and clones expressing the desired constructs were identified by immunoblots 10-14 days after identification.

Cell lysis, co-immunoprecipitation, and SDS-PAGE analyses

Following treatment, cells were harvested by centrifugation, washed once with 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 Na₄P₂O₇, 2 mm Na₃ VO₄, 1% Triton X-100 (v/v), 0.5% sodium deoxycholate (w/v), and 10% glycerol supplemented with protease inhibitors. Lysates were incubated for 30 min on ice and cleared by centrifugation at 16,000 × g for 10 min at 4 °C. Protein concentrations in cleared lysates were determined using Dc protein assay kit (Bio-Rad). Cleared supernatants were rocked with the indicated antibodies and 30 μl of protein A/G slurry overnight. Immunocomplexes were washed five times and resuspended in gel loading buffer. Proteins were fractionated on SDS-PAGE gel and transferred to a nitrocellulose membrane, which was probed with the indicated primary antibodies and corresponding secondary antibodies. Membranes were imaged with an Odyssey IR imaging system (LICOR Biosciences).

Immunocytochemistry and confocal microscopy

Stable HEK cell lines were plated on poly-d-lysine–coated coverslips. When roughly 50% confluent cells were fixed using ice-cold methanol at −20 °C or paraformaldehyde at room temperature for 10 min. Subsequently, coverslips were washed with PBS and cells were blocked in goat serum or BSA for 1.5 h. Following blocking, cells were incubated in primary antibody (IP₃R1 or IP₃R2) and DAPI overnight at 4 °C. The following day, primary antibody was removed, and coverslips were washed 3 times with PBS for 10 min with gentle rocking. Subsequently, the secondary antibody conjugated to Alexa Fluor 488 or 594 was incubated for 1 h at room temperature with gentle rocking. After incubation, coverslips were washed with PBS and mounted on slides. After allowing slides to dry, coverslips were sealed onto slides and imaged using confocal microscopy.

Measurement of cytosolic Ca²⁺ in intact cells

Single cell Ca²⁺ imaging was performed as described previously (66). DT40-3KO and HEK-3KO cells expressing IP₃R constructs were washed once with imaging buffer (10 mm HEPES, 1.26 mm Ca²⁺, 137 mm NaCl, 4.7 mm KCl, 5.5 mm glucose, 1 mm Na₂HPO₄, 0.56 mm MgCl₂, pH 7.4). Cells were then resuspended in imaging buffer containing 2 μm Fura2-AM and placed on a coverslip for 20 min at room temperature to allow for fluorescent dye loading and attachment of cells to the glass coverslip. Cells were then perfused with imaging buffer and stimulated with the desired agonist. Ca²⁺ imaging was performed using an inverted epifluorescence Nikon microscope with a ×40 oil immersion objective. 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 × 4 binning using a digital camera driven by TILL Photonics software. Image acquisition was performed using TILLvisION software and data were exported to Microsoft excel. Experiments were repeated at least three times.

Population-based Ca²⁺ imaging was performed in intact HEK cells by loading cells with 4 μm Fura-2 AM in culture medium. After 1 h, cells were harvested and subsequently washed, resuspended in imaging buffer, and dispensed into a black-walled flat-bottom 96-well–plate (∼300,000 cells/well). The plate was spun at 200 × g for 2 min to plate cells to the bottom of each well. The plate rested for 30 min prior to commencing the assay. Fluorescence imaging was carried out using FlexStation 3 from Molecular Devices (excitation alternated between 340 and 380 nm and emission 510 nm) and analyzed by using SoftMax© Pro Microplate Data Acquisition and Analysis software to determine peak of the response. Data were averaged from at least 3 individual plates and curves fitting was done using a logistic dose-response equation in the OriginPro 6.1 software.

