Agonist-Induced Ca²⁺ Signaling in HEK-293-Derived Cells Expressing a Single IP₃ Receptor Isoform

Abstract

In mammals, three genes encode IP₃ receptors (IP₃Rs), which are involved in agonist-induced Ca²⁺ signaling in cells of apparently all types. Using the CRISPR/Cas9 approach for disruption of two out of three IP₃R genes in HEK-293 cells, we generated three monoclonal cell lines, IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK, with the single functional isoform, IP₃R1, IP₃R2, and IP₃R3, respectively. All engineered cells responded to ACh with Ca²⁺ transients in an “all-or-nothing” manner, suggesting that each IP₃R isotype was capable of mediating CICR. The sensitivity of cells to ACh strongly correlated with the affinity of IP₃ binding to an IP₃R isoform they expressed. Based on a mathematical model of intracellular Ca²⁺ signals induced by thapsigargin, a SERCA inhibitor, we developed an approach for estimating relative Ca²⁺ permeability of Ca²⁺ store and showed that all three IP₃R isoforms contributed to Ca²⁺ leakage from ER. The relative Ca²⁺ permeabilities of Ca²⁺ stores in IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK cells were evaluated as 1:1.75:0.45. Using the genetically encoded sensor R-CEPIA1er for monitoring Ca²⁺ signals in ER, engineered cells were ranged by resting levels of stored Ca²⁺ as IP3R3-HEK ≥ IP3R1-HEK > IP3R2-HEK. The developed cell lines could be helpful for further assaying activity, regulation, and pharmacology of individual IP₃R isoforms.

1. Introduction

Extracellular cues regulate diverse cellular functions by involving a variety of surface receptors and intracellular signaling pathways. The mobilization of intracellular Ca²⁺ is central to transduction of many first messengers that act through G-protein-coupled receptors (GPCRs) and receptor tyrosine kinases (RTKs) coupled to the phosphoinositide cascade [1,2,3]. GPCRs stimulate the phosphoinositide cascade by involving the β1–4 isoforms of phospholipase C (PLC) in a αGq- and/or βγGi-dependent manner, while RTKs employ PLCγ [2,3,4,5,6]. Once stimulated, PLC produces two second messengers, IP₃ and diacylglycerol, by hydrolyzing the precursor lipid phosphatidylinositol 4,5-bisphosphate. The primary mode of action of IP₃ is to release Ca²⁺ from Ca²⁺ store through IP₃ receptors (IP₃Rs) [1,3]. Three genes encode IP₃R subunits IP₃R1, IP₃R2, and IP₃R3, which form primarily homotetrameric IP₃-gated Ca²⁺ channels in the endoplasmic reticulum (ER). Moreover, evidence exists that different IP₃R subunits also can form heterotetrameric complexes with specific functional and regulatory properties [3,7,8,9,10]. This and alternative splicing of the IP₃R genes increase the functional heterogeneity of IP₃-gated channels, posing an additional complexity to studying physiological functions and regulations of IP₃Rs in cells [11,12].

Most cell types express two or even all three IP₃R genes [12], implying that any particular IP₃R subtype cannot cover the reported diversity of intracellular IP₃/Ca²⁺ signaling [1]. The expression pattern of IP₃Rs is not uniform among different tissues and cell types, suggesting a specific role for individual IP₃R isoforms or their combinations in cell physiology [12]. For instance, IP₃R1 predominates in Purkinje neurons, cardiac myocytes rely largely on IP₃R2, and insulin-secreting β-cells, taste cells, and vomeronasal sensory neurons express primarily IP₃R3 [13,14,15,16,17].

Since agonist-induced IP₃/Ca²⁺ signaling is pivotal to the physiology of apparently all cell types, IP₃Rs were subjected to intensive studies at molecular, biophysical, and functional levels [1,3,9,10,18]. Reportedly, all IP₃R isoforms are regulated by cytosolic Ca²⁺ in a bimodal manner, suggesting that IP₃Rs hold two allosteric Ca²⁺-binding sites, both activatory and inhibitory [1,3,19,20]. Although the activatory site has been identified, the structure and location of the inhibitory site is still debatable [3]. Characteristic of IP₃R regulation is that IP₃ binding increases affinity of the activatory site, and its occupation by Ca²⁺ increases the open probability of the IP₃-gated channel [19]. Owing to this interdependent regulation of IP₃R gating by the primary co-agonists, Ca²⁺ ions released from the ER through IP₃Rs can facilitate their activity. This positive feedback underlies Ca²⁺-induced Ca²⁺ release (CICR), the regenerative process that is ubiquitously involved in the generation of diverse Ca²⁺ signals [1,6]. Serving as a co-agonist at a relatively low level, cytosolic Ca²⁺ suppresses the activity of IP₃Rs by occupying the inhibitory site at higher concentrations [1,3,19]. This multimodal control of IP₃Rs by IP₃ and Ca²⁺ is central to diverse modes of intracellular Ca²⁺ signaling [1,20].

Although IP₃R1, IP₃R2, and IP₃R3 share 60–80% homology at the amino acid level, they are dissimilar in sensitivity towards the primary agonists IP₃ and Ca²⁺ and differ in their regulatory mechanisms [9,10,11,12,18]. Multiple intracellular regulators, from small molecules to proteins, have been reported to control IP₃R activity in an isoform-specific manner and depending on intracellular context. The list of regulators includes ATP [21], cAMP [22], NADH [23], H⁺ [24], calmodulin, and several other Ca²⁺-binding proteins [18,25], as well as the IP₃R binding proteins IRBIT and IRAG [26,27]. A variety of protein kinases and phosphatases is also involved in the regulation of IP₃Rs [10,18,28].

Previously, we studied Ca²⁺ signaling induced by GPCR agonists, including ATP, UTP, adenosine, ACh, 5-HT, and glutamate, in cells of diverse types [29,30,31,32]. The assayed cells universally responded to Ca²⁺-mobilizing agonists in an “all-or-nothing” manner: They either were irresponsive to a particular agonist at subthreshold concentrations or generated quite similar Ca²⁺ signals at different agonist doses above the threshold (see Figure 1 below). The body of evidence suggests that being a trigger-like self-driven process, CICR finalized agonist transduction by forming cellular responses of a virtually universal shape, regardless of agonist doses [31]. Given that assayed cells express multiple IP₃R isoforms, it remains unclear whether the “all-or-nothing” responsiveness is mediated by a specific IP₃R subtype or whether each IP₃R isoform can endow agonist-induced Ca²⁺ signaling in cells with this feature.

