1. Cell Biology
  2. Neuroscience

Regulation of store-operated Ca2+ entry by IP3 receptors independent of their ability to release Ca2+

  1. Pragnya Chakraborty
  2. Bipan Kumar Deb
  3. Vikas Arige
  4. Thasneem Musthafa
  5. Sundeep Malik
  6. David I Yule
  7. Colin W Taylor  Is a corresponding author
  8. Gaiti Hasan  Is a corresponding author
  1. National Centre for Biological Sciences, Tata Institute of Fundamental Research, India
  2. SASTRA University, India
  3. Department of Pharmacology and Physiology, University of Rochester, United States
  4. Department of Pharmacology, University of Cambridge, United Kingdom
https://doi.org/10.7554/eLife.80447
Share this article
Cite this article
  1. Pragnya Chakraborty
  2. Bipan Kumar Deb
  3. Vikas Arige
  4. Thasneem Musthafa
  5. Sundeep Malik
  6. David I Yule
  7. Colin W Taylor
  8. Gaiti Hasan
(2023)
Regulation of store-operated Ca2+ entry by IP3 receptors independent of their ability to release Ca2+
eLife 12:e80447.
https://doi.org/10.7554/eLife.80447
  2. Download BibTeX
  3. Download .RIS

Abstract

Loss of endoplasmic reticular (ER) Ca2+ activates store-operated Ca2+ entry (SOCE) by causing the ER localized Ca2+ sensor STIM to unfurl domains that activate Orai channels in the plasma membrane at membrane contact sites (MCS). Here, we demonstrate a novel mechanism by which the inositol 1,4,5 trisphosphate receptor (IP3R), an ER-localized IP3-gated Ca2+ channel, regulates neuronal SOCE. In human neurons, SOCE evoked by pharmacological depletion of ER-Ca2+ is attenuated by loss of IP3Rs, and restored by expression of IP3Rs even when they cannot release Ca2+, but only if the IP3Rs can bind IP3. Imaging studies demonstrate that IP3Rs enhance association of STIM1 with Orai1 in neuronal cells with empty stores; this requires an IP3-binding site, but not a pore. Convergent regulation by IP3Rs, may tune neuronal SOCE to respond selectively to receptors that generate IP3.

Editor's evaluation

This paper proposes a fundamental new role for IP3 receptors in the regulation of store-operated calcium entry in neurons, in which IP3-bound receptors enhance the association of STIM1 and Orai1 independently of their ability to release Ca from the ER. While the evidence for this phenomenon is solid, experimental support for an underlying mechanism is incomplete and will require additional studies. The paper will appeal to cell biologists and neurobiologists interested in calcium signaling pathways, particularly store-operated calcium entry.

https://doi.org/10.7554/eLife.80447.sa0

Introduction

The activities of all eukaryotic cells are regulated by increases in cytosolic-free Ca2+ concentration ([Ca2+]c), which are almost invariably evoked by the opening of Ca2+-permeable ion channels in biological membranes. The presence of these Ca2+ channels within the plasma membrane (PM) and the membranes of intracellular Ca2+ stores, most notably the endoplasmic reticulum (ER), allows cells to use both intracellular and extracellular sources of Ca2+ to evoke Ca2+ signals. In animal cells, the most widely expressed Ca2+ signaling sequence links extracellular stimuli, through their specific receptors and activation of phospholipase C, to formation of inositol 1,4,5-trisphosphate (IP3), which then stimulates Ca2+ release from the ER through IP3 receptors (IP3R) (Foskett et al., 2007; Prole and Taylor, 2019). IP3Rs occupy a central role in Ca2+ signaling by releasing Ca2+ from the ER. IP3Rs thereby elicit cytosolic Ca2+ signals, and by depleting the ER of Ca2+ they initiate a sequence that leads to activation of store-operated Ca2+ entry (SOCE) across the PM (Putney, 1986; Thillaiappan et al., 2019). SOCE occurs when loss of Ca2+ from the ER causes Ca2+ to dissociate from the luminal Ca2+-binding sites of an integral ER protein, stromal interaction molecule 1 (STIM1). STIM1 then unfolds its cytosolic domains to expose a region that binds directly to a Ca2+ channel within the PM, Orai, causing it to open and Ca2+ to flow into the cell across the PM (Parekh and Putney, 2005; Prakriya and Lewis, 2015; Lewis, 2020). The interactions between STIM1 and Orai occur across a narrow gap between the ER and PM, a membrane contact site (MCS), where STIM1 puncta trap Orai channels. While STIM1 and Orai are undoubtedly the core components of SOCE, many additional proteins modulate their interactions (Rosado et al., 2000; Palty et al., 2012; Deb et al., 2016; Srivats et al., 2016) and other proteins contribute by regulating the assembly of MCS (Chang et al., 2013; Giordano et al., 2013; Kang et al., 2019).

It is accepted that IP3-evoked Ca2+ release from the ER through IP3Rs is the usual means by which extracellular stimuli evoke SOCE. Here, the role of the IP3R is widely assumed to be restricted to its ability to mediate Ca2+ release from the ER and thereby activate STIM1. Evidence from Drosophila, where we suggested an additional role for IP3Rs in regulating SOCE (Agrawal et al., 2010; Chakraborty et al., 2016), motivated the present study, wherein we examined the contribution of IP3Rs to SOCE in mammalian neurons. We show that in addition to their ability to activate STIM1 by evoking ER Ca2+ release, IP3Rs also facilitate interactions between active STIM1 and Orai1. This additional role for IP3Rs, which is regulated by IP3 but does not require a functional pore, reveals an unexpected link between IP3, IP3Rs and Ca2+ signaling that is not mediated by IP3-evoked Ca2+ release. We speculate that dual regulation of SOCE by IP3Rs may allow Ca2+ release evoked by IP3 to be preferentially coupled to SOCE.

Results

Loss of IP3R1 attenuates SOCE in human neural stem cells and neurons

We investigated the effects of IP3Rs on SOCE by measuring [Ca2+]c in human neural stem cells and neurons prepared from embryonic stem cells. Human neural progenitor cells (hNPCs) were derived from H9 embryonic stem cells using small molecules that mimic cues provided during human brain development (Gopurappilly et al., 2018). We confirmed that hNPCs express canonical markers of neural stem cells (Figure 1A) and that IP3R1 is the predominant IP3R subtype (GEO accession no. GSE109111; Gopurappilly et al., 2018). An inducible lentiviral shRNA-miR construct targeting IP3R1 reduced IP3R1 expression by 93 ± 0.4% relative to a non-silencing (NS) construct (Figure 1B and C). Carbachol stimulates muscarinic acetylcholine receptors, which are expressed at low levels in hNPCs (Gopurappilly et al., 2018). In Ca2+-free medium, carbachol evoked an increase in [Ca2+]c in about 10% of hNPCs, consistent with it stimulating Ca2+ release from the ER through IP3Rs. Restoration of extracellular Ca2+ then evoked an increase in [Ca2+]c in all cells that responded to carbachol. Both carbachol-evoked Ca2+ release and SOCE were abolished in hNPCs expressing IP3R1-shRNA, confirming the effectiveness of the IP3R1 knockdown (Figure 1—figure supplement 1A–C).

Figure 1 with 1 supplement see all
Download asset Open asset
Loss of IP3R1 attenuates SOCE in human neural stem cells.

(A) Confocal images of hNPCs (passage 6) stained for DAPI and neural stem cell proteins: Pax6 and Ki67 (proliferation marker). Scale bars, 50 μm. (B) WB for IP3R1 of hNPCs expressing non-silencing (NS) or IP3R1-shRNA. (C) Summary results (mean ±s.d., n=3) show IP3R1 expression relative to actin. **p < 0.01, Student’s t-test with unequal variances. (D) Changes in [Ca2+]c evoked by thapsigargin (Tg, 10 µM) in Ca2+-free HBSS and then restoration of extracellular Ca2+ (2 mM) in hNPCs expressing NS or IP3R1-shRNA. Mean ± s.e.m. from hree independent experiments, each with four replicates that together included 100–254 cells. Inset shows the target of Tg. (E–G) Summary results (individual cells, median (bar), 25th and 75th percentiles (box) and mean (circle)) show Ca2+ signals evoked by Tg or Ca2+ restoration (E), rate of Ca2+ entry (F) and resting [Ca2+]c (G). ***p < 0.001, Mann-Whitney U-test. (H) Changes in [Ca2+]c evoked by Tg (10 µM) in Ca2+-free HBSS and after restoring extracellular Ca2+ (2 mM) in neurons (differentiated hNPCs) expressing NS or IP3R1-shRNA. Mean ± s.e.m. from three experiments with ~200 cells. (I,J) Summary results (presented as in E-G) show Ca2+ signals evoked by Tg or Ca2+ restoration (I) and rate of Ca2+ entry (J). ***p < 0.001. Mann-Whitney U-test. See also Figure 1—figure supplement 1. Source data in Figure 1—source data 1.

Figure 1—source data 1

Loss of IP3R1 attenuates SOCE in human neural stem cells.

https://cdn.elifesciences.org/articles/80447/elife-80447-fig1-data1-v2.zip

Thapsigargin, a selective and irreversible inhibitor of the ER Ca2+ pump (sarcoplasmic/endoplasmic reticulum Ca2+-ATPase, SERCA), was used to deplete the ER of Ca2+ and thereby activate SOCE (Figure 1D; Parekh and Putney, 2005). Restoration of extracellular Ca2+ to thapsigargin-treated hNPCs evoked a large increase in [Ca2+]c, reflecting the activity of SOCE (Figure 1D). The maximal amplitude and rate of SOCE were significantly reduced in cells lacking IP3R1, but the resting [Ca2+]c and thapsigargin-evoked Ca2+ release were unaffected (Figure 1D–F and Figure 1—figure supplement 1D and E). STIM1 and Orai1 expression were also unaltered in hNPC lacking IP3R1 (Figure 1—figure supplement 1G). After spontaneous differentiation of hNPC, cells expressed markers typical of mature neurons, and the cells responded to depolarization with an increase in [Ca2+]c (Figure 1—figure supplement 1F and Figure 1—figure supplement 1H–J). Thapsigargin evoked SOCE in these differentiated neurons; and expression of IP3R1-shRNA significantly reduced the SOCE response without affecting depolarization-evoked Ca2+ signals (Figure 1H–J and Figure 1—figure supplement 1H–L).

Loss of IP3R1 attenuates SOCE in human neuroblastoma cells

IP3Rs link physiological stimuli that evoke Ca2+ release from the ER to SOCE, but the contribution of IP3Rs is thought to be limited to their ability to deplete the ER of Ca2+. We have reported that in Drosophila neurons there is an additional requirement for IP3Rs independent of ER Ca2+ release (Venkiteswaran and Hasan, 2009; Agrawal et al., 2010; Chakraborty et al., 2016). Our results with hNPCs and stem cell-derived neurons suggest a similar requirement for IP3Rs in regulating SOCE in mammalian neurons. To explore the mechanisms underlying this additional role for IP3Rs, we turned to a more tractable cell line, SH-SY5Y cells. These cells are derived from a human neuroblastoma; they exhibit many neuronal characteristics (Agholme et al., 2010); they express M3 muscarinic acetylcholine receptors that evoke IP3-mediated Ca2+ release and SOCE (Grudt et al., 1996); and they express predominantly IP3R1 (Wojcikiewicz, 1995; Tovey et al., 2001), with detectable IP3R3, but no IP3R2 (Figure 2A). We used inducible expression of IP3R1-shRNA to significantly reduce IP3R1 expression (by 74 ± 1.2%), without affecting IP3R3 (Figure 2A and B). As expected, carbachol-evoked Ca2+ signals in individual SH-SY5Y cells were heterogenous and the carbachol-evoked Ca2+ release was significantly reduced by knockdown of IP3R1 (Figure 2C and D and Figure 2—figure supplement 1A and B). Thapsigargin evoked SOCE in SH-SY5Y cells (Grudt et al., 1996), and it was significantly attenuated after knockdown of IP3R1 without affecting resting [Ca2+]c, the Ca2+ release evoked by thapsigargin or expression of STIM1 and Orai1 (Figure 2E–G and Figure 2—figure supplement 1C–E).

Figure 2 with 3 supplements see all
Download asset Open asset
Loss of IP3R1 attenuates SOCE in SH-SY5Y cells.