Single cell-permeabilized DT40 cell IP₃-induced Ca²⁺ release (IICR) assays

DT40-3KO cells stably expressing defined IP₃R constructs were loaded with 20 μm MgFura-2 AM for 50–60 min and subsequently permeabilized by superfusion of 40 μm β-escin in imaging buffer as previously described (57, 64). The duration of permeabilization depended on the flow rate, and care was taken to prevent excessive (internal membrane) permeabilization by careful monitoring of the fluorescence. Permeabilized cells were then washed in ICM without β-escin for 15 min to facilitate removal of cytosolic dye. The internal stores of permeabilized cells were loaded by activating SERCA through superfusion with ICM containing 1.4 mm MgCl₂, 3 mm NaATP, and 0.65 mm CaCl₂ (free [Ca²⁺] of 200 nm (MaxChelator freeware)). Upon stabilization of fluorescence, MgCl₂ was removed from the superfused solution to disable SERCA. Varying concentrated IP₃ were applied through superfusion to induce unidirectional Ca²⁺ release from internals stores. After washing out IP₃, stores were repeatedly refilled and released to allow for repeated stimulations. Each experiment contained between 30 and 60 cells in a field of view and performed a minimum of three times. The initial rates of Ca²⁺ release were determined by fitting the release curves of each individual cell to a single exponential function (OriginPro 6.1). Concentration-response curves were plotted using the determined rates. All imaging was performed on an inverted epifluorescence Nikon microscope with a ×40 oil immersion objective. Cells were alternately excited at 340 and 380 nm, and emission was monitored at 505 nm. Images were captured every 5 s during permeabilization, loading, and disabling and every second during release. This was done with an exposure of 10 ms and 4 × 4 binning using a digital camera driven by TILL Photonics software.

Competitive IP₃-binding assay

DT40-3KO cells stably expressing the indicated IP₃R constructs were washed once with PBS and lysed in Igepal lysis buffer containing (120 mm NaCl, 50 mm Tris-HCl, 0.5% Igepal (v/v), 1 mm EDTA) supplemented with a mixture of protease inhibitors. Lysates were incubated on ice for 30 min and cleared by centrifugation at 16,000 × g for 10 min at 4 °C. IP₃R proteins were immunoprecipitated overnight using anti-IP₃R and protein A/G PLUS-agarose beads. Immunocomplexes were washed three times with lysis buffer and twice with binding buffer (50 mm Tris-base, 1 mm EDTA, pH 8, plus 1 mm β-mercaptoethanol) and equally distributed into 1.5-ml microcentrifuge tubes. One pair of tubes was immunoblotted using SDS-PAGE to ensure equal immunoprecipitation of WT and mutant protein, whereas all others underwent the binding reaction consisting of a 100-μl volume containing equal amounts of immunoprecipitated proteins, 2.5 nm tritiated IP₃ ([³H]IP₃) in the presence of different concentrations of unlabeled IP₃. Tubes were incubated for 1 h at 4 °C with mixing every 10 min. Finally, tubes were centrifuged at 16,000 × g, supernatants were removed, and 500 µl of 1% of SDS was added to each tube. The contents of the tubes were transferred to vials containing scintillation liquid after 12 h and bound radioactivity were measured using liquid scintillation counter. Nonspecific binding was calculated as the amount of bound radioactivity in the presence of 50 μm unlabeled IP₃. Specific binding is determined by subtracting nonspecific binding from the CPM values obtained in other conditions. Total specific binding was determined as the binding observed in the absence of unlabeled IP₃. Normalized specific binding from 3 experiments were averaged and curves were fit using a logistic dose-response equation using the OriginPro 6.1 software.

Data analysis and statistical analysis

Data obtained were exported into Microsoft Excel where appropriate ratios, normalizations, and averages were calculated. Statistical analysis was performed in GraphPad Prism for experiments comparing two cell lines (unpaired t test) and three or more cell lines (one-way ANOVA with Tukey's test). Logistic fits, as well as IC₅₀ were calculated using OriginPro 6.1.

Data availability

All constructs used are available freely on request. All data needed to evaluate the conclusions are presented in the paper or Supporting material.

Edited by Roger J. Colbran