As demonstrated in previous studies, a cell line expressing a particular IP₃R isotype represents a promising cellular model for the systematic assay of gating, regulation, pharmacology, and physiology of IP₃R isoforms [33,34,35,36,37,38,39]. Here, we generated several such cell lines by inactivating two out of three IP₃R genes in HEK-293 cells using the CRISPR/Cas9 approach. Different aspects of intracellular Ca²⁺ signaling in these cells were analyzed with Ca²⁺ imaging. It was particularly shown that the engineered cells responded to ACh in the “all-or-nothing” manner and that the ACh sensitivity of a given cellular subclone strongly correlated with EC₅₀ for IP₃ characteristic of an IP₃R isoform it expressed. A mathematical model of intracellular Ca²⁺ signals induced by thapsigargin, a SERCA inhibitor, was proposed, based on which we developed an approach for estimating relative Ca²⁺ permeability of Ca²⁺ store. Altogether, our results suggest that the engineered cell lines could provide a relatively simple and effective assay of activity, regulation, and pharmacology of individual IP₃R isoforms.

2. Materials and Methods

2.1. Cell Culture and Transfection

WT-HEK cells and their derivatives were routinely cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Corning, NY, USA) containing 10% (v/v) fetal bovine serum (Cytiva, Marlborough, MA, USA) and the antibiotic gentamicin (100 μg/mL) (Corning, NY, USA) on 12-well culture plates. Cells were grown in a humidified atmosphere containing 5% CO₂ at 37 °C.

To induce transient expression of genetically encoded Ca²⁺ sensor R-CEPIA1er with reticular location, cells were transfected with the plasmid vector pCMV R-CEPIA1er kindly provided by Masamitsu Iino (Addgene plasmid # 58216; http://n2t.net/addgene:58216, accessed on 1 October 2019; RRID:Addgene_58216) [40]. Before the day of transfection, 3–5 × 10⁵ cells were placed in the well of 12-well culture plates. For the transfection of cultured cells, the growth medium was replaced with the transfection mixture, containing 800 μL of the growth medium as well as 200 μL OptiMEM media, 2 μL P3000 Reagent, 2 μL Lipofectamine 3000 (all from Invitrogen, Waltham, MA, USA), and 2 μg pCMV R-CEPIA1er. After 24 h of incubation, the transfection mixture was replaced with the normal culture medium. The transfection was considered effective if at least 30% of the transfected cells exhibited sufficient fluorescence in the red spectral range. Next, the transfected cells were subjected to selection in the presence of antibiotic G418 (700 μg/mL) (Corning) for two weeks. The survivor cells yielded a population, which was maintained in the presence of 300 μg/mL G-418, wherein a nearly 70% cell fraction stably expressed R-CEPIA1er.

2.2. Ca²⁺ Imaging and IP₃ Uncaging

Isolated cells were plated on a photometric chamber (~150 μL), which contained a disposable coverslip (Menzel-Glaser, Braunschweig, Germany) with an attached ellipsoidal resin wall. The chamber bottom was coated with Cell-Tak (Corning, NY, USA), enabling strong cell adhesion. Attached cells were loaded with Fluo-8 at room temperature (23–25 °C) by adding Fluo-8 AM (3 μM) and Pluronic^® F-127 (0.02%) (both from AAT Bioquest, Pleasanton, CA, USA) to the following bath solution (mM): 140 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES-NaOH (pH 7.4), 10 glucose. After 30 min of incubation, the cells were rinsed several times with the bath solution and stored for 40 min to complete dye de-esterification. For IP₃ uncaging, the cells were treated with 3 μM Fluo-8 AM + 4 μM caged-Ins(145)P3/PM (SiChem, Bremen, Germany) + 0.02% Pluronic as described above. When necessary, 2 mM CaCl₂ in the bath were replaced with 0.5 mM EGTA + 0.4 mM CaCl₂, thus reducing free Ca²⁺ to nearly 260 nM at 23 °C, as calculated with the Maxchelator program (http://maxchelator.stanford.edu, accessed on 1 October 2019). All used chemicals were applied with a gravity-driven perfusion system, which enabled the complete replacement of the bath solution in the photometric chamber for nearly 2 s. The used salts and buffers were from Sigma-Aldrich (St. Louis, MO, USA); the ACh and inhibitors were from Tocris Bioscience (Bristol, UK).

Experiments were carried out using an inverted fluorescent microscope Axiovert 135 equipped with an objective Plan NeoFluar 20×/0.75 (Zeiss, Oberkochen, Germany) and a digital ECCD camera LucaR (Andor Technology, Belfast, Northern Ireland). Apart from a transparent light illuminator, the microscope was equipped with a handmade system for epi-illumination via an objective. The epi-illumination was performed using a bifurcational glass fiber. One channel transmitted irradiation of computer-controlled LEDs, which provided sequential excitation of Fluo-8 or R-SEPIA1er at 480 ± 5 and 572 ± 17 nm, respectively. Their emission was collected at 535 ± 20 and 630 ± 30 nm, respectively. Serial fluorescent images were captured every second and analyzed using NIS-Elements software (version AR 5.30.01) (Nikon, Tokyo, Japan). Deviations of cytosolic and reticular Ca²⁺ in individual cells were quantified by a relative change in intensity of Fluo-8 and R-SEPIA1er fluorescence (ΔF/F₀), respectively. Another channel was connected to a TECH-351 Advanced pulsed solid laser (680 mW) (Laser-Export, Moscow, Russia). This unit operated in a two-harmonic mode and generated not only 351 nm UV light used for Ca²⁺ uncaging but also visible light at 527 nm. The last could penetrate an emission channel through non-ideal optical filters and elicited optical artifacts during uncaging.

2.3. Generation of Cell Lines with Inactivated IP₃R Genes

Two out of three IP₃R genes in HEK-293 cells were inactivated sequentially using CRISPR/Cas9 technology. Firstly, three monoclonal cell lines were generated, HEK-∆IP3R1, HEK-∆IP3R2, and HEK-∆IP3R3, wherein IP₃R1-, IP₃R2-, and IP₃R3 were disrupted, respectively. Next, by inactivating either of two remaining IP₃R genes in these lines, IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK cells were obtained with solely IP₃R1, IP₃R2, or IP₃R3, respectively, to be functional. The used methods, protocols, and controls are sufficiently detailed in the Supplementary Materials.