(A) WB for IP3R1-3 of SH-SY5Y cells expressing non-silencing (NS) or IP3R1-shRNA. (B) Summary results (mean ± s.d., n=4) show IP3R expression relative to actin normalized to control NS cells. **p < 0.01, Student’s t-test with unequal variances. (C) Ca2+ signals evoked by carbachol (CCh, 3 µM) in SH-SY5Y cells expressing NS or IP3R1-shRNA. Mean ± s.e.m. from three experiments with 70–90 cells. (D) Summary results show peak changes in [Ca2+]c (Δ[Ca2+]c) evoked by CCh. ***p < 0.001, Mann-Whitney U-test. (E) Ca2+ signals evoked by thapsigargin (Tg, 10 µM) in Ca2+-free HBSS and then after restoration of extracellular Ca2+ (2 mM) in cells expressing NS or IP3R1-shRNA. Mean ± s.e.m. from three experiments with ~50 cells. (F, G) Summary results (individual cells, mean ± s.e.m., n=3, ~50 cells) show peak changes in [Ca2+]c evoked by Ca2+ restoration (Δ[Ca2+]c) (F) and rate of Ca2+ entry (G). ***p < 0.001, Mann-Whitney U-test. (H) Ca2+ signals evoked by Tg and then Ca2+ restoration in cells expressing NS-shRNA, or IP3R1-shRNA alone or with IP3R1 or IP3R3. Traces show mean ± s.e.m. (50–115 cells from three experiments). (I, J) Summary results (mean ± s.e.m, 50–115 cells from three experiments) show peak increases in [Ca2+]c (Δ[Ca2+]c) evoked by Ca2+ restoration (I) and rates of Ca2+ entry (J) evoked by restoring extracellular Ca2+. (K) Effects of thapsigargin (Tg, 10 µM) in Ca2+-free HBSS and then after Ca2+ restoration (2 mM) in cells expressing IP3R1-shRNA alone or with IP3R1 or mCh-STIM1. Traces show mean ± s.e.m. (100–150 cells from three experiments). (L, M) Summary results (mean ± s.e.m.) show peak increase in [Ca2+]c after Ca2+ restoration (Δ[Ca2+]c) (L) and rate of Ca2+ entry (M). Different letters indicate significant differences (panels I, J, L, M), p <0.001, one-way ANOVA with pair-wise Tukey’s test. See also Figure 2—figure supplements 1–3. Source data in Figure 2—source data 1.

Figure 2—source data 1

Loss of IP3R1 attenuates SOCE in SH-SY5Y cells.

https://cdn.elifesciences.org/articles/80447/elife-80447-fig2-data1-v2.zip

We also used CRISPR/Cas9n and Cas9 to disrupt one or both copies of the IP3R1 gene, subsequently referred to as IKO (one copy knockout) and IKO null (both copies knocked out) in SH-SY5Y cells. IP3R1 expression was absent in the IKO null (Figure 2—figure supplement 1F) whereas expression of STIM1, STIM2 and Orai1 were unperturbed (Figure 2—figure supplement 1G). Carbachol-evoked Ca2+ release and thapisgargin-evoked SOCE were significantly reduced (Figure 2—figure supplement 1H–J). Since the IKO null cells were fragile and grew slowly, we examined SOCE in SH-SY5Y cells with disruption of one copy of the IP3R1 gene. In the IKO cells, IP3R1 expression, carbachol-evoked Ca2+ signals and thapsigargin-evoked SOCE were all reduced (Figure 2—figure supplement 1K–Q).

These observations, which replicate those from hNPCs and neurons (Figure 1), vindicate our use of SH-SY5Y cells to explore the mechanisms linking IP3Rs to SOCE in human neurons.

Expression of IP3R1 or IP3R3 in SH-SY5Y cells expressing IP3R1-shRNA restored both carbachol-evoked Ca2+ release and thapsigargin-evoked SOCE without affecting resting [Ca2+]c or thapsigargin-evoked Ca2+ release (Figure 2H–J and Figure 2—figure supplement 2A–D). Over-expression of STIM1 in cells expressing NS-shRNA had no effect on SOCE (Figure 2—figure supplement 2E and F), but it restored thapsigargin-evoked SOCE in cells expressing IP3R1-shRNA, without affecting resting [Ca2+]c or thapsigargin-evoked Ca2+ release (Figure 2K–M). We conclude that IP3Rs are required for optimal SOCE, but they are not essential because additional STIM1 can replace the need for IP3Rs (Figure 3A).

Figure 3 with 1 supplement see all
Download asset Open asset
Regulation of SOCE by IP3R requires IP3 binding but not a functional pPore in SH-SY5Y cells.

(A) SOCE is activated when loss of Ca2+ from the ER through IP3Rs activates STIM1 (i). Our results suggest an additional role for IP3Rs (ii). (B) SH-SY5Y cells expressing IP3R1-shRNA alone or with IP3R1 or IP3R1DA were stimulated with thapsigargin (Tg, 1 µM) in Ca2+-free HBSS before restoring extracellular Ca2+ (2 mM). Traces show mean ± s.e.m, for 100–150 cells from three experiments. (C) Cells expressing IP3R1-shRNA and IP3R1DA were treated with NS-siRNA or Orai1-siRNA before measuring Tg-evoked Ca2+ entry. Traces show mean ± s.e.m. for 85–100 cells from three experiments. (D) Summary results (mean ± s.e.m.) show peak increases in [Ca2+]c (Δ[Ca2+]c) evoked by Ca2+ restoration. (E) Tg-evoked Ca2+ entry in cells expressing IP3R1-shRNA with IP3R1, IP3R1RQ or IP3R1RQ/KQ. Traces show mean ± s.e.m, for 90–150 cells from three experiments. (F) Summary results (mean ± s.e.m.) show peak increases in [Ca2+]c (Δ[Ca2+]c) evoked by Ca2+ restoration. Different letter codes (panels D, F) indicate significantly different values, p<0.001, for multiple comparison one-way ANOVA and pair-wise Tukey’s test and for two genotype comparison Mann Whitney U-test. See also Figure 3—figure supplement 1—source data 1. Source data in Figure 3—source data 1.

Figure 3—source data 1

Regulation of SOCE by IP3R requires IP3 binding but not a functional pPore in SH-SY5Y cells.

https://cdn.elifesciences.org/articles/80447/elife-80447-fig3-data1-v2.zip

It has been reported that SOCE is unaffected by loss of IP3R in non-neuronal cells (Ma et al., 2001; Chakraborty et al., 2016). Consistent with these observations, the SOCE evoked in HEK cells by stores emptied fully by treatment with thapsigargin was unaffected by expression of IP3R1 shRNA (Figure 2—figure supplement 3A–3C) or by knockout of all three IP3R subtypes using CRISPR/cas9 (HEK-TKO cells; Figure 2—figure supplement 3D and E). The association of STIM1 with Orai1 in wild type HEK cells and HEK TKO cells after thapsigargin-evoked store depletion also appeared identical as tested by a proximity ligation assay (PLA, described further in Figure 5 and Figure 2—figure supplement 3F). Neuronal and non-neuronal cells may, therefore, differ in the contribution of IP3R to SOCE. We return to this point later.

Binding of IP3 to IP3R without a functional pore stimulates SOCE

IP3Rs are large tetrameric channels that open when they bind IP3 and Ca2+, but they also associate with many other proteins (Prole and Taylor, 2019), and many IP3Rs within cells appear not to release Ca2+ (Thillaiappan et al., 2019). A point mutation (D2550A, IP3R1D/A) within the IP3R1 pore prevents it from conducting Ca2+ (Dellis et al., 2008). As expected, expression of IP3R1D/A in cells lacking IP3R1 failed to rescue carbachol-evoked Ca2+ release, but it unexpectedly restored thapsigargin-evoked SOCE (Figure 3B-D; and Figure 3—figure supplement 1). We confirmed that rescue of thapsigargin-evoked Ca2+ entry by this pore-dead IP3R was mediated by a conventional SOCE pathway by demonstrating that it was substantially attenuated by siRNA-mediated knockdown of Orai1 (Figure 3C and D and Figure 3—figure supplement 1F–H).

Activation of IP3Rs is initiated by IP3 binding to the N-terminal IP3-binding core of each IP3R subunit (Prole and Taylor, 2019). Mutation of two conserved phosphate-coordinating residues in the α-domain of the binding core (R568Q and K569Q of IP3R1, IP3R1RQ/KQ) almost abolishes IP3 binding (Yoshikawa et al., 1996; Iwai et al., 2007), while mutation of a single residue (R568Q, IP3R1RQ) reduces the IP3 affinity by ~10-fold (Dellis et al., 2008). Expression of rat IP3R1RQ/KQ rescued neither carbachol-evoked Ca2+ release nor thapsigargin-evoked SOCE in cells lacking IP3R1 (Figure 3E and F and Figure 3—figure supplement 1C and I). However, expression of IP3R1RQ substantially rescued thapsigargin-evoked SOCE (Figure 3E and F and Figure 3—figure supplement 1J). Expression of an N-terminal fragment of rat IP3R (IP3R11-604), to which IP3 binds normally (Iwai et al., 2007), failed to rescue thapsigargin-evoked SOCE (Figure 3—figure supplement 1K and L). These results establish that a functional IP3-binding site within a full-length IP3R is required for IP3Rs to facilitate thapsigargin-evoked SOCE. Hence in cells with empty Ca2+ stores, IP3 binding, but not pore-opening, is required for regulation of SOCE by IP3Rs. In cells stimulated only with thapsigargin and expressing IP3Rs with deficient IP3 binding, basal levels of IP3 are probably insufficient to meet this need.

We further examined the need for IP3 by partially depleting the ER of Ca2+ using cyclopiazonic acid (CPA), a reversible inhibitor of SERCA, to allow submaximal activation of SOCE (Figure 3—figure supplement 1M and N). Under these conditions, addition of carbachol in Ca2+-free HBSS to SH-SY5Y cells expressing IP3R1-shRNA caused a small increase in [Ca2+]c (Figure 4A–C). In the same cells expressing IP3R1DA, the carbachol-evoked Ca2+ release was indistinguishable from that observed in cells without IP3RDA (Figure 4B and C), indicating that the small response was entirely mediated by residual native IP3R1 and/or IP3R3. Hence, the experiment allows carbachol to stimulate IP3 production in cells expressing IP3R1DA without causing additional Ca2+ release. The key result is that in cells expressing IP3R1DA, carbachol substantially increased SOCE from sub maximal to higher levels (Figure 4AC). Moreover, addition of carbachol to control shRNA expressing SH-SY5Y cells with maximal store depletion (thapsigargin, Tg, 2 µM) resulted in a small increase in SOCE (Figure 4—figure supplement 1A). We conclude that in neuronal cells IP3, through IP3Rs, regulates coupling of empty stores to SOCE. This is the first example of an IP3R mediating a response to IP3 that does not require the pore of the channel.

Figure 4 with 1 supplement see all
Download asset Open asset
Receptor-regulated IP3 production stimulates SOCE in cells with empty Ca2+ stores and expressing pore-dead IP3R.

(A, B) SH-SY5Y cells expressing IP3R1-shRNA alone (A) or with IP3R1DA (B) were treated with a low concentration of CPA (2 µM) in Ca2+-free HBSS to partially deplete the ER of Ca2+ and sub-maximally activate SOCE (see Figure 3—figure supplement 1M–N). Carbachol (CCh, 1 µM) was then added to stimulate IP3 formation through muscarinic receptors, and extracellular Ca2+ (2 mM) was then restored. Traces (mean ± s.e.m of 68–130 cells from three experiments) show responses with and without the CCh addition. (C) Summary results show the peak increases in [Ca2+]c (Δ[Ca2+]c) after addition of CCh (CCh-induced Ca2+ release) and then after restoring extracellular Ca2+ (SOCE). (D–F) SH-SY5Y cells wild type (WT) (D) and expressing NS-shRNA (E) or IP3R1-shRNA (F) were treated with YM-254890 (YM, 1 µM, 5 min) in Ca2+-free HBSS to inhibit Gαq and then with thapsigargin (Tg, 1 µM) before restoring extracellular Ca2+ (2 mM). Traces show mean ± s.e.m of ~120 cells from three experiments. (G–I) Similar analyses of HEK cells. Summary results (mean ± s.e.m, 50–100 cells from three experiments) are shown in (I). Different letter codes (panels C and I) indicate significantly different values within the store Ca2+ release or SOCE groups, p<0.001, one-way ANOVA and pair-wise Tukey’s test. See also Figure 4—figure supplement 1. Source data in Figure 4—source data 1.