2.3.1. Inactivation of IP₃R1 Gene

The construct used for IP₃R1 inactivation was engineered on the basis of the AIO-GFP vector that provided expression of Cas9-D10A nickase fused with the enhanced green fluorescent protein (EGFP) [41]. This vector was kindly provided by Steve Jackson (Addgene plasmid # 74119; http://n2t.net/addgene:74119, accessed on 30 March, 2017; RRID: Addgene_74119). The appropriate protospacer locus was identified using IP₃R1 mRNA sequences (GenBank NM_001099952.4; NM_001168272.2; NM_001378452.1; NM_002222.7). Target-specific sgRNAs were designed and cloned into the AIO-GFP vector as described in the Supplementary Materials. The final cAIO-GFP-sgRNA construct was verified by sequencing. WT-HEK cells were transfected with cAIO-GFP-sgRNA using Lipofectamin 3000 (Invitrogen, Waltham, MA, USA). Seventy-two hours after transfection, EGFP-expressing cells were sorted using a FACSAria SORP sorter (BD Biosciences, NJ, USA) and grown as single cells. Once a particular cell monoclone achieved a monolayer, the cells were collected, their genomic DNA was isolated, and a gene fragment containing the target site was amplified using PCR with specific primers. Each amplicon was subjected to in vitro hydrolysis using commercial nuclease Cas9 and synthesized sgRNA to reveal induced mutations in the IP₃R1 gene in a source clone. Finally, 4 cell clones (HEK-∆IP3R1) were found to contain necessary biallelic mutations inactivating the IP₃R1 gene, which also were verified by sequencing. Due to exhibiting the highest fraction (~90%) of cells responsive to ACh with Ca²⁺ transients, one clone was chosen for future experimentations.

2.3.2. Inactivation of IP₃R2 and IP₃R3 Genes

The IP₃R2 and IP₃R3 genes were inactivated using the pGuide-it-tdTomato vector (Takara Bio, San Jose, CA, USA) that encoded nuclease Cas9 fused with the red fluorescent protein tdTomato. The appropriate protospacer locus was identified using the IP₃R2/IP₃R3 mRNA sequence (GenBank NM_002223.4/NM_002224.4). Target-specific sgRNA-IP₃R2/IP₃R3 was designed and cloned into the pGuide-it-tdTomato vector as described in the Supplementary Materials. After verification by sequencing, the final pGuide-it-tdTomato-sgRNA-IP₃R2/IP₃R3 construct was transfected into WT-HEK cells. In seventy-two hours, tdTomato-positive cells were collected using a FACSAria SORP sorter and then grown as single cells. Genomic DNA was isolated from a particular monoclone, and a gene fragment containing a target site was amplified using PCR with specific primers. Each amplicon was subjected to in vitro hydrolysis using commercial nuclease Cas9 and synthesized sgRNA to reveal mutations in the IP₃R2/IP₃R3 gene in cells of source clones. Overall, 3 HEK-∆IP3R2 and 4 HEK-∆IP3R3 clones were identified to contain inactivating indels in both alleles of the IP₃R2 and IP₃R3 genes, which were verified by sequencing. All these HEK-∆IP3R2/HEK-∆IP3R3 clones were assayed with Ca²⁺ imaging, and one exhibiting the highest fraction (~90%) of ACh-responsive cells was taken for future experimentations.

2.3.3. Generation of IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK Lines

Cells expressing the only IP₃R1 gene were generated by disrupting the IP₃R2 gene in HEK-∆IP3R3 cells using the strategy described above. In a variety of HEK-∆IP3R3-derived subclones, solely one cell clone, IP3R1-HEK, was eventually identified to contain proper biallelic inactivating mutations in both IP₃R2 and IP₃R3 genes.

Cells, wherein solely IP₃R2 or IP₃R3 was functional, were generated by inactivating the IP₃R3 or IP₃R2 gene in HEK-∆IP3R1 cells, respectively. Single-cell clones of each type, IP3R2-HEK and IP3R3-HEK, were eventually obtained (see Supplementary Materials).

2.4. RT-PCR and RT-qPCR

For the expression analysis, total RNA was routinely isolated from a particular cell colony (~10⁶ cells), using the Gen Elute Mammalian Total RNA Miniprep Kit (Sigma, St. Louis, MO, USA) according to the manufacturer’s protocol. Reverse transcription was performed using SuperScript IV VILO Master Mix (Thermo Fisher Scientific, Waltham, MA, USA). PCR amplification of the target sequences was performed using Phusion Hot Start II High-Fidelity DNA Polymerase (Thermo Fisher Scientific, Waltham, MA, USA) and gene-specific primers (Table S2). The primers were intron-spanning and designed to recognize transcript sequences of all known splice variants, should they be characteristic of an assayed gene. Different RNA transcription levels were quantified with the RT-qPCR approach using a real-time PCR instrument DTlight (DNA Technology, Protvino, Russia) and Maxima SYBR Green/ROX qPCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA). Amplifications were performed starting with a 3 min template denaturation step at 94 °C, followed by 45 cycles of denaturation at 94 °C for 20 s and combined primer annealing/extension at the gene specific primer temperature for 30 s (Table S2). All samples were amplified in triplicate and the mean was obtained for further calculations. Relative quantification of gene expression was calculated using the 2^–ΔCt method and β-Actin gene as an endogenous reference.

2.5. Western Blot

Expression of a particular IP₃R isoform in the WT-, IP3R1-, IP3R2-, and IP3R3-HEK cells was verified with the Western blot approach (Figure S11). In each case, cells (~10⁷) were pelleted and directly transferred to 300 µL of 1x Laemmli buffer. Samples were incubated for 10 min at 95 °C, and each probe (20 µL) was applied on 4–15% BIS-TRIS gradient gel. Protein transfer was performed on 0.45 µM nitrocellulose membranes (BioRad Laboratories, Hercules, CA, USA) using the PowerBlotter semi-dry transfer system (BioRad Laboratories, Hercules, CA, USA). Membranes were probed with primary antibodies to IP₃R1 (rabbit polyclonal antibodies against rat IP₃R1 2732–2750 aa, Alomone Labs, Jerusalem Israel, Cat# ACC-019), type 2 (rabbit polyclonal antibodies against rat IP₃R2 (2683–2696 aa, Alomone Labs, Jerusalem, Israel, Cat# ACC-116), or human IP₃R3 (mouse monoclonal antibodies against 22–230 aa, BD Transduction Laboratories, NJ, USA, Cat# 610312). Next, probes were incubated with horseradish peroxidase-conjugated secondary antibodies Anti-Rabbit gG Peroxidase Conjugate (Sigma-Aldrich, St. Louis, MO, USA, Cat# A-9169) and Anti-Mouse IgG, IgA, and IgM Peroxidase Conjugate (IMTEC, Moscow, Russia, Cat# P-GAM Iss). Anti-actin was from United States Biological (Cat# A0760-40). Blots were imaged using iBright™ CL750 Imaging System (Thermo Fisher Scientific, Waltham, MA, USA) and enhanced chemiluminescent substrate SuperSignal™ West Pico PLUS Chemiluminescent Substrate (Thermo Fisher Scientific, Waltham, MA, Cat#34580).

3. Results

Reportedly, cells of the HEK-293 line endogenously express GPCRs of multiple types, including muscarinic, purinergic, chemokine, lysophospholipid, and protease receptors, many of which are coupled to the phosphoinositide cascade [42]. We therefore assayed the responsiveness of basic HEK-293 cells (WT-HEK) and their genetically modified offspring to a variety of GPCR agonists with Ca²⁺ imaging. Among agonists capable of mobilizing intracellular Ca²⁺, acetylcholine (ACh) was most effective in that it mobilized cytosolic Ca²⁺ in most (80–90%) of the assayed cells. So, cell responsiveness to ACh was predominantly analyzed in the experiments described below. WT-HEK cells that expressed all three IP₃R isotypes (Figure S6) were subjected to gene editing using the CRISPR/Cas9 technology, and three monoclonal cell lines, each expressing a particular IP₃R isoform, were generated (see Supplementary Materials). Here, we performed the comparative physiological analysis of Ca²⁺ signaling in WT-HEK cells as well as in cells of three lines, called IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK, which functionally express solely IP₃R1, IP₃R2, and IP₃R3, respectively.

3.1. ACh Responses of WT-HEK Cells

In a particular experiment, 100–120 cells loaded with Fluo-8 were assayed simultaneously using Ca²⁺ imaging. When applied shortly, ACh elicited Ca²⁺ transients in WT-HEK (Figure 1A) and in cells of the derived lines (Figure 2A). In general, both Ca²⁺ release from intracellular Ca²⁺ store and Ca²⁺ entry mediated by a variety of Ca²⁺-permeable channels contributed to agonist-induced mobilization of cytosolic Ca²⁺. Although to our knowledge, HEK-293 cells do not express nicotinic ACh receptors, receptor-operated channels and store-operated channels (SOCs), which serve in apparently all cell types [6], contributed to the ACh responses.

It turned out that the decrease in bath Ca²⁺ from 2 mM to 260 nM affected cellular responses to ACh insignificantly (Figure 1A,B), thus indicating that IP₃-driven Ca²⁺ release was mostly responsible for ACh-induced Ca²⁺ signals. The rationale for the abovementioned Ca²⁺ protocol was that the complete removal of bath Ca²⁺ with EGTA initiated dramatic rundown of cell responsivity during prolonged recordings. On the other hand, cells tolerated the reduction in extracellular Ca²⁺ to 260 nM, which proportionally, i.e., by the four orders of magnitude, decreased Ca²⁺ influx, in fact nullifying its contribution. In addition, the inhibition of PLC with U73122 (3 μM) rendered cells irresponsive to ACh (Figure 1A). These findings indicate that ACh transduction involved primarily muscarinic receptors that were coupled by the phosphoinositide cascade to IP₃-driveen Ca²⁺ release rather than to Ca²⁺ entry.

The previous transcriptome analysis suggests that the M3-receptor was the predominant muscarinic isoform expressed in WT-HEK [42]. We confirmed this finding and also evaluated the M3 transcripts to be much more abundant compared to the other M-receptors (M1, M4, and M5) expressed in WT-HEK cells (Figure 1C and Figure S7). Consistently, ACh responses completely disappeared in the presence of 10 nM 4-DAMP (Figure 1C), an antagonist specific to the human M3 and M5 receptors [43].

The peculiar feature of ACh responses was their dose dependence. At concentrations below the threshold of 150–200 nM, ACh insignificantly affected intracellular Ca²⁺ in WT-HEK cells but the agonist elicited Ca²⁺ transients of similar magnitudes at a variety of higher doses (Figure 1E). This “all-or-nothing” fashion was also characteristic of agonist-induced Ca²⁺ responses in cells of several other types [29,30,31,32]. Such a step-like dose dependence should be intrinsic to agonist transduction that involves CICR, the mechanism capable of producing a large and global Ca²⁺ signal of universal shape, regardless of agonist concentrations [31]. The involvement of IP₃R-mediated CICR in the generation of ACh responses (Figure 1E) was verified by the observation that 0.5 s and 1.5 s UV pulses elicited ACh response-like Ca²⁺ transients that were similar kinetically and by magnitude (Figure 1F,G), although the former should have uncaged a nearly three times smaller amount of IP₃.

3.2. Responses of IP3R1-, IP3R2-, and IP3R3-HEK Cells to ACh

In designated experiments, we assayed IP3R1-HEK-, IP3R2-HEK-, and IP3R3-HEK cells and found that irrespective of the line, most (~90%) of them responded to brief ACh pulses with Ca²⁺ transients (Figure 2A). Similar to WT-HEK cells (Figure 1E), cells of each particular line responded to the agonist in an “all-or-nothing” manner (Figure 2A). Being associated with CICR (Figure 1F), the “all-or-nothing” responsivity of IP3R1-HEK-, IP3R2-HEK-, and IP3R3-HEK cells was consistent with the idea that each IP₃R isoform was capable of mediating CICR [44,45]. Given the nearly step-like responsiveness of individual cells (Figure 1E and Figure 2A), their sensitivity to ACh could not be characterized by Ca²⁺ response magnitude. As an alternative dose dependence, each assayed population was evaluated by a fraction of cells responsive to the agonist at a particular concentration. After being averaged over all the experiments (n = 7–11) (Figure 2B, symbols), the data were fitted using the nonlinear regression approach and the Hill equation (Figure 2B, continuous lines):

$$F\left(C\right)=F_0\frac{C^n}{C_{0.5}^n+C^n}$$

(1)

where F(C) is the fraction of responsive cells at the given ACh concentration C, F₀ is the maximal fraction of ACh responsive cells, C_0.5 is the EC₅₀ dose, and n is the Hill coefficient. Based on this approximation, EC₅₀ for WT-, IP3R1-, IP3R2-, and IP3R3-HEK cells were estimated as 0.16 ± 0.006, 0.21 ± 0.007, 0.41 ± 0.013, and 1.01 ± 0.05 μM, respectively. Thus, by sensitivity to ACh, the assayed cell lines were ranked as WT-HEK ≈ IP3R2-HEK > IP3R1-HEK > IP3R3-HEK.