Figure 4—source data 1

Receptor-regulated IP3 production stimulates SOCE in cells with empty Ca2+ stores and expressing pore-dead IP3R.

https://cdn.elifesciences.org/articles/80447/elife-80447-fig4-data1-v2.zip

G-protein-coupled receptors are linked to IP3 formation through the G-protein Gq, which stimulates phospholipase C β (PLC β). We used YM-254890 to inhibit Gq (Kostenis et al., 2020; Patt et al., 2021). As expected, addition of YM-254890 to wild type (WT) or NS-shRNA transfected SH-SY5Y cells abolished the Ca2+ signals evoked by carbachol (Figure 4—figure supplement 1C), but it also reduced the maximal amplitude and rate of thapsigargin-evoked SOCE (Figure 4D–E and Figure 3—figure supplement 1O). YM-254890 had no effect on the residual thapsigargin-evoked SOCE in SH-SY5Y cells expressing IP3R1-shRNA (Figure 4F and Figure 3—figure supplement 1O). The latter result is important because it demonstrates that the inhibition of SOCE in cells with functional IP3Rs is not an off-target effect causing a direct inhibition of SOCE.

In wild type or HEK-TKO (lacking all three IP3Rs) cells, YM-254890 had no effect on thapsigargin-evoked SOCE, but it did inhibit SOCE in HEK cells lacking only IP3R1 (Figure 4G–I and Figure 4—figure supplement 1D–G). These results suggest that in HEK cells, which normally express all three IP3R subtypes (Mataragka and Taylor, 2018), neither loss of IP3R1 nor inhibition of Gαq is sufficient on its own to inhibit thapsigargin-evoked SOCE, but when combined there is a synergistic loss of SOCE.

IP3Rs promote interaction of STIM1 with Orai1 within MCS

Our evidence that IP3Rs intercept coupling between empty stores and SOCE (Figure 3A) prompted us to investigate the coupling of STIM1 with Orai1 across the narrow junctions between ER and PM (Carrasco and Meyer, 2011). An in situ proximity ligation assay (PLA) is well suited to analyzing this interaction because it provides a signal when two immunolabeled proteins are within ~40 nm of each other (Derangère et al., 2016), a distance comparable to the dimensions of the junctions wherein STIM1 and Orai1 interact (Poteser et al., 2016). We confirmed the specificity of the PLA and demonstrated that it reports increased association of STIM1 with Orai1 after treating SH-SY5Y cells with thapsigargin by measuring the surface area of PLA spots (Figure 5A and Figure 5—figure supplement 1A–F) and not the number, because the latter did not change upon store-depletion (Figure 5—figure supplement 1O). In cells expressing IP3R1-shRNA, thapsigargin had no effect on the STIM1-Orai1 interaction reported by PLA, but the interaction was rescued by expression of IP3R1 or IP3R1DA. There was no rescue with IP3R1RQ/KQ (Figure 5B–E). WT SH-SY5Y cells that were depleted of basal IP3 by treatment with the Gq inhibitor YM-254890, showed significantly reduced STIM1-Orai1 interaction after thapsigargin-evoked depletion of Ca2+ stores (Figure 4—figure supplement 1B). The results with PLA exactly mirror those from functional analyses (Figures 1—4), suggesting that IP3 binding to IP3R enhances SOCE by facilitating interaction of STIM1 with Orai1 (Figure 3A).

Figure 5 with 1 supplement see all
Download asset Open asset
IP3Rs promote interaction of STIM1 with Orai1.

(A–E) PLA analyses of interactions between STIM1 and Orai1 in SH-SY5Y cells expressing NS-shRNA (A) or IP3R1-shRNA alone (B) or with IP3R1 (C), IP3R1DA (D) or IP3R1RQ/KQ (E). Confocal images are shown for control cells or after treatment with thapsigargin (Tg, 1 µM) in Ca2+-free HBSS. PLA reaction product is red, and nuclei are stained with DAPI (blue). Scale bars, 5 µm. Summary results show the surface area of the PLA spots for 8–10 cells from two independent analyses. Individual values, median (bar) and 25th and 75th percentiles (box). ***p < 0.001, Student’s t-test with unequal variances. See also Figure 5—figure supplement 1. Source data in Figure 5—source data 1.

Figure 5—source data 1

IP3Rs promote interaction of STIM1 with Orai1.

https://cdn.elifesciences.org/articles/80447/elife-80447-fig5-data1-v2.zip

In independent experiments we tested the effect of fluorescent-tagged and ectopically expressed ligand bound (wild type rat IP3R1) and mutant (rat IP3R1RQ/KQ; Figure 6—figure supplement 1A) IP3R1 on SOCE dependent STIM1 oligomerization and translocation to ER-PM junctions in SH-SY5Y cells (Figure 6). In agreement with PLA data (Figure 5), ER-PM translocation of mVenus-STIM1 upon SOCE induction was reduced significantly in mCherry-IP3R1RQ/KQ expressing cells compared to mCherry-IP3R1 expressing SH-SY5Y cells (Figure 6A, B, D and E and Figure 6—figure supplement 1B and D). SOCE also brought about a small increase in the surface intensity of over-expressed wild type mCherry-IP3R1 and mCherry-IP3R1RQ/KQ in the regions where we observe formation of SOCE-dependent STIM1 puncta (Figure 6A–C and Figure 6—figure supplement 1C and E). Moreover, the intensity of mCherry-IP3R1RQ/KQ appeared marginally lower than mCherry-IP3R1 (Figure 6—figure supplement 1C and E). The significance, if any, of these small changes in surface localization between over-expressed mCherry-IP3R1 and mCherry-IP3R1RQ/KQ upon SOCE induction, need further verification by alternate methods.

Figure 6 with 1 supplement see all
Download asset Open asset
Ligand-bound IP3R1 supports SOCE-dependent STIM1 movement to ER-PM contact sites.

(A–B) Representative TIRF images of mVenus STIM1 co-transfected with either wild type mcherry-rat IP3R1 (A) or IP3R1RQ/KQ (ligand binding mutant), (B) in wild type SH-SY5Y cells before (Basal) and after CPA induced store depletion (CPA treated) at 4 min and 7 min. On the right are shown RGB profile plots of STIM1 (green) and IP3R1, wild type or mutant (magenta) corresponding to the rectangular selections (Cell 1 and Cell 2). Scale bar is 10 µm.(C–D) Changes in number of IP3R1 (C) and STIM1 (D) puncta upon CPA-induced store depletion over a period of 10 min in the indicated genotypes. Mean ± s.e.m from seven cells from n=6 independent experiments. (E) Summary result (mean ± s.e.m) showing the change in the number of maximum STIM1 puncta formed after CPA-induced store depletion in the indicated genotypes. Mean ± s.e.m. of seven cells from n=6 independent experiments. Different letters indicate significant differences, p<0.05, Mann-Whitney U-test. See also Figure 6—figure supplement 1. Source data in Figure 6—source data 1.

Figure 6—source data 1

Ligand-bound IP3R1 supports SOCE-dependent STIM1 movement to ER-PM contact sites.

https://cdn.elifesciences.org/articles/80447/elife-80447-fig6-data1-v2.zip

Extended synaptotagmins (E-Syts) are ER proteins that stabilize ER-PM junctions including STIM1-Orai1 MCS (Maléth et al., 2014; Kang et al., 2019; Woo et al., 2020). Over-expression of E-Syt1 in SH-SY5Y cells expressing IP3R1-shRNA rescued thapsigargin-evoked Ca2+ entry without affecting resting [Ca2+]c or thapsigargin-evoked Ca2+ release (Figure 7A–C). The rescued Ca2+ entry is likely to be mediated by conventional SOCE because it was substantially attenuated by knockdown of STIM1 (Figure 7D–F). Over-expression of E-Syt1 had no effect on SOCE in cells with unperturbed IP3Rs (Figure 7G–I). These results suggest that attenuated SOCE after loss of IP3Rs can be restored by exaggerating ER-PM MCS.

Figure 7
Download asset Open asset
Extended synaptotagmins rescue SOCE in cells lacking IP3R1.

(A) SH-SY5Y cells expressing IP3R1-shRNA alone or with E-Syt1 were stimulated with Tg (1 µM) in Ca2+-free HBSS before restoring extracellular Ca2+ (2 mM). Traces show mean ± s.e.m, for 20–80 cells from three experiments. (B) Summary results show Δ[Ca2+]c evoked by restoring Ca2+ (SOCE). Mean ± s.e.m, ***p < 0.001, Mann-Whitney U- test. (C) Summary results (mean ± s.e.m, n=20–80 cells) show resting [Ca2+]c (left) and the peak Ca2+ signals (Δ[Ca2+]c) evoked by thapsigargin (Tg, 1 µM) in Ca2+-free HBSS for SH-SY5Y cells expressing IP3R1-shRNA alone or with human E-Syt1. (D) Cells over-expressing E-Syt1 and treated with IP3R1-shRNA in combination with either NS or STIM1 siRNA were stimulated with Tg (1 µM) in Ca2+-free HBSS before restoration of extracellular Ca2+ (2 mM). Mean ± s.e.m. from three experiments with 30–40 cells. (E, F) Summary results (mean ± s.e.m, n=30–40 cells) show SOCE evoked by Tg (E), resting [Ca2+]c and the Tg-evoked Ca2+ release from intracellular stores (F). ***p< 0.001, Mann-Whitney U- test. (G) Similar analyses of cells expressing NS shRNA alone or with human E-Syt1 and then treated with Tg (1 µM) in Ca2+-free HBSS before restoring extracellular Ca2+ (2 mM). Mean ± s.e.m. from three experiments with 115–135 cells. (H, I) Summary results (mean ± s.e.m, n=115–135 cells) show resting [Ca2+]c (H) and Δ[Ca2+]c evoked by Tg (store release) or Ca2+ restoration (SOCE) (I). No significant difference, Mann Whitney U-test. Source data in Figure 7—source data 1.

Figure 7—source data 1

Extended synaptotagmins rescue SOCE in cells lacking IP3R1.

https://cdn.elifesciences.org/articles/80447/elife-80447-fig7-data1-v2.zip

Discussion

After identification of STIM1 and Orai1 as core components of SOCE (Prakriya and Lewis, 2015; Thillaiappan et al., 2019), the sole role of IP3Rs within the SOCE pathway was assumed to be the release of ER Ca2+ that triggers STIM1 activation. The assumption is consistent with evidence that thapsigargin-evoked SOCE can occur in avian (Sugawara et al., 1997; Ma et al., 2002; Chakraborty et al., 2016) and mammalian cells without IP3Rs (Prakriya and Lewis, 2001). Although SOCE in mammalian HEK cells was unaffected by loss of IP3Rs in our study (Figure 2—figure supplement 3), it was modestly reduced in other studies of mammalian cells (Bartok et al., 2019; Yue et al., 2020). However, additional complexity is suggested by evidence that SOCE may be reduced in cells without IP3Rs (Chakraborty et al., 2016; Bartok et al., 2019; Yue et al., 2020), by observations implicating phospholipase C in SOCE regulation (Rosado et al., 2000; Broad et al., 2001), by evidence that SOCE responds differently to IP3Rs activated by different synthetic ligands (Parekh et al., 2002) and by some, albeit conflicting reports (Woodard et al., 2010; Santoso et al., 2011; Béliveau et al., 2014; Sampieri et al., 2018; Ahmad et al., 2022), that IP3Rs may interact with STIM and/or Orai (Woodard et al., 2010; Santoso et al., 2011; Béliveau et al., 2014; Sampieri et al., 2018).

We identified two roles for IP3Rs in controlling endogenous SOCE in human neurons. As widely reported, IP3Rs activate STIM1 by releasing Ca2+ from the ER, but they also, and independent of their ability to release Ca2+, enhance interactions between active STIM1 and Orai1 (Figure 8). The second role for IP3Rs can be supplanted by over-expressing other components of the SOCE complex, notably STIM1 or ESyt1 (Figure 2K–M and Figure 7A and B). It is intriguing that STIM1 (Carrasco and Meyer, 2011; Lewis, 2020), ESyt1 (Giordano et al., 2013) and perhaps IP3Rs (through the IP3-binding core) interact with phosphatidylinositol 4,5-bisphosphate (PIP2), which is dynamically associated with SOCE-MCS (Kang et al., 2019). We suggest that the extent to which IP3Rs tune SOCE in different cells is probably determined by the strength of Gq signaling, the proximity of IP3Rs to nanodomains of PLC signaling and endogenous interactions between STIM1 and Orai1. The latter is likely to depend on the relative expression of STIM1 and Orai1 (Woo et al., 2020), the STIM isoforms expressed, expression of proteins that stabilize STIM1-Orai1 interactions (Darbellay et al., 2011; Rana et al., 2015; Rosado et al., 2015; Knapp et al., 2022), and the size and number of the MCS where STIM1 and Orai1 interact (Kang et al., 2019). The multifarious contributors to SOCE suggest that cells may differ in whether they express “spare capacity”. In neuronal cells, loss of IP3 (Figure 4D) or of the dominant IP3R isoform (IP3R1-shRNA; Figures 1 and 2) is sufficient to unveil the contribution of IP3R to SOCE, whereas HEK cells require loss of both IP3 and IP3R1 to unveil the contribution (Figure 4H and I). The persistence of SOCE in cells devoid of IP3Rs (Figure 2—figure supplement 3D and E; Prakriya and Lewis, 2001; Ma et al., 2002) possibly arises from adaptive changes within the SOCE pathway. This does not detract from our conclusion that under physiological conditions, where receptors through IP3 initiate SOCE, IP3Rs actively regulate SOCE.