Consistent with our previous studies of agonist-induced Ca²⁺ signaling [30,31], cells of all assayed lines responded to ACh with evident delay relative to the moment of agonist application, depending on the dose of the agonist (Figure 2C). We determined the characteristic time of the response lag (τ_d) as a time interval necessary for a Ca²⁺ transient to reach the half magnitude (Figure 2C). The common feature of response lags was that τ_d markedly decreased with increasing agonist concentration (Figure 2C,D) in a cell-line-specific manner (Figure 2D). Being similarly sensitive to ACh (Figure 2B), WT-HEK and IP3R2-HEK cells showed quite similar dependencies of response lag on ACh concentration (Figure 2B). In the case of IP3R1-HEK and IP3R3-HEK cells, response lag versus ACh dose was shifted to the right (Figure 2D), consistent with lower sensitivities of both lines to ACh (Figure 2B).

Note that recently, we developed a mathematical model of agonist-induced Ca²⁺ signaling, which properly simulated cell responsivity of the “all-or-nothing” type [31]. This model predicted that for the phosphoinositide cascade with IP₃R as the only type, its affinity to IP₃ should be a key factor that determines the lag of Ca²⁺ responses. Note that based on their affinity to IP₃, different IP₃R isoforms follow the sequence IP₃R2 > IP₃R1 > IP₃R3 [10,45,46,47]. It is therefore likely that only sensitivity of the particular IP₃R isoform to IP₃ determines the dose–response curve characteristic of the cell line wherein it is expressed (Figure 2B).

3.3. Thapsigargin Test of WT-, IP3R1-, IP3R2-, and IP3R3-HEK Cells

Several Ca²⁺-transporting systems are universally involved in Ca²⁺ homeostasis in resting and stimulated cells [46,47]. To interpret the experiments described below, we employed a simplified model of Ca²⁺ homeostasis in assayed cells (Figure 3A). It was suggested that a level of cytosolic Ca²⁺ was mainly determined by Ca²⁺ fluxes through the plasmalemma and ER membrane, although Ca²⁺-accumulating organelles, such as mitochondria, also could shape Ca²⁺ signals. As suggested in Figure 3A, IP₃Rs, Ca²⁺ leak channels, and Ca²⁺-ATPase SERCA were pivotal players in ER, while SOCs, the Na⁺/Ca²⁺ exchanger, and Ca²⁺-ATPase PMCA mediated Ca²⁺ fluxes through the plasma membrane.

The inhibition of SERCA with thapsigargin is conventionally used to empty Ca²⁺ store through spontaneous Ca²⁺ release, thus initiating store-operated Ca²⁺ entry (SOCE) in unstimulated cells. We employed the classical thapsigargin test to clarify whether resting activity of IP₃Rs was a factor of Ca²⁺ leakage from ER in assayed cells. In a typical experiment, cells were initially stimulated with 1 μM ACh, and Ca²⁺ homeostasis in a given cell was considered sufficiently robust if its Ca²⁺ response to the agonist was fast and exceeded 2 in terms of ΔF/F₀ (Figure 3B–E, upper traces). Next, cells were treated with 1 μM thapsigargin with 260 nM Ca²⁺ in the bath that nullified a contribution of Ca²⁺ entry to intracellular Ca²⁺ signals. Under these conditions, thapsigargin-elicited Ca²⁺ transients were produced by Ca²⁺ leakage from ER, which emptied Ca²⁺ store and stimulated activity of SOCs, albeit SOCE was not evident at low bath Ca²⁺. Thapsigargin was applied for 600 s, which was a sufficient interval for intracellular Ca²⁺ to return apparently to the initial level. The restoration of bath Ca²⁺ to 2 mM initiated significant SOCE associated with a marked Ca²⁺ response in the cell cytosol (Figure 3B–E; upper fluorescence trace). To quantify Ca²⁺ release and Ca²⁺ entry, Ca²⁺ traces from individual cells (Figure 3B–E; upper traces) were differentiated, and maximal rates of Ca²⁺ release (R_r) and Ca²⁺ entry (R_e) were determined as appropriate local maximums in the d(F/F₀)/dt curves (Figure 3B–E; upward peaks in the bottom traces).

Based on these data, we generated a number of histograms to characterize distributions of R_r and R_e among robust cells of different lines. Note that satisfactory fitting for all obtained histograms could not be achieved using a single Gaussian function. Instead, it normally required a combination of two or three Gaussians (Figure 4). This implied that each particular cellular line could include two to three cell subpopulations that differed in Ca²⁺ leakage and SOCE. The level of luminal Ca²⁺, activities of Ca²⁺ leak channels and Ca²⁺ pumps, and coupling of Ca²⁺store to SOCs may have varied from cell to cell.

The experimental histograms revealed dissimilarity between WT-, IP3R1-, IP3R2-, and IP3R3-HEK cells in both thapsigargin-induced Ca²⁺ release and SOCE. In particular, the R_r distributions for WT-HEK and IP3R3-HEK cells were wider and shifted positively compared to the R_r histograms obtained for IP3R1- and IP3R2-HEK cells (Figure 4, left panels). On average, R_r in WT-, IP3R1-, IP3R2-, and IP3R3-HEK cells was 0.037 ± 0.011, 0.017 ± 0.005, 0.026 ± 0.008, and 0.049 ± 0.013 (ΔF/F₀)s⁻¹ (mean ± S.D.), respectively. This indicates that Ca²⁺ leakage in WT- and IP3R3-HEK cells was more intensive than that in IP3R1- and IP3R2-HEK cells.

For SOCE observed after the 600 s depletion of Ca²⁺ store (Figure 4, right panels), averaged R_r was 0.031 ± 0.009, 0.023 ± 0.007, 0.038 ± 0.012, and 0.018 ± 0.006 (ΔF/F₀)s⁻¹ in WT-, IP3R1-, IP3R2-, and IP3R3-HEK cells, respectively. The histograms generated for R_r also demonstrated cell-line specificity (Figure 4, right panels). Although IP3R1- and IP3R3-HEK cells were comparable to the distributions (Figure 4, right IP3R1 and IP3R3 panels) and averaged values of SOCE rates, SOC activity was higher in IP3R2-HEK cells and in WT-HEK cells (Figure 4, right panels of WT and IP3R2). These findings point at the possibility that IP₃R2 could be functionally coupled to Ca²⁺ channels that mediate SOCE in WT- and IP3R2-HEK cells. In contrast, in H4IIE liver cells, which also express all three IP3R isoforms, primarily type 1 and, to a lesser extent, type 3, but not type 2, participated in the activation of a CRAC current associated with SOCE [48].

In resting cells, spontaneous activity of IP₃Rs could be a factor of Ca²⁺ leakage from Ca²⁺ store [46]. Being least active at rest, IP₃R3 should have contributed to Ca²⁺ leakage to a lesser extent compared to the other IP₃R subtypes. However, it turned out that just Ca²⁺ store in IP3R3-HEK cells was most leaky in terms of the initial rate of thapsigargin-induced Ca²⁺ release (Figure 4, left panels). Note, however, that the Ca²⁺ release rate depended not only on Ca²⁺ permeability of the reticular membrane but also on a level of stored Ca²⁺. Thus, the Ca²⁺ release rate could not serve as an independent measure of ER permeability to Ca²⁺. To address this issue, we developed a simplified kinetics model of Ca²⁺ signals triggered by thapsigargin at low bath Ca²⁺ (Figure S13) and found the Ca²⁺ permeability P of the ER membrane to be proportional to the ratio (Equation (S11)):

$$P~\frac{\frac{dC}{dt}\left(0\right)}{{\int }_0^{T}Cdt}$$

(2)

where C is the concentration of cytosolic Ca²⁺, $\frac{dC}{dt}\left(0\right)$ is the initial rate of a Ca²⁺ rise triggered by thapsigargin applied at t = 0, ${\int }_0^{T}Cdt$ is the area under Ca²⁺ release curve, and T is the time interval necessary for cytosolic Ca²⁺ to return to the initial level (Figure 5A).

To employ this formalism, we suggested that Fluo-8 fluorescence was far below saturation. Indeed, normally, thapsigargin-induced Ca²⁺ responses did not exceed 2 (Figure 3A), while the dynamic range of Fluo-8 should have been 20 at least [49]. Therefore, the measured parameter F/F₀ was nearly proportional to the concentration of cytosolic Ca²⁺. For each assayed cell, we evaluated both the initial rate of Ca²⁺ release (Figure 3B–E, bottom panels) and the area under the Ca²⁺-release trace (Figure 5A, hatched area).

The appropriate histograms of $\frac{dC}{dt}\left(0\right)$ and ${\int }_0^{T}Cdt$ obtained for particular IP₃R isoforms are presented in Figure 5B–D (left and middle panels). As illustrated, thapsigargin induced much more massive Ca²⁺ release in IP3R3-HEK cells (Figure 5D, middle panel) compared to IP3R1-HEK and IP3R2-HEK cells (Figure 5B,C, middle panels). The individual ratios (Equation (2)) were calculated, and their distributions among cells of the particular subtype were generated (Figure 5B,C, right panels). For IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK cells, the averaged ratios (Equation (2)) were 1.42 ± 0.49, 2.49 ± 0.94, and 0.64 ± 0.17, respectively, and the differences between them were statistically significant (p < 0.05, ANOVA test). Based on these values of the ratio (2), relative Ca²⁺ permeability of the ER membrane was estimated for IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK cells as 1:1.75:0.45, respectively. Thus, based on Ca²⁺ permeability of the ER, assayed cells were arranged as IP3R2-HEK > IP3R1-HEK > IP3R3-HEK. This order was rather consistent with the values of the steady-state open probabilities of IP₃R2, IP₃R1, and IP₃R3 found to be ~0.3, 0.1, and <0.1, respectively, at nearly resting conditions (100 nM Ca²⁺, 1 μM IP₃) [19]. In agreement with the previous report [39], this conformity suggested that spontaneous activity of IP₃Rs was an essential factor of Ca²⁺ leakage from the ER. Note that the total Ca²⁺ release for 600 s (Figure 5B–D, middle panels) indicated that even being least permeable to Ca²⁺ (Figure 5B–D, right panels), Ca²⁺ store in IP3R3-HEK cells lost a larger number of Ca²⁺ ions compared to IP3R2-HEK and IP3R1-HEK cells. It was possible only if a resting level of Ca²⁺ in IP₃-regulated Ca²⁺ store in IP3R3-HEK cells was essentially higher than one in IP3R2-HEK and IP3R1-HEK cells, provided that the ER volume was invariable among the cell lines.

3.4. IP3R1-, IP3R2-, and IP3R3-HEK Cells with the Ca²⁺ Sensor R-CEPIA1er

To extend the experimental capability of the engineered lines, R-CEPIA1er, the Ca²⁺sensor with a reticular location [40], was heterologously expressed in IP3R1-, IP3R2-, and IP3R3-HEK cells. Being loaded with Fluo-8, R-CEPIA1er-positive cells allowed for simultaneous monitoring of cytosolic and reticular Ca²⁺ (Figure 6). ACh stimulated Ca²⁺ transients in the cell cytosol (Figure 5A–C, upper panels) and a synchronous drop in reticular Ca²⁺, which relaxed close to the resting level despite the agonist still being present in the bath (Figure 6A–C, bottom panels). The Ca²⁺ ionophore ionomycin (5 μM) applied at low bath Ca²⁺ (260 nM) also triggered cytosolic Ca²⁺ signals (Figure 6A–C), presumably by penetrating through plasmalemma and increasing Ca²⁺ permeability of the reticular membrane. In this case, the low steady-state level of the ionomycin response was achievable if Ca²⁺ fluxes mediated by SERCA and ionomycin were precisely balanced. The relative effects of the agonist and ionophore on reticular Ca²⁺ were quantified by the A₁/A₂ ratio, where magnitudes of Ca²⁺ signals elicited by ACh (A₁) and ionomycin (A₂) were determined as indicated in Figure 6B. As summarized in Figure 6D, ionomycin emptied Ca²⁺ store in IP3R1- and IP3R3-HEK cells to a much higher extent than ACh did (Figure 6A,C, bottom panels). In contrast, ACh and ionomycin dropped luminal Ca²⁺ in IP3R2-HEK cells to comparable levels (Figure 6B, bottom panels).