Figure 8
Download asset Open asset
Dual regulation of SOCE by IP3Rs.

(A) SOCE is activated when loss of Ca2+ from the ER, usually mediated by opening of IP3Rs when they bind IP3, causes STIM to unfurl cytosolic domains (2). The exposed cytosolic domains of STIM1 reach across a narrow gap between the ER and PM at a MCS to interact with PIP2 and Orai1 in the PM. Binding of STIM1 to Orai1 causes pore opening, and SOCE then occurs through the open Orai1 channel. We show that IP3Rs when they bind IP3 also facilitate interactions between Orai1 and STIM, perhaps by stabilizing the MCS (1). Receptors that stimulate IP3 formation thereby promote both activation of STIM (by emptying Ca2+ stores) and independently promote interaction of active STIM1 with Orai1. (B) Other mechanisms, including ryanodine receptors (RyR), can also release Ca2+ from the ER. We suggest that convergent regulation of SOCE by IP3R with bound IP3 allows receptors that stimulate IP3 formation to selectively control SOCE.

The IP3Rs that initiate Ca2+ signals reside in ER immediately beneath the PM and alongside, but not within, the MCS where STIM1 accumulates after store depletion (Thillaiappan et al., 2017; Figure 6A and B). In migrating cells too, IP3Rs and STIM1 remain separated as they redistribute to the leading edge (Okeke et al., 2016). Furthermore, there is evidence that neither STIM1 nor STIM2 co-immmunoprecipitate with IP3R1 (Ahmad et al., 2022). We suggest, and consistent with evidence that SOCE in cells without IP3Rs can be restored by over-expressing E-Syt1 (Figure 7A–C), that ligand-bound IP3Rs facilitate SOCE either by stabilizing the MCS wherein STIM1 and Orai1 interact, or by indirectly supporting STIM1 movement towards the MCS, rather than by directly regulating either protein. Stabilization of the MCS is analogous with similar structural roles for IP3Rs in maintaining MCS between ER and mitochondria (Bartok et al., 2019) or lysosomes (Atakpa et al., 2018; Figure 8). Alternately, our observation that SOCE-dependent STIM1 movement to the MCS is reduced in presence of IP3R1RQ/KQ (Figure 6 and Figure 6—figure supplement 1), suggests that ligand-bound IP3R1s could help in STIM1 mobilization to the MCS. The mechanism(s) by which ligand bound IP3R1s might stabilize the MCS or stimulate STIM1 movement to the MCS remain to be elucidated by methods that can directly assay the MCS such as electron microscopy.

Since both contributions of IP3Rs to SOCE require IP3 binding (Figure 3E and F), each is ultimately controlled by receptors that stimulate IP3 formation (Figure 4B and C). Convergent regulation by IP3Rs at two steps in the SOCE pathway may ensure that receptor-regulated PLC activity provides the most effective stimulus for SOCE; more effective, for example, than ryanodine receptors, which are also expressed in neurons (Figure 8B). By opening IP3Rs parked alongside SOCE MCS (Thillaiappan et al., 2017; Ahmad et al., 2022), IP3 selectively releases Ca2+ from ER that is optimally placed to stimulate SOCE, and by facilitating Orai1-STIM1 interactions IP3 reinforces this local activation of SOCE (Figure 8A and B).

We conclude that IP3-regulated IP3Rs regulate SOCE by mediating Ca2+ release from the ER, thereby activating STIM1 and/or STIM2 (Ahmad et al., 2022) and, independent of their ability to release Ca2+, IP3Rs facilitate the interactions between STIM and Orai that activate SOCE. Dual regulation of SOCE by IP3 and IP3Rs allows robust control by cell-surface receptors and may reinforce local stimulation of Ca2+ entry.

Materials and methods

Culture of human neural precursor cells

Request a detailed protocol

Human neural precursor cells (hNPCS) were derived from a human embryonic stem cell (hESC) line, H9/WA09 (RRID: CVCL_9773), using a protocol that inhibits dual SMAD signaling and stimulates Wnt signaling (Reinhardt et al., 2013) as described previously (Gopurappilly et al., 2018, 2019). hNPCs were grown as adherent dispersed cells on growth factor-reduced Matrigel (0.5%, Corning, Cat#356230) in hNPC maintenance medium (NMM) at 37 °C in humidified air with 5% CO2. NMM comprised a 1:1 mixture of Dulbecco’s Modified Eagle Medium with Nutrient Mixture F-12 (DMEM/F-12, Invitrogen, Cat#10565018) and Neurobasal medium (ThermoFisher, Cat#21103049), supplemented with GlutaMAX (0.5 x, Thermo Fisher, Cat#35050061), N2 (1:200, Thermo Fisher, 17502048), B27 without vitamin A (1:100, Thermo Fisher, Cat#12587010), Antibiotic-Antimycotic (Thermo Fisher, Cat#15240112), CHIR99021 (3 μM, STEMCELL Technologies, Cat#72052), purmorphamine (0.5 mM, STEMCELL Technologies, Cat#72202), and ascorbic acid (150 μM, Sigma, Cat#A92902). Doubling time was ~24 hr. Cells were passaged every 4–5 days by treatment with StemPro Accutase (Thermo Fisher, Cat#A1110501), stored in liquid nitrogen, and thawed as required. Cells were confirmed to be mycoplasma-free by monthly screening (MycoAlert, Lonza, Cat#LT07-318). hNPCs between passages 16 and 19 were used.

All experiments performed with hESC lines were approved by the Institutional Committee for Stem Cell Research, registered under the National Apex Committee for Stem Cell Research and Therapy, Indian Council of Medical Research, Ministry of Health, New Delhi.

Stable knockdown of IP3R1

Request a detailed protocol

An UltramiR lentiviral inducible shRNA-mir based on the shERWOOD algorithm (Auyeung et al., 2013; Knott et al., 2014) was used to inhibit IP3R1 expression. The all-in-one pZIP vector, which allows puromycin-selection and doxycycline-induced expression of both shRNA-mir and Zs-Green for visualization, was from TransOMIC Technologies (Huntsville, AL). Lentiviral pZIP transfer vectors encoding non-silencing shRNA (NS, NT#3-TTGGATGGGAAGTTCACCCCG) or IP3R1-targeting shRNA (ULTRA3316782- TTTCTTGATCACTTCCACCAG) were packaged as lentiviral particles using packaging (pCMV- dR8.2 dpvr, Addgene, plasmid #8455) and envelope vectors (pCMV-VSV-G, Addgene, plasmid #8454) by transfection of HEK293T cells (referred as HEK, ATCC, Cat# CRL-3216). Viral particles were collected and processed and hNPCs (passage 9) or SH-SY5Y cells were transduced (multiplicity of infection, MOI = 10) using Lipofectamine LTX with PLUS reagent (Thermo Fisher, Cat#15338100). Cells were maintained in media containing doxycycline (2 μg/ml, Sigma, Cat# D3072) to induce shRNA expression, and puromycin to select transduced cells (1 μg/ml for hNPCs; 3 μg/ml for SH-SY5Y cells; Sigma, Cat# P9620). Cells were passaged 4–5 times after lentiviral transduction to select for stable expression of shRNAs.

Derivation of neurons from hNPCs

Request a detailed protocol

Neurons were differentiated from hNPCs stably transduced with shRNA. hNPCs were seeded at 50–60% confluence in NMM on coverslips coated with poly-d-lysine (0.2 mg/ml, Sigma, Cat#P7280). After 1–2 days, the medium was replaced with neuronal differentiation medium, which comprised a 1:1 mixture of DMEM/F-12 with Neurobasal supplemented with B27 (1:100), N2 (1:200), GlutaMAX (0.5 x) and Antibiotic-Antimycotic solution. Medium was replaced on alternate days. Neurons were used after 15–20 days.

Culture and transfection of SH-SY5Y cells

Request a detailed protocol

SH-SY5Y cells (ATCC, USA, Cat# CRL-2266) were grown on culture dishes in DMEM/F-12 with 10% fetal bovine serum (Sigma, Cat# F4135) at 37°C in humidified air with 5% CO2. Cells were passaged every 3–4 days using TrypLE Express (ThermoFisher, Cat# 12605036) and confirmed to be free of mycoplasma. Cells expressing shRNA were transiently transfected using TransIT-LT1 reagent (Mirus, Cat# MIR-2300) in Opti-MEM (ThermoFisher, Cat# 31985062). Plasmids (250 ng) and/or siRNA (200 ng) in transfection reagent (1 µg/2.5 µl) were added to cells grown to 50% confluence on glass coverslips attached to an imaging dish. Cells were used after 48 hr. The siRNAs used were to human Orai1 (100 nM, Dharmacon, Cat# L-014998-00-0005) or non-silencing (NS, Dharmacon, Cat# D-495 001810-10-05), to human STIM1 (Santa Cruz Biotechnology, Cat# sc-76589) or NS (Santa Cruz Biotechnology, Cat# sc-37007). The expression plasmids were IP3R1 (rat type 1 IP3R1 in pcDNA3.2/V5DEST vector) (Dellis et al., 2008), rat IP3R1DA (D2550 replaced by A in pcDNA3.2 vector) (Dellis et al., 2008), rat IP3R1RQ (R568 replaced by Q of type 1 IP3R in pCDNA3.2/V5DEST vector) (Dellis et al., 2008), rat IP3R1RQ/KQ (R568 and K569 replaced by Q of type 1 IP3R in pCDNA3.2/V5DEST vector), rat IP3R11-604 (residues 1–604 of IP3R with N-terminal GST tag in pCDNA3.2/V5DEST vector; Dellis et al., 2008), rat IP3R3 (rat type 3 IP3R in pcDNA3.2/V5DEST vector; Saleem et al., 2013), human mCherry-STIM1 (N terminal mCherry tagged human STIM1 in pENTR1a vector; Nunes-Hasler et al., 2017) and human extended synaptotagmin 1 (E-Syt1), a kind gift from Dr S. Muallem, NIDCR, USA (Maléth et al., 2014).

CRISPR/Cas9 and Cas9n editing of SH-SY5Y cells

Request a detailed protocol

To allow either CRISPR/Cas9 or Cas9n-mediated disruption of IP3R1 expression, we used a published method to clone gRNAs into the backbone vector (pSpCas9n(BB)–2A-Puro PX462 V2.0, Addgene, Cat#62987; Ran et al., 2013). Forward and reverse sgRNA oligonucleotides (100 µM) were annealed and ligated using T4 DNA ligase by incubation (10 µl, 37 °C, 30 min) before slow cooling to 20 °C. Plasmids encoding Cas9n were digested with BbsI-HF (37 °C, 12 hr), gel-purified (NucleoSpin Gel and PCR Clean-up kit from Takara) and the purified fragment was stored at –20 °C. A mixture (final volume 20 µl) of gRNA duplex (1 µl, 0.5 µM), digested px459 (for IKO null) or pX462 vector (for IKO) (30 ng), 10× T4 DNA ligase buffer (2 µl) and T4 DNA ligase (1 µl) was incubated (20 °C, 1 hr). After transformation of DH5-α competent E. coli with the ligation mixture, plasmids encoding Cas9 or Cas9n and the sgRNAs were extracted, and the coding sequences were confirmed (Ran et al., 2013). The plasmid (2 µg) was then transfected into SH-SY5Y cells (50–60% confluent) in a six-well plate using TransIT LT-1 reagent (Mirus Bio, Cat# MIR-2300). After 48 hr, puromycin (3 µg/ml, 72 hr) was added to kill non-transfected cells. IKO colonies were propagated and screened for Ca2+ signals evoked by carbachol and for the presence of the IP3R gene by genomic DNA PCR and droplet digital PCR using primers close to the region targeted by the gRNAs (Miotke et al., 2014). Three independently derived IKO lines, each with one residual IP3R1 gene, were used for analyses of Ca2+ signaling (see Figure 2—figure supplement 1N-Q). For one of the cell lines (IKO 2), disruption of one copy of the IP3R1 gene was confirmed by genomic PCR, droplet digital PCR and western blotting (see Figure 2—figure supplement 1K-M). For the IKO null line, single-cell selection was done in a 96-well plate setup followed by screening for carbachol-evoked Ca2+ signals from multiple clones. A single clone was selected (Figure 2—figure supplement 1H) and a western blot performed to confirm absence of IP3R1 expression (Figure 2—figure supplement 1F). All the oligonucleotide sequences are described in Supplementary file 1.