This phenomenon could be plausibly interpreted based on the recent finding that luminal Ca²⁺ inhibited activity of IP₃Rs and related IP₃-dependent Ca²⁺ signals in the cell cytosol [50]. This inhibitory effect was presumably mediated by the Ca²⁺-binding protein annexin A1 (ANXA1). With no ANXA1 bound, IP₃-gated channels were sufficiently active, while at high luminal Ca²⁺ (>100 μM), ANXA1 interacted with IP₃Rs, promoting their inhibition [50]. Given that ANXA1 is expressed in WT-HEK cells (Figure S8), the abovementioned ANXA1-mediated regulation could explain why ACh and ionomycin reduced luminal Ca²⁺ to close levels in IP3R2-HEK cells but not in IP3R1-HEK and IP3R3-HEK cells. Indeed, our findings suggest that at rest, ER permeability to Ca²⁺ was highest in IP3R2-HEK cells (Figure 5B–D, right panels), implying that a resting level of luminal Ca²⁺ should have been lower in these cells compared to IP3R1-HEK and IP3R3-HEK cells, provided that SERCA was similarly active in all cell subgroups. If so, IP₃Rs in IP3R2-HEK cells operated in a mode characterized by a higher open probability at the same IP₃ level [50], thus mediating higher Ca²⁺ release despite lower luminal Ca²⁺.

4. Discussion

The Ca²⁺ homeostasis in the ER is governed by a complex protein network, which ensures the steady-state Ca²⁺ level in the ER lumina at rest, precise Ca²⁺ release upon cell stimulation, and effective refilling of Ca²⁺ store [47,51,52,53]. At rest, the level of free Ca²⁺ in the ER lumina, which ranges between 300 and 800 μM [51], is determined by balance between passive Ca²⁺ leakage and active ER refilling by the SERCA pump [47]. The inhibition of SERCA with thapsigargin completely depletes a Ca²⁺ pool in the ER within minutes, indicating that Ca²⁺ constantly leaks from the ER [54]. Several mechanisms have been implicated in mediating Ca²⁺ leakage, including proteins from the transmembrane BAX inhibitor motif-containing (TMBIM) family, the antiapoptotic protein BCL-2, and a truncated version of SERCA pump SERCA1T [47,54]. The effective feedback mechanism preventing the overload of ER with Ca²⁺ involves TMCO1 proteins, which are capable of oligomerizing at high luminal Ca²⁺ to form transient Ca²⁺ leak channels [55]. Evidence exists that in resting cells, spontaneous activity of IP₃Rs and ryanodine receptors (RyRs) could be responsible for a fraction of Ca²⁺ influx from the ER [47,54]. For instance, in HEK-293 cells, knockout of all three IP₃R genes markedly slowed spontaneous Ca²⁺ release from the ER [39], indicating that resting activity of IP₃Rs was a significant factor in Ca²⁺ leakage. On the other hand, our observations suggest that RyRs contributed negligibly to Ca²⁺ leakage from the ER in HEK-293 cells (Figure S12).

The sustained depletion of luminal Ca²⁺ is detrimental to cells, as it causes ER stress, inhibits protein synthesis, and provokes apoptosis [56,57]. Cells employ a number of mechanisms to counteract the prolonged depletion of Ca²⁺ store, including Ca²⁺-binding proteins buffering luminal Ca²⁺ [58], coupling of Ca²⁺ store depletion to activated SOCE [59], and active reloading of ER with SERCA [60]. The core mechanism of SOCE involves stromal interaction molecules (STIMs), basically STIM1 but also STIM2, which serve as sensors of ER Ca²⁺ and SOCE regulators [36]. Being initiated by Ca²⁺ store depletion, the dissociation of Ca²⁺ from STIM1 proteins causes their oligomerization and relocation to the specialized membrane contact sites between the plasma membrane and the ER, called the ER–PM junction. In this junction, the cytosolic domains in STIM1 oligomers have an unfurling and elongated conformation necessary to capture Orai channels located in the plasmalemma and enable their opening [36].

In cells of diverse types, IP₃Rs represents the main conduit for stimulus-dependent Ca²⁺ release [3,9,46], and therefore they should be coupled to SOCs, at least functionally. The functional interaction was indeed demonstrated for IP₃Rs and TRPC channels involved in SOCE [56,57,58,61,62] as well as between IP₃Rs and ORAI1 [59,60]. Moreover, evidence exists that STIM proteins can directly interact with IP₃Rs [36,63]. The recent findings point out that STIM1 forms a complex predominantly with activated IP₃R [63]. Indeed, being a trigger of STIM1 oligomerization, a transient fall in intraluminal Ca²⁺ initiated by IP₃ should be most pronounced just in the close vicinity of open IP₃R.

The cell lines expressing merely one IP₃R subtype represent a promising cellular model for the systematic assay of gating, regulation, pharmacology, and physiology of individual IP₃R isoforms. The first vertebrate cellular model suitable for functional analysis of individual IP₃R isotypes in the same cellular background was established based on chicken lymphoma-derived DT40 cells, wherein all three IP₃R genes were disrupted (DT40-TKO cells) [33]. The heterologous expression of mammalian IP₃Rs in DT40-TKO cells provided a deep insight into the functionality of individual IP₃R isotypes. It was particularly demonstrated that constitutive Ca²⁺ release through IP₃R to mitochondria is required for mitochondrial respiration and maintenance of normal cell bioenergetics [34]. It turned out that IP₃R2 delivered Ca²⁺ to mitochondria most effectively, although each IP₃R isotype can support local contact sites of ER and mitochondria [38]. The permeabilized cells expressing a particular IP₃R subunit were employed to assay Ca²⁺ release stimulated by IP₃ and its synthetic analogs. The dose–response curves generated for all IP₃R subtypes yielded the first structure–activity relationships for the key IP₃ analogues [35]. Reportedly, IP₃-induced Ca²⁺ release was effective only if each IP₃R monomer within the tetramer was occupied by IP₃ [36].

Apart from DT40-TKO cells, IP₃R-defficient HEK-293 and HeLa cells have also been generated by using CRISPR/Cas9 genome editing [36,37,39]. Being employed as a heterologous system for systematic expression and assay of mammalian IP₃Rs, these model lines facilitated the acquisition of a number of interesting findings. It was particularly reported that upon IP₃ uncaging, all three IP₃R subtypes were capable of mediating Ca²⁺ puffs, relatively small and localized Ca²⁺ transients. The pathological mutations associated with dysfunction of IP₃R1 disrupted IP₃ binding, IP₃-mediated gating, and its regulation by IP₃R-modulatory proteins [37]. Being significantly reduced in HEK293-TKO cells, Ca²⁺ leakage from the ER and its refilling were rescued by overexpression of recombinant IP₃R1 or IP₃R3 [39].

In the present work, we developed our own cell lines suitable for the analysis of a role of individual IP₃R isotypes in agonist-induced Ca²⁺ signaling. By inactivating two out of three IP₃R genes in HEK-293 cells with CRISPS/Cas9 technology and employing cell selection methods, we generated three monoclonal cell lines, IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK, with IP₃R1, IP₃R2, and IP₃R3 being solely functional, respectively. The functional consequence of this meddling in the natural pattern of IP₃R expression was evaluated by studying certain aspects of Ca²⁺ signaling in these genetically modified cells.