Plasmid construction

Request a detailed protocol

Mutagenesis and all DNA modifications were carried out using Q5 Hot Start high-fidelity 2 X Master Mix (New England BioLabs, Cat# M0494L) using the recommendations of the manufacturer. Primers used in this study (details given in Supplementary file 1) were synthesized by Integrated DNA Technologies (IDT). Mutations in the Ligand binding domain (R568Q and K569Q) of IP3R1 were generated on the rat mCherry-IP3R1 cDNA in pDNA3.1 Mutations in all the constructs were confirmed by sequencing.

Ca2+ imaging

Request a detailed protocol

Methods for single-cell Ca2+ imaging were described previously (Gopurappilly et al., 2019). Briefly, cells grown as a monolayer (~70% confluence) on homemade coverslip-bottomed dishes were washed and loaded with Fura 2 by incubation with Fura 2 AM (4 μM, 45 min, 37 °C, Thermo Fisher, Cat# F1221), washed and imaged at room temperature in HEPES-buffered saline solution (HBSS). HBSS comprised: 20 mM HEPES, 137 mM NaCl, 5 mM KCl, 2 mM MgCl2, 2 mM CaCl2, 10 mM glucose, pH 7.3. CaCl2 was omitted from Ca2+-free HBSS. Treatments with carbachol (CCh, Sigma, Cat# C4382), thapsigargin (Tg, ThermoFisher, Cat# 7458), cyclopiazonic acid (CPA, Sigma Cat# C1530) or high-K+ HBSS (HBSS supplemented with 75 mM KCl) are described in legends.

Responses were recorded at 2 s intervals using an Olympus IX81-ZDC2 Focus Drift-Compensating Inverted Microscope with 60×oil immersion objective (numerical aperture, NA = 1.35) with excitation at 340 nm and 380 nm. Emitted light (505 nm) was collected with an Andor iXON 897E EMCCD camera and AndoriQ 2.4.2 imaging software (RRID: SCR_014461). Maximal (Rmax) and minimal (Rmin) fluorescence ratios were determined by addition of ionomycin (10 μM, Sigma, Cat# 407953) in HBSS containing 10 mM CaCl2 or by addition of Ca2+-free HBSS containing BAPTA (10 mM, Sigma, Cat# 196418) and Triton X100 (0.1%). Background-corrected fluorescence recorded from regions of interest (ROI) drawn to include an entire cell was used to determine mean fluorescence ratios (R = F340/F380) (ImageJ), and calibrated to [Ca2+]c from Grynkiewicz et al., 1985:

[Ca2+]c=KD.F380min/F380max.(R−Rmin)/(Rmax−R)

where, KD = 225 nM (Forostyak et al., 2013).

Western blots

Request a detailed protocol

Proteins were isolated in RIPA buffer (Sigma, Cat# R0278) with protease inhibitor cocktail (Sigma, Cat# P8340) or, for WB of Orai1, in medium containing 150 mM NaCl, 50 mM Tris, 1% Triton-X-100, 0.1% SDS and protease inhibitor cocktail. After 30 min on ice with intermittent shaking, samples were collected by centrifugation (11,000×g, 20 min) and their protein content was determined (Thermo Pierce BCA Protein Assay kit, ThermoFisher, Cat# 23225). Proteins (~30 µg/lane) were separated on 8% SDS-PAGE gels for IP3R or 10% SDS-PAGE gels for STIM1 and Orai1, and transferred to a Protran 0.45 μm nitrocellulose membrane (Merck, Cat# GE10600003) using a TransBlot semi-dry transfer system (BioRad, Cat# 1703940). Membranes were blocked by incubation (1 hr, 20 °C) in TBST containing skimmed milk or bovine serum albumin (5%, Sigma, Cat# A9418). TBST (Tris-buffered saline with Tween) comprised: 137 mM NaCl, 20 mM Tris, 0.1% Tween-20, pH 7.5. Membranes were incubated with primary antibody in TBST (16 hr, 4 °C), washed with TBST (3 ×10 min), incubated (1 hr, 20 °C) in TBST containing HRP-conjugated secondary antibody (1:3000 anti-mouse, Cell Signaling Technology Cat# 7076 S; or 1:5000 anti-rabbit, ThermoScientific Cat# 32260). After 3 washes, HRP was detected using Pierce ECL Western Blotting Substrate (ThermoFisher, Cat# 32106) and quantified using ImageQuant LAS 4000 (GE Healthcare) and Image J. The primary antibodies used were to: IP3R1 (1:1000, ThermoFisher, Cat# PA1-901, RRID: AB_2129984); β-actin (1:5000, BD Biosciences, Cat# 612656, RRID: AB_2289199); STIM1 (1:1000, Cell Signaling Technology, Cat# 5668 S, RRID: AB_10828699); Orai1 (1:500, ProSci, Cat# PM-5205, RRID: AB_10941192); IP3R2 (1:1000, custom made by Pocono Rabbit Farm and Laboratory; Mataragka and Taylor, 2018); and IP3R3 (1:500, BD Biosciences, Cat# 610313, RRID: AB_397705).

Immunocytochemistry

Request a detailed protocol

After appropriate treatments, cells on a coverslip-bottomed plate were washed twice with cold PBS, fixed in PBS with paraformaldehyde (4%, 20 °C, 20 min), washed (3×5 min) with PBS containing Triton-X100 (0.1%, PBST) and blocked by incubation (1 hr, 20 °C) in PBST containing goat serum (5%). After incubation with primary antibody in PBST (16 hr, 4 °C) and washing with PBST (3×5 min), cells were incubated (1 hr, 20 °C) with secondary antibody in PBST containing goat serum, washed (3×5 min), stained (10 min, 20 °C) with DAPI (1 µg/ml in PBS; Sigma, Cat# D9542) and washed (5 min, PBST). Cells were then covered with glycerol (60% v/v) and imaged using an Olympus FV300 confocal laser scanning microscope with 20×or 60×oil-immersion objectives. Fluorescence was analyzed using ImageJ. The primary antibodies used were to: PAX6 (1:500, Abcam, Cat# ab195045, RRID: AB_2750924); Nestin (1:500, Abcam, Cat# 92391, RRID: AB_10561437); Ki67 (1:250, Abcam, Cat# ab16667, RRID: AB_302459); SOX1 (1:1000, Abcam, Cat# ab87775, RRID: AB_2616563); Tuj1 (βIII Tubulin 1:1000, Promega, Cat# G712, RRID: AB_430874); NeuN (1:300, Abcam, Cat# ab177487, RRID: AB_2532109); Doublecortin (1:500, Abcam, Cat# 18723, RRID: AB_732011); MAP2 (1:200, Abcam, Cat# ab32454, RRID: AB_776174); STIM1 (1:1000, Cell Signaling Technology, Cat# 5668 S, RRID: AB_10828699); and Orai1 (1:500, ProSci, Cat# PM-5205, RRID: AB_10941192).

Proximity ligation Assay

Request a detailed protocol

The Duolink In Situ Red Starter Mouse/Rabbit kit was from Sigma (#Cat DUO92101) and used according to the manufacturer’s protocol with primary antibodies to Orai1 (mouse 1:500) and STIM1 (rabbit 1:1000). Cells (~30% confluent) were treated with thapsigargin (1 μM, 5 min) in Ca2+-free HBSS before fixation, permeabilization, and incubation with primary antibodies (16 hr, 4 °C) and the PLA reactants. Red fluorescent PLA signals were imaged using an Olympus FV300 confocal laser scanning microscope, with excitation at 561 nm, and a 60×oil-immersion objective. Quantitative analysis of the intensity and surface area of PLA spots used the ‘Analyze particle’ plugin of Fiji. Results are shown for 8–10 cells from two biological replicates of each genotype. Number of PLA spots in all genotypes and conditions were counted manually.

Detection of STIM1 and IP3R1 puncta using TIRF microscopy

Request a detailed protocol

SHSY5Y cells were cultured on 15 mm glass coverslips coated with poly-D-lysine (100 μg/ml) in a 35 mm dish for 24 h. Cells were co-transfected with 500 ng of mCherry rIP3R1 and 200 ng mVenus STIM1 plasmids using TransIT-LT1 transfection reagent in Opti-MEM. Following 48 hr of transfection and prior to imaging, cells were washed 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, at pH 7.4). The coverslips were mounted in a chamber and imaged using an Olympus IX81 inverted total internal reflection fluorescence microscope (TIRFM) equipped with oil-immersion PLAPO OTIRFM 60×objective lens/1.45 numerical aperture and Hamamatsu ORCA-Fusion CMOS camera. Olympus CellSens Dimensions 2.3 (Build 189987) software was used for imaging. The angle of the excitation beam was adjusted to achieve TIRF with a penetration depth of ∼130 nm. Images were captured from a final field of 65 µm × 65 µm (300×300 pixels, one pixel = 216 nm, binning 2×2). Cells positive for both mCherry rIP3R1 and mVenus STIM1 were identified using 561 nm and 488 nm lasers, respectively. The cells were incubated in zero calcium buffer (10 mM HEPES, 1 mM EGTA, 137 mM NaCl, 4.7 mM KCl, 5.5 mM glucose, 1 mM Na2HPO4, 0.56 mM MgCl2, at pH 7.4) for 2 min followed by addition of 30 µM CPA in zero calcium buffer. IP3R1 and STIM1 puncta prior to CPA addition and after CPA addition were captured at 1 min intervals. Raw images were filtered for background correction and same setting was used across all samples. Regions where fresh STIM1 puncta (2–10 pixels) appeared post-CPA treatment at 10 mins were marked and subsequently IP3R1 puncta (2–10 pixels) were captured from the same region. Change in the intensity of either STIM1 or IP3R1 puncta was calculated from puncta of >2 pixel by deducting the basal intensity at 0 min from the maximum intensity after CPA treatment using ImageJ ROI based mean grey value measurement. Particle analysis and RGB profile plot were done using ImageJ.

Statistical analyses

Request a detailed protocol

All experiments were performed without blinding or prior power analyses. Independent biological replicates are reported as the number of experiments (n), with the number of cells contributing to each experiment indicated in legends. The limited availability of materials for PLA restricted the number of independent replicates (n) to 2 (each with 8–10 cells). Most plots show means ± s.e.m. (or s.d.). Box plots show 25th and 75th percentiles, median and mean (see legends). Where parametric analyses were justified by a Normality test, we used Student’s t-test with unequal variances for two-way comparisons and ANOVA followed by pair-wise Tukey’s test for multiple comparisons. Non-parametric analyses used the Mann-Whitney U-test. Statistical significance is shown by ***p < 0.001, **p < 0.01, *p < 0.05, or by letter codes wherein different letters indicate significantly different values (p<0.001, details in legends). All analyses used Origin 8.5 software.

Details of the plasmids and recombinant DNAs are given in Supplementary file 1.

Resource availability

Lead contact

Request a detailed protocol

All requests for resources and reagents should be directed to the lead contact, Dr. Gaiti Hasan (gaiti@ncbs.res.in).

Materials availability

Request a detailed protocol

Constructs and cell lines are available upon request. MTA required for cell lines.

Data availability

This study did not generate any computer code. The data supporting the findings of this study are available within the manuscript. All other data supporting the findings of this study are available in source data file of respective figures.

References

    1. Grynkiewicz G
    2. Poenie M
    3. Tsien RY
    (1985)
    A new generation of Ca2+ indicators with greatly improved fluorescence properties
    The Journal of Biological Chemistry 260:3440–3450.
    1. Kostenis E
    2. Pfeil EM
    3. Annala S
    (2020)
    Heterotrimeric Gq proteins as therapeutic targets
    Journal of Biological Chemistry 295:5206–5215.
    1. Thillaiappan NB
    2. Chakraborty P
    3. Hasan G
    4. Taylor CW
    (2019) IP3 receptors and Ca2+ entry
    Biochimica et Biophysica Acta (BBA) - Molecular Cell Research 1866:1092–1100.
    https://doi.org/10.1016/j.bbamcr.2018.11.007

Decision letter

  1. Richard S Lewis
    Reviewing Editor; Stanford University School of Medicine, United States
  2. Kenton J Swartz
    Senior Editor; National Institute of Neurological Disorders and Stroke, National Institutes of Health, United States
  3. Khaled Machaca
    Reviewer; Weill Cornell Medicine Qatar, Qatar
  4. Nicolas Demaurex
    Reviewer; Department of Cell Physiology and Metabolism, University of Geneva, Switzerland

Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Regulation of Store-Operated ca2+ Entry by IP3 Receptors Independent of Their Ability to Release ca2+" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Richard Aldrich as the Senior Editor. The following individuals involved in the review of your submission have agreed to reveal their identity: Khaled Machaca (Reviewer #2); Nicolas Demaurex (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

– Are WT and IP3 binding deficient receptors recruited equivalently to membrane contact sites? This should be examined via high-resolution imaging.