In each line, IP3R1-HEK, IP3R2-HEK, or IP3R3-HEK, cells responded to ACh with Ca²⁺ mobilization and exhibited “all-or-nothing” responsivity to the agonist (Figure 2), as was the case with WT-HEK cells (Figure 1D). Given this feature, a dose dependence of cell responses was characterized at the population level by the number of cells responsive to ACh at a particular concentration. Based on the EC₅₀ doses obtained from these dose-response relationships (Figure 2B), ACh sensitivities of the genetically modified cells ranked as IP3R2-HEK > IP3R1-HEK > IP3R3-HEK. Response lags basically obeyed the same sequence (Figure 2C,D).

It should be noted that genome editing with RNA-guided nucleases, such as Cas9, can entail off-target DNA cleavage [61,62]. As an additional control, we compared assayed cells based on the expression of muscarinic receptors and several downstream signaling proteins that could be involved in ACh transduction, including G_q- and G_i1-3-proteins, which are known to couple a variety of GPCRs to G-protein-regulated PLCβ1-4 [56]. The relative expression levels were assessed using RT-qPCR (see Supplementary Materials).

In mammals, five genes encode muscarinic receptors of the M1–M5 subtypes. We identified M1, M3, M4, and M5 transcripts in WT-HEK cells, while M2 transcripts were undetectable (Figure S7A). Among them, the M3 isotype was dominant at the transcript level (Figure S7B). Consistently, physiological evidence validated the central role of the M3 receptor in mediating ACh-induced Ca²⁺ mobilization (Figure 1D). Although expression of the identified M receptors somewhat varied from line to line (Figure S9A), M3 transcript levels were statistically indistinguishable (Figure S9A, M3 panel). It thus appears that distinct sensitivities of IP3R1-, IP3R2-, and IP3R3-HEK cells to ACh (Figure 2B) were determined by a mechanism downstream of the M receptors.

Given that the M3 receptor primarily couples to the G_q protein [57], the distinct sensitivities of IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK cells to ACh may be attributed to lineage-specific expression of G_q. It was found that the IP3R1-HEK/IP3R2-HEK and IP3R2-HEK/IP3R3-HEK pairs exhibited statistically indistinguishable levels of G_q expression (Figure S9B, G_q panel). On the other hand, the level of G_q transcripts in IP3R3-HEK cells was ~50% lower than in IP3R1-HEK cells (Figure S9B, G_q panel). These data revealed no correlation between a level of G_q expression in a particular cell line (Figure S9B, G_q panel) and its responsiveness to ACh (Figure 2B). In addition, levels of G_i1-3 transcripts in all assayed lines were statistically indistinguishable (Figure S9B, G_i panels). Hence, the different sensitivities of IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK cells to ACh (Figure 2B) did not seem correlated with differences in their expression of G proteins (Figure S9B), which coupled M receptors to PLC.

Although expression of PLCβ1–4 was more scattered among cell populations, statistically significant deviations exhibited solely PLCβ3 transcripts (Figure S6C). Given, however, that compared to PLCβ1 and PLCβ2 the level of PLCβ3 transcripts was lower by a factor of 3–10 (Figure S8C), this PLC isoform was presumably a minor contributor to ACh signaling (Figure 2). In summary, the abovementioned results (Figures S7 and S8) indicate that the IP₃R gene editing has had insignificant impact on the expression of proteins potentially crucial to ACh transduction.

Previously, we developed a mathematical model of agonist transduction that included the phosphoinositide signaling cascade and IP₃-driven Ca²⁺ release through the only type of IP₃Rs [31]. This model properly simulated the characteristic features of agonist-induced Ca²⁺ signaling, such as the “all-or-nothing” responsivity of cells (Figure 2A) and the dose dependence of the response lag (Figure 2D). In line with this model, a threshold agonist concentration in a step-like dose dependence of Ca²⁺ responses (Figure 4 in [31]) was determined by both a rate of IP₃ production stimulated by an agonist and the affinity of IP₃ binding to IP₃Rs. Note that the expression analysis (Figure S9) suggests that the efficacy of the ACh transduction pathway, i.e., muscarinic receptor–G protein–PLC–IP₃ production, was likely similar or nearly identical in IP3R1-, IP3R2-, and IP3R3-HEK cells. On the other hand, based on the revealed levels of IP₃R transcripts (Figure S10), these cells were ranked as IP3R3-HEK > IP3R2-HEK > IP3R3-HEK, although based on ACh responsivity, they were ordered as IP3R2-HEK > IP3R1-HEK > IP3R3-HEK (Figure 2B). The last sequence aligned well with the affinities of the corresponding IP₃Rs to IP₃, which followed the order IP₃R2 > IP₃R1 > IP₃R3 [7,11,58,59,60]. It thus appears that the IP₃ affinity of an IP₃R isoform operating in a particular cell line was a decisive factor that determined sensitivity of the assayed cells to ACh.

5. Conclusions

In this study, we generated our own monoclonal cell lines, IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK, for further analysis of the role of individual IP₃R isotypes in mediating various aspects of agonist-induced Ca²⁺ signaling. For instance, the IP3R3-HEK line could provide sufficient insight into why type II taste cells exclusively involve IP₃R3 in sweet, bitter, and umami transduction [16]. We utilized the generated cellular models to verify several important findings reported earlier. Specifically, our data confirmed that each IP₃R isoform could mediate the CICR process and contribute to Ca²⁺ leakage from Ca²⁺ store at rest. Furthermore, we developed a mathematical model, which allowed for the relative Ca²⁺ permeability of Ca²⁺ store to be estimated based on Ca²⁺ signals induced by thapsigargin in the cell cytosol. With this approach, relative Ca²⁺ permeabilities of Ca²⁺ store in IP3R1-HEK, IP3R2-HEK, and IP3R3-HEK cells were evaluated to be 1:1.75:0.45.

The engineered cells responded to ACh in strong correlation with the IP₃ sensitivity of an IP₃R isoform they expressed. Based on this correlation, any alteration in responsivity of the particular cell line, induced by a pharmacological agent targeting IP₃Rs, could be attributed to a change in activity of the related IP₃R isoform. We therefore anticipate that the developed cell lines, in combination with genetically encoded sensors (Figure 6), could provide a relatively straightforward and efficient means for assaying the activity, regulation, and pharmacology of individual IP₃R isoforms.