– Are phosphoinositides involved in the recruitment of IP3R receptors?

– Does IP3R deficiency or mutant IP3R cause changes in the morphology/anatomy of the MCS? Is this independent of STIM1/Orai1? This should be addressed using high-resolution imaging and appropriate probes that do not perturb the contact sites.

– What is the dependence of IP3R occupancy with IP3 for SOCE activation? Does elevation of IP3 promote SOCE directly (independent of store-depletion)? This could be tested by examining what happens with CCh, the prediction is that this should further increase SOCE in response to TG (i.e. full depletion).

– Does STIM2 directly interact with IP3?

– What is the basis of the cell specificity of the IP3R effect in neuronal cells over HEK293 cells and immune cells where no effects on SOCE and CRAC currents were detected when all IP3Rs are deleted? This is a major point of difference from long-standing results and needs to be addressed.

Methodological issues:

– The conclusions regarding ER Ca levels and I3R activity are questionable. These should be improved using direct measures of ER [ca2+] using the now widely available ER targeted indicators.

– Determine if levels of STIM1 and Orai1 are altered in the IP3R1 KO cells.

– PLA studies should be done to show that YM and E-syt1 affect SOCE by modulating STIM-Orai interactions.

– PLA analysis is faulty and needs to be fixed.

– Orai1 Ab needs to be validated using an Orai1 KO HEK293 line.

In addition to the above points, the authors could consider addressing these additional questions as they would provide interesting additional insights in support of the paper's conclusions:

– Does IP3R deficiency alter (impair) phosphoinositide homeostasis at the PM?

– What is the phenotype of the full IP3R1 KO for MCS and Ca signaling?

– What about IP3R2 and 3 which are also present in neurons?

Reviewer #1 (Recommendations for the authors):

1. Based on the effects of YM, the authors suggest that only background levels of IP3 are required to maintain the normal amount of SOCE. Under these conditions, only a fraction of IP3R would be occupied. A prediction is that elevation of IP3 through a PLC-linked receptor should increase the occupancy of the IP3R and enhance SOCE in response to TG, even though TG by itself evokes full store depletion. This would strengthen support for the authors' model.

2. More direct evidence is needed to support the hypothesis that the IP3R increases STIM-Orai coupling and SOCE by stabilizing ER-PM junctions. MAPPER should be used to monitor the number and size of junctions after inhibitory treatments (IP3R1 KO, YM) or stimulatory treatments (CCh, IP3R1 restoration, STIM1 overexpression, E-Syt1). Because it links the ER and plasma membrane, a low expression of MAPPER may be required to avoid perturbing MCS formation. Perturbation would be indicated by rescue of SOCE by MAPPER in IP3R KO cells.

3. The cell specificity of the IP3R effects on SOCE is important. The STIM-Orai PLA experiments (or better yet, the MAPPER experiments in #2 above) should be done in HEK and SH-SY5Y cells in parallel. The prediction is that PLA and MCS will be reduced by IP3R KO in SH-SY5Y but not HEK cells. This would help establish a basis for explaining the cell-specific effects.

4. In Figure 4D, E, the effect of the Gq inhibitor on SOCE in NS shRNA cells is much less than in WT. What accounts for this difference?

5. As a control, the authors should confirm that STIM1 and Orai1 levels are not altered in the IP3R KO cells.

6. To show that YM and E-syt1 affect SOCE by modulating STIM-Orai interactions, PLA experiments should be done after YM inhibition of Gq, and after overexpression of E-syt1 as in Figure 5.

Reviewer #2 (Recommendations for the authors):

Were any controls performed to validate the specificity of the Orai1 antibody used to assess the expression levels of Orai1, as anti-Orai1 antibodies are notoriously non-specific? Has it been tested on Orai1-KO cells for example?

Line 198 should be Figure S6A.

Reviewer #3 (Recommendations for the authors):

1. IP3R is proposed to act as tethers at ER-PM junctions, yet no evidence is provided that WT and mutated receptors are recruited differentially to ER-PM junctions. To establish this point the authors should document the localization of the different receptors expressed (by immunofluorescence and PLA) to show that WT and mutant IP3R are differentially recruited to ER-PM contact sites and to show whether they localize near STIM and ORAI proteins.

2. The lack of an identified target for PM-bound IP3Rs is a significant limitation of the current study. While all the components of the STIM/ORAI machinery are potential targets the requirement for IP3 binding speaks for phosphoinositides. The authors should establish whether phosphoinositides are involved in the recruitment of IP3R receptors. Excellent tools are available to measure and manipulate the levels of phosphoinositides that can be used to document the effect of altered PIP2 levels on the localization of WT and mutated receptors.

3. The observation that the SOCE defect caused by IP3R deficiency can be rescued by the enforced expression of Esyt1 suggests that stabilizing membrane contact sites bypass the need for IP3R. This raises two questions: (1) Does IP3R deficiency indeed alters MCS structure or stability and (2) does IP3R deficiency alters MCS functions other than calcium signaling? ER-PM junctions are critical for the lipid replenishment of the PM thus the correction of the SOCE defect by ESyt1 could reflect changes in PM lipids rather than restored MCS tethering. The authors should provide morphological evidence that IP3R depletion and Esyt expression alter MCS and test whether the expression of lipid transport proteins (VAPs, NIRs, ORP) and of synthetic tethers such as MAPPERs can recapitulate the effect of ESyt1 on SOCE.

4. Why where the levels of the STIM2 protein not examined? STIM2 plays an important role in neurons and is thought to regulate and to be regulated by basal ER calcium levels. The lack of a ca2+ release channel is expected to increase the basal ca2+ levels within the ER, which in turn might prevent the activity of STIM2 and potentially reduce its expression levels. This control is particularly important since STIM2 was shown to interact with IP3R at membrane contact sites.

Methodological aspects:

5. As indicated in the public review I'm not convinced by the assays used to measure the filling state of intracellular calcium store. The authors report the amplitude of the calcium elevations evoked by thapsigargin in a ca2+-free medium to estimate the ca2+ content of intracellular stores. This parameter is considered to be equivalent in all the conditions presented despite some significant differences between conditions (e.g. Figure S5E, H, J). A better estimate would be provided by measuring the integrated response (area under the curve) preferably using non-calibrated traces as the non-linearity of the calibration augments variability. Inspection of the traces suggests that differences in ER ca2+ content are the norm rather than the exception because the averaged Tg responses differ visually between conditions in Figures 1H, 3B, C, E, S1D-E, S1K-L, S2K, S5K. The differentiated NPC shown in Figure 1H also appears to have a higher basal ca2+ which was not quantitatively evaluated. I would therefore ask the authors to re-analyse the data and to report the AUC of the Tg-evoked ca2+ responses. Given the importance of the ER ca2+ levels in controlling SOCE, the free ca2+ concentration within the ER should be determined for the critical conditions (control, IP3R KD, and reexpression of pore-dead and IP3-binding-deficient mutant). Excellent ratiometric cameleon indicators are available that allow quantitative recordings of [ca2+]ER. This is important because if the knockdown of IP3R impacts the resting ER ca2+ levels then the effects reported here might simply reflect changes in the activation levels of endogenous STIM proteins.

6. The PLA experiments are not analyzed properly. Reporting the total area of the dots is not common practice as these data are usually quantified by counting the number of dots per cell using Z stacks of 0.5µMm or less. The number of dots should be provided. The reference number for the STIM1 and ORAI1 antibodies used for PLA are not provided and their specificity should be validated by duolink using siRNA or KO for these proteins. These experiments should be repeated as an N of 2 biological replicates is too limited to support the conclusions.

7. Although CRISPR heterogeneous clones are validated by WB, I wonder why the authors did not present the sequences of change, or the impact. Also, I wonder why they decided to work with heterozygous and not with full KOs, which would give them a clean background. Is there any reason for choosing heterozygous over full KOs?

8. What about the other IP3R isoforms IP3R2 and 3? IP3R3 is predominant in neuronal systems but is not even discussed. Also, it is strange that its expression is not observed in the lines used in Figure 1. Is there any explanation for this? Do IP3R2 and 3 have a similar effect on MCS formation? Is the IP3 binding site conserved?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Regulation of Store-Operated ca2+ Entry by IP3 Receptors Independent of Their Ability to Release ca2+" for further consideration by eLife. Your revised article has been evaluated by Richard Aldrich (Senior Editor), a Reviewing Editor, and three referees.

The manuscript has been improved, but all three reviewers noted that many of the essential revisions in the first review that called for experiments had not been done and that the rebuttal arguments given in their stead did not resolve the issues. Because of the potential importance of the main findings, we will consider a second revision, but only if the essential experiments to support the main findings can be performed. These include:

1) Are WT and IP3-binding-deficient mutant IP3Rs recruited equivalently to the MCS? This could be done by expressing labeled WT or binding-deficient mutant IP3R in IKO null cells or cells pretreated with IP3R shRNA.

2) Does IP3R deletion or expression of IP3-binding-deficient mutant IP3R cause changes in the number or dimensions of MCS? The previous review noted that a low level of MAPPER must be used in order to avoid perturbing the system and that rescue of SOCE by MAPPER would indicate too high a level (as was seen in the rebuttal data).

3) Does additional IP3 (from CCh or caged IP3) increase SOCE in cells with fully depleted stores (e.g., treated with TG)? The experiment using partial depletion with CPA (Figure 4) does not answer this question. The strong prediction of the model is that even with fully depleted stores (TG), increasing IP3 should increase SOCE.

4) How can the cell specificity of the IP3R effect on SOCE in neuronal cells over HEK293 and immune cells be explained? The results with SH-SY5Y cells are a major departure from longstanding results in other cell types that need to be addressed. The new data from HEK cells lack internal consistency and do not support the model proposed by the authors.

5) The PLA analysis should include both the number and area of spots.

Please see the individual reviews below for more detail on these essential points.

Reviewer #1 (Recommendations for the authors):

– Are WT and IP3 binding deficient receptors recruited equivalently?

They did not assess this. The rebuttal argument that "there is no reason to believe that IP3R mutants change their localization" does not apply, because they attribute an SOCE-enhancing function to IP3-bound receptors that cannot be filled with the non-binding mutant. So the question is whether the binding of IP3 generates this function by regulating localization.

– Does IP3R deficiency or mutant IP3R cause changes in the morphology/anatomy of the MCS?

They did not look at the effect of IP3R KO or KD on ER-PM contacts, stating that MAPPER would not work because it restores junctions on its own. This is certainly true for moderate/high expression levels (I am guessing those are the conditions for what they show in the rebuttal), but my original comment suggested specifically to look at low levels that may not perturb the number of junctions. In fact, MAPPER has been used to monitor changes in ER-PM junctions following store depletion with TG (Chang et al., Cell Rep 2013), and it seems well suited since they show an increased number of STIM-Orai PLA signals following the expression of IP3R. It may be a bit tricky to use properly but would not require as much effort as EM.

– What is the dependence of IP3R occupancy with IP3 for SOCE activation?

Their model predicts that increasing IP3 should increase SOCE even in cells treated with TG to fully deplete stores. This experiment was suggested, but they did not do it. The previous data (referenced in the rebuttal) used a partially depleting dose of CPA, and they conclude that IP3R enhances SOCE without enhancing store depletion. But this is an indirect argument that does not cleanly address the question. This is an important experiment to do because it is extremely unlikely that IP3R are saturated with IP3 at resting levels, and they see a large effect even with this low level of occupancy. Presumably, it would be even larger with higher IP3 levels. This would make the conclusion much more convincing in my view.

– What is the basis of the cell specificity of the IP3R effect in neuronal cells over HEK293 cells and immune cells where no effects on SOCE and CRAC currents were detected when all IP3Rs are deleted?

The new data from HEK cells about differences between neuronal/non-neuronal cells raise new questions. They state "In wild type or HEK-TKO cells, YM-254890 had no effect on thapsigargin-evoked SOCE, but it did inhibit SOCE in HEK cells lacking IP3R1" (lines 218-219). This does not make sense, as TKO cells lack IP3R1. Also, why would the addition of YM (and presumably reduction of basal IP3) reduce SOCE in cells lacking IP3R1, if it acts through IP3R1? These data do not seem self-consistent and do not make sense to me.

Reviewer #2 (Recommendations for the authors):

Cannot see the STIM2 WB in the supplemental figures? Figure 2 S1 is cropped and missing panels including the STIM2 WB.

The data for the inhibition of Cch-induced ca2+ release in the presence of YM is not shown in Figure 3 S1 as indicated in the text.

Figure 4D shows a dramatic inhibition of Tg-induced SOCE in the presence of YM in SH cells. This is quite surprising and not addressed, especially since with the ns shRNA the inhibition is much smaller (Figure 4E). It actually looks like SOCE inhibition in WT cells is more dramatic than the residual SOCE remaining after Ip3R1 shRNA treatment. This is quite confusing.

The new HEK293 data with YM to inhibit Gq is also confusing. Figure 4 S1 the panels are mislabeled and the statistical significance in the last panel is not clear. In TKO cells YM doesn't have any effect on SOCE, yet partial loss of IP3R1 (shRNA) with a reduction in IP3 levels decreases SOCE. It is concluded that loss of both IP3R and IP3 production is required to lower SOCE, yet both are lost in TKO cells with no reduction. Is the reduction in SOCE in TKO cells in the absence of YM compared to WT significant? This is not clear from the data in Figure S1F. Please clarify. As it stands conclusions in HEK293 cells are not justified.

Figure 2 S1 the panels are confusing: panel F is not described in the legend and for other panels, there is a mismatch between the Figure legend and figure. Also the full IP3R KO mentioned in the response to reviewers is not clear.

For the PLA analyses, the authors use the area of PLA spots as a measure of STIM-Orai interactions. They should also consider the number of spots as those as well reflect STIM1-Orai1 interactions. A more reliable way to encompass both would be to quantify the percent of cell footprint occupied by spots as a total measure of STIM1-Orai1 interactions.

Please carefully review the supplemental figures labelling and legends as it is currently difficult to follow.

Reviewer #3 (Recommendations for the authors):

This has been a difficult rebuttal to handle because the authors did not respond directly to my queries to explain why they did not consider the experiments suggested. To avoid ambiguities, I would appreciate it if they could clarify the points below and provide a point-by-point response to the initial queries

Location of receptors: I fail to understand the argument that "there is no reason to believe that IPTR mutants changed their localization". The receptors are proposed to facilitate SOCE by stabilizing contact sites (lines 285-286). Altered localization of mutated receptors can thus logically explain the signaling defect and in the absence of experimental evidence, we cannot exclude this possibility. Visualization of IPTR recruitment to STIM/ORAI clusters could be performed without tagging endogenous receptors, by PLA, or by TIRF imaging of re-expressed tagged receptors.

Role of phosphoinositides: The question here was whether altering PiP levels differentially impacts the recruitment of WT and mutated receptors to MCS, but this was not tested. Instead, the authors provide evidence that CCh-induced PIP2 depletion is enhanced in neuronal cells lacking IPTR1. This phenotype is consistent with a tethering function of IPTR1 that would facilitate PIP2 replenishment by stabilizing ER-PM contact sites but does not provide information as to whether phosphoinositides are involved in IPTR1 recruitment. Again, tagged receptors could be used to measure the impact of PIP2 depletion on their recruitment to the TIRF plane. Why was this not attempted?

Morphology of MCS: The new data show that MAPPER expression restores SOCE in IPTR1 null cells, confirming that stabilizing contact sites with artificial tethers corrects the signaling defect. Further evidence for a tethering function would require EM which is beyond the scope of this study.

Role of STIM2: A WB showing STIM2 levels is mentioned in the rebuttal as Figure 2 supplement 1M but this Figure is not included in the pdf provided. The key point here is whether the STIM2 levels are similar in WT and KO cells as the low intensity of bands in WB could reflect the poor affinity of the antibody.

Methodological issues:

ER ca2+ levels: The authors did not perform the suggested ER [ca2+] recordings, arguing that "they have not drawn conclusions regarding changes in ER-Ca" and that "there is no reason to believe that loss of IPTR1 affects ER-Ca". Yet in the Ms the Tg-evoked ca2+ release, an indicator of the ER ca2+ content, is repeatedly said to be unaltered or minimally perturbed (lines 99, 128, 132, 138, 141, 199, 229). The problem here is that, as detailed in the initial review, the fura-2 recordings show substantial differences in the amount of ca2+ mobilized from stores by thapsigargin. Please address the queries in point #5 of the first review regarding the re-analysis of the fura-2 data as differences in ER [ca2+] would impact the conclusions of the study.

PLA experiments: please provide the number of dots for each condition.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Regulation of Store-Operated ca2+ Entry by IP3 Receptors Independent of Their Ability to Release ca2+" for further consideration by eLife. Your revised article has been evaluated by Kenton Swartz (Senior Editor) and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

While the revised manuscript is improved, several of the experiments that were requested by reviewers were not feasible or raised further questions, weakening the support for proposed mechanisms. However, the reviewers all agree that the main phenomenon described in the paper – a novel role for IP3R in modulating SOCE independent of their ability to release ca2+ – is exciting and worthy of publishing without further experiments to define an underlying mechanism. They also agree that given the uncertainty as to the mechanism, the authors will need to address the remaining points below clearly in the paper.

1) Are WT and IP3-binding-deficient mutant IP3Rs recruited equivalently to the MCS?

As Reviewer 2 noted, Figure 6A, B shows that the WT IP3R fluorescence in TIRF increases after CPA, but the RQ/KQ mutant does not. This does not appear to support the authors' conclusion that the two receptors are recruited equivalently to ER-PM junctions. The consensus view of the reviewers is that the time course of IP3R intensity in puncta after CPA should be plotted in Figure 6. This result has mechanistic implications, so it is important to show it clearly and discuss its interpretation in the paper.

2) Does IP3R deletion or expression of IP3-binding-deficient mutant IP3R cause changes in the number or dimensions of MCS?

The authors attempted to use MAPPER to monitor the number and size of MCS, but the lowest concentration of MAPPER DNA (200 ng) that produced detectable puncta also rescued SOCE in IKO null cells, suggesting that it altered the number of MCS by itself. It is surprising that 150 ng DNA did not label any MCS, while a slightly higher amount (200 ng) had a large enough effect on MCS stability to fully rescue the SOCE response. A more graded response would be expected, raising questions about whether MAPPER puncta at low transfection levels were overlooked. Nevertheless, difficulties in using MAPPER to track numbers and dimensions of native contact sites has been noted by other groups (including one of the reviewers). EM could be used to address this point, but is beyond the scope of the paper. Unfortunately, this means there is no direct evidence that IP3R increases the number of junctions, a key part of the hypothetical mechanism. For this reason, the reviewers agree that the authors should discuss the limitations of the available tools and propose alternative mechanisms (Discussion, around line 322). One such alternative would be that the IP3R directly interact with STIM1 rather than promoting junction formation. This might explain why the effects of YM in Figure 4 are so rapid (5 min), which may be too short a time for a profound loss of MCS.

3) Does additional IP3 (from CCh or caged IP3) increase SOCE in cells with fully depleted stores (e.g., treated with TG)?

(Figure 4 suppl 1A) Raising [IP3] with CCh after TG does increase the SOCE response but the effect is quite small, and as such does not offer strong support for the model. In fact, similar differences in peak Ca from SOCE were described as not significant in other experiments (e.g., Figure 2 supplement 3D). It would seem that if endogenous IP3 levels, which would be expected to only minimally occupy IP3R, are so potent at supporting SOCE (e.g., Figure 2E), raising IP3 significantly should cause a sizable increase in SOCE. It seems remarkable that resting [IP3], which is not enough to open the IP3R, can do so much, and even more so that the RQ mutant, with 10x lower affinity for IP3, rescues more than half the SOCE response (Figure 3E). It is difficult to imagine how this mutant would be binding any significant amount of IP3 in resting cells, unless it is sampling IP3 in a nanodomain close to PLC. Or perhaps as Reviewer 2 suggested, IP3Rs might be less important once STIM-Orai complexes are already formed. In any event, the small size of the CCh effect on SOCE needs to be acknowledged and discussed, as it appears to run counter to expectations given the hypothesis that IP3-bound receptors are needed for the effect.

4) How can the cell specificity of the IP3R effect on SOCE in neuronal cells over HEK293 and immune cells be explained?

The authors have provided a plausible explanation; however, there is no evidence that normal SOCE seen in TKO HEK cells "probably" arises from adaptive changes within the SOCE pathway (l. 310-311). "Possibly" would be more justified here (also considering that the response in TKO HEK cells appears somewhat reduced in Figure 2 Suppl 3D,E).

5) The PLA analysis should include both the number and area of spots.

The new data in Figure 5 suppl 1 show the number of PLA spots, but the quantification in panel O is confusing. The number of spots in the bar graph seems much lower than the number of spots visible in the PLA images of Figure 5 or the STIM/Orai puncta in Figure 5 suppl 1E-N. The authors should explain this apparent discrepancy (or choose more representative images to display).

Reviewer #2 (Recommendations for the authors):

For the new experiments shown in Figure 6 to address the recruitment of IP3R to MCS (ie the TIRF plane) the authors conclude that there is no difference in the recruitment of the WT vs RQ/KQ mutants. However, the intensity data for the two cells shows suggests otherwise: whereas there is a clear increase in the TIRF ROI for WT it is not apparent in the RQ/KQ mutant. Quantification of IP3R intensity in the TIRF plane on a per cell or ROI basis would quantitatively answer this question. It is clear that the number of IP3R puncta is not different between WT and the mutant (Figure 6C), but the intensity change is not clear given the way the data is presented. Need to show a time course of normalize intensity changes for WT and the mutant IP3R following CPA. This is important as it would argue for modulation of ER-PM MCS and as such provide a potential mechanism.

Figure 4—figure supplement 1A. Please elaborate on the finding of the relatively small increase with CCh compared to the significant decrease in SOCE following knockdown of IP3R (Figure 1D). This is an important finding as throughout the manuscript it is argued that ligand binding is critical for IP3R to support SOCE. Why the differential then with knockdown versus engaging the receptors after establishment of the STIM1-Orai1 complex? Are IP3Rs less important once the STIM1-Orai1 complexes are fully formed? Does TG treatment induce IP3 production?

Reviewer #3 (Recommendations for the authors):

Figure 6 shows TIRF images of mCherry-tagged receptors co-expressed with YFP-STIM1. The fluorescence pattern of the tagged receptors did not change appreciably during store depletion while additional STIM1 clusters appeared in cells expressing WT receptors. These data rely on overexpression with a substantial fraction of STIM1 pre-recruited in the TIRF plane that could explain the presence of the tagged IP3R in the TIRF plane, but nonetheless suggest that the WT and mutated receptors are not differentially recruited to contact sites. Whether changes in PPI impacts the distribution of receptors in the TIRF plane was not tested (using carbachol instead of CPA) but I agree that this experiment should be done with endogenously tagged receptors which would require considerable efforts. I thank the authors for addressing the points raised. The STIM2 levels are now documented and PLA data quantified. I have no further suggestions for changes.

https://doi.org/10.7554/eLife.80447.sa1

Author response

Essential revisions:

– Are WT and IP3 binding deficient receptors recruited equivalently to membrane contact sites? This should be examined via high-resolution imaging.

In this manuscript we have not attempted direct visualization of the IP3R at membrane contact sites. This is because previous publications from one of our groups (CWT) and others (Smith, Wiltgen and Parker, Cell Calcium 2009; Thillaiappan et al., Nat commun. 2017; Lock et al., Sci. Signal 2018) demonstrate that endogenous immobile clusters of IP3Rs that generate ca2+ puffs reside in ER-PM junctions alongside STIM whereas mobile IP3R clusters are present on the ER membrane. There is no reason to believe that IP3R1 mutants change their localization. Moreover, direct visualization of the IP3R1 mutants at MCS requires fluorescently tagged mutant and wild type IP3R1 by CRISPR editing that are currently not available in SH-SY5Y cells. Instead, we demonstrate that WT and IP3 binding mutant IP3R1s recruit SOCE molecules STIM1 and Orai1 differentially to the MCS (Figure 5). Possible roles for the IP3R1 and ligand binding during SOCE have been discussed in detail with relevant references to past work (Lines 291-311).

– Are phosphoinositides involved in the recruitment of IP3R receptors?

In order to attempt to answer this point we visualized an important PM phosphoinositide, PIP2, by expressing a PIP2 biosensor PH-PLCD1-GFP (Várnai and Balla, 2006 Biochim. Biophys. Acta – Mol. Cell Biol. Lipids) in SH-SY5Y cells lacking IP3R1 (Author response image 1A, IKO null) and HEK cells lacking all three IP3Rs (TKO) (Author response image 1B). In wild type SH-SY5Y cells a submaximal stimulus of carbachol (100µM) hydrolyzed plasma membrane (PM) bound PIP2 to approximately 60% of basal PM PIP2 whereas in IKO null cells PIP2 levels went down to approximately 35% of basal PM PIP2 after carbachol stimulation. Importantly, we did not observe a change in PM-localized PIP2 levels post-carbachol stimulation between HEK control and HEK-TKO cells. These data indicate higher PIP2 hydrolysis and/or reduced re-synthesis dynamics of membrane bound PIP2 in SH-SY5Y neuronal cells lacking IP3R1 but not in non-neuronal cells. We humbly submit that the relevance of altered PIP2 dynamics observed in IP3R1 knockout neuronal cells (Figure 1, below) to SOCE needs detailed future investigation involving genes that regulate PIP2 synthesis and hydrolysis. Our preliminary observation does not provide any new information in the context of our novel and important observation that ligand bound IP3Rs are the first step for initiating SOCE through STIM-Orai coupling at the ER-PM junction. Hence, we have not included these data in the manuscript.

Author response image 1
Download asset Open asset
Differential PIP2 dynamics in neuronal and non-neuronal cells.

(A-B) Confocal images of cells expressing a PIP2 biosensor PH-PLCD1-GFP. SH-SY5Y (WT and IKO null) (A) and HEK- (WT and TKO) cells (B) before (basal) and after carbachol treatment (CCh, 100μM). Scale bar is 10µm. Summary results (mean+ s.e.m, from 3 independent experiments with a total of 15-18 cells) show the relative intensity of plasma membrane (PM) bound PH-PLCD1-GFP normalized to total fluorescence in their respective WT cells. Different alphabet indicate P<0.01, Student’s t-test with unequal variances.

– Does IP3R deficiency or mutant IP3R cause changes in the morphology/anatomy of the MCS? Is this independent of STIM1/Orai1? This should be addressed using high-resolution imaging and appropriate probes that do not perturb the contact sites.

This is an important point that we are unfortunately unable to answer due to the lack of appropriate probes that do not also affect ER-PM contact sites. Established methods by expression of MAPPER or Esyt1 will not work because both molecules restore ER-PM contact sites and rescue SOCE in IP3R1 shRNA or KO (IKO null) cells (Figure 6 for Esyt1 in IP3R1 shRNA cells) and see Author response image 2 for MAPPER in IP3R KO (IKO null) cells. The relationship of IP3Rs to MCS morphology has been discussed in detail (Lines 274-318).

Author response image 2
Download asset Open asset

– What is the dependence of IP3R occupancy with IP3 for SOCE activation? Does elevation of IP3 promote SOCE directly (independent of store-depletion)? This could be tested by examining what happens with CCh, the prediction is that this should further increase SOCE in response to TG (i.e. full depletion).

We examined the need for IP3 by partially depleting the ER of ca2+ using cyclopiazonic acid (CPA), a reversible inhibitor of SERCA, to allow submaximal activation of SOCE (Figure 3 —figure supplement 1M and 1N). Under these conditions, addition of carbachol in ca2+-free HBSS to cells expressing IP3R1-shRNA caused a small increase in [ca2+]c (Figures 4A-4C). In the same cells expressing IP3R1DA, the carbachol-evoked ca2+ release was indistinguishable from that observed in cells without IP3RDA (Figures 4B and 4C), indicating that the small response was entirely mediated by residual native IP3R1 and/or IP3R3. Hence, the experiment allows carbachol to stimulate IP3 production in cells expressing IP3R1DA without causing additional ca2+ release. The key result is that in cells expressing IP3R1DA, carbachol substantially increased SOCE (Figures 4A-4C). We conclude that IP3, through IP3Rs, regulates coupling of empty stores to SOCE.

– Does STIM2 directly interact with IP3?

Western Blot analysis showed relatively low expression of STIM2 compared to STIM1 in SH-SY5Y cells (Figure 2 —figure supplement 1M) unlike non-neuronal cells (Ahmad et al., PNAS 2022). Therefore we did not investigate STIM2 function and interactions in SH-SY5Y cells any further.

– What is the basis of the cell specificity of the IP3R effect in neuronal cells over HEK293 cells and immune cells where no effects on SOCE and CRAC currents were detected when all IP3Rs are deleted? This is a major point of difference from long-standing results and needs to be addressed.

We provide new data showing that in HEK cells, neither loss of IP3R1 nor inhibition of Gq/11 uncouples empty ca2+ stores form SOCE, but together they do uncouple (Figure 4). In agreement with these data, we now show that STIM-Orai interactions (visualized by PLA) are similar in HEK WT and HEK-TKO cells (Figure 2 —figure supplement 1F). The new results contribute to the expanded Discussion (p13-14) in which we suggest that multifarious regulation of SOCE is likely to provide different cells with different levels of ‘surplus capacity’ and so different susceptibilities to disabling single elements. Hence, in neurons, loss of IP3R is sufficient to inhibit SOCE, while HEK cells require loss of both IP3R and IP3. Our new results suggest that regulation of SOCE by IP3R is likely to be a widespread feature of mammalian cells.

Methodological issues:

– The conclusions regarding ER Ca levels and I3R activity are questionable. These should be improved using direct measures of ER [ca2+] using the now widely available ER targeted indicators.

ca2+ measurements in this manuscript are all related to changes in cytosolic ca2+ either in response to store depletion (Thapsigargin) or an IP3 generating ligand, Carbachol. We have not drawn any conclusions regarding changes in ER-ca2+. To the best of our knowledge there is no reason to believe that loss of IP3R1 (the predominantly expressed IP3R subtype in many mammalian neurons) affects ER-ca2+. There are small changes in ER-ca2+ when ALL three subtypes of IP3Rs are knocked out in either HEK cells or mouse fibroblasts. Our experiments have not addressed triple IP3R knock outs in neuronal cells. The study focuses on knockdown and knockout of IP3R1 alone.

– Determine if levels of STIM1 and Orai1 are altered in the IP3R1 KO cells.

As suggested we have performed the Westerns Blots (Figure S2M). The data show that levels of STIM1, STIM2 and Orai1 are not altered in IP3R1 KO (IKO null) cells.

– PLA studies should be done to show that YM and E-syt1 affect SOCE by modulating STIM-Orai interactions.

We have performed PLA analyses for YM treated wild type SH-SY5Y cells (Figure 4 —figure supplement 1A). They show reduced interaction between STIM1 and Orai1 after Tg-induced store depletion similar to our observation with IP3R1 KD SH-SY5Y cells (Figure 5). The ability of E-Syt1 to restore SOCE by enhancing STIM-Orai interactions by creating more ER-PM junctions is published (Chang et al., Cell Rep, 2013; Giordano et al., Cell 2013; Kang et al., Sci Rep. 2019).

– PLA analysis is faulty and needs to be fixed.

Upon store depletion by thapsigargin the size of PLA spots grew significantly bigger due to recruitment of multiple STIM1 molecules to the ER-PM junctions and their interaction with more Orai1 molecules similar to what has been observed in non-excitable cells (Shen et al., PNAS 2021). The thapsigargin-mediated response was robust and significant (Figure 5). Hence, we have analyzed the surface area of each PLA spot to represent greater STIM1-Orai1 interactions at the MCS (Figure 5). In our studies we did not observe a significant change between the number of smaller PLA spots in resting cells vs the larger spots in thapsigargin-treated cells.

– Orai1 Ab needs to be validated using an Orai1 KO HEK293 line.

We validated the Orai1 Ab using siOrai1 expressing SH-SY5Y cells (Figure 3 —figure supplement 1F).

In addition to the above points, the authors could consider addressing these additional questions as they would provide interesting additional insights in support of the paper's conclusions.

In addition to the above points, the authors could consider addressing these additional questions as they would provide interesting additional insights in support of the paper's conclusions:

– Does IP3R deficiency alter (impair) phosphoinositide homeostasis at the PM?

Yes, we have tested this and the results are described above (Figure 1 in this letter). While the results are indeed of interest they do not add to the conclusions of this manuscript. Hence, we have not included them here.

– What is the phenotype of the full IP3R1 KO for MCS and Ca signaling?

Full IP3R1 KO lines were made in SH-SY5Y cells and their phenotype is now included in Figure 2 —figure supplement 1. Their SOCE phenotype is similar to the single-copy IP3R1 knockout. However, the IP3R1 KO null cells were extremely fragile and grow very slowly. This is not surprising since IP3R1 is the predominant (~99%) IP3R isoform in SH-SY5Y cells. Hence, we have not used the complete knock out cells extensively in this study.

– What about IP3R2 and 3 which are also present in neurons?

IP3R3 overexpression rescues the SOCE phenotype of IP3R1 knockdown cells (Figure 2H). As shown in Figure 2A IP3R2 is not expressed in SH-SY5Y cells.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been improved, but all three reviewers noted that many of the essential revisions in the first review that called for experiments had not been done and that the rebuttal arguments given in their stead did not resolve the issues. Because of the potential importance of the main findings, we will consider a second revision, but only if the essential experiments to support the main findings can be performed. These include:

1) Are WT and IP3-binding-deficient mutant IP3Rs recruited equivalently to the MCS? This could be done by expressing labeled WT or binding-deficient mutant IP3R in IKO null cells or cells pretreated with IP3R shRNA.

As suggested we tested the recruitment of over-expressed mcherry-wild type IP3R1 and mcherry-IP3 binding deficient mutant (IP3R1RQ/KQ) to ER-PM junctions by TIRF microscopy in SH-SY5Y cells. Consistent with our PLA findings from Figure 5, we show that SOCE-dependent STIM1 translocation to the TIRF layer is significantly reduced in SH-SY5Y cells transfected with IP3R1RQ/KQ as compared with cells transfected with WT IP3R1. In the same experiment we tested localisation of WT-IP3R1 and IP3R1RQ/KQ to the TIRF layer upon store-depletion and SOCE. We see no change in EITHER IP3R1WT or IP3R1RQ/KQ localisations monitored for 10 minutes after store depletion (Figure 6), possibly due to overexpression of the constructs. Hence the second experiment suggested was not attempted. The current data do NOT support differential recruitment of WT and LBD IP3R1 to the MCS. Rather, as shown in Figure 8, they suggest that IP3 bound IP3Rs help create the MCS through interaction with as yet unidentified molecule(s) (see discussion lines 323-335).

2) Does IP3R deletion or expression of IP3-binding-deficient mutant IP3R cause changes in the number or dimensions of MCS? The previous review noted that a low level of MAPPER must be used in order to avoid perturbing the system and that rescue of SOCE by MAPPER would indicate too high a level (as was seen in the rebuttal data).

We have tried this experiment by transfecting 50ng, 150ng and 200ng of MAPPER. In SH-SY5Ycells we do not see MCS formation either with 50ng MCS or with 150ng of MAPPER (please see panel A in Author response image 3 for 150 ng). At 200ng we do see MCS but we also see rescue of SOCE. The concentrations used by Chang et al., Cell Rep 2013 of 15-50 ng MAPPER in HeLa cells appear not to work in SH-SY5Y cells. This is possibly because the ER-PM architecture of neuronal cells is different from non-neuronal cells like HeLa (PMID: 14493991; PMID: 30739879; PMID: 28559323).

Author response image 3
Download asset Open asset

3) Does additional IP3 (from CCh or caged IP3) increase SOCE in cells with fully depleted stores (e.g., treated with TG)? The experiment using partial depletion with CPA (Figure 4) does not answer this question. The strong prediction of the model is that even with fully depleted stores (TG), increasing IP3 should increase SOCE.

We have done the SOCE experiment in control SH-SY5Y cells by fully depleting stores using 2 µM thapsigargin (Tg) followed by 1µM carbachol (Cch) and we could see a small but significant potentiation of SOCE by addition of carbachol (Figure 4 —figure supplement 1A).

4) How can the cell specificity of the IP3R effect on SOCE in neuronal cells over HEK293 and immune cells be explained? The results with SH-SY5Y cells are a major departure from longstanding results in other cell types that need to be addressed. The new data from HEK cells lack internal consistency and do not support the model proposed by the authors.

We suggest that the extent to which IP3Rs tune SOCE in different cell types is determined by their “spare capacity” for SOCE. This is likely dependent on strength of Gq signaling, expression level and sub-cellular localisation of IP3R isoforms and interactions between STIM and Orai, based on cell specific regulators of SOCE (see discussion lines 300-315). In neuronal cells, loss of either IP3 (Figure 4D) or of the dominant IP3R isoform (IP3R1-shRNA; Figures 1 and 2) is sufficient to unveil the contribution of IP3R to SOCE, whereas HEK cells requires reduction in BOTH IP3 (+YM condition) and IP3R1 (IP3R1 shRNA) to unveil the contribution of ligand bound IP3Rs to SOCE (Figures 4H and 4I; Figure 4 - figure supplement 1E-1G ). The persistence of SOCE in HEK cells devoid of all three IP3Rs (HEK TKO; Figure 2 —figure supplement 3D and 3E) (Prakriya and Lewis, 2001; Ma et al., 2002) probably arises from adaptive changes within the SOCE pathway in the prolonged absence of all three IP3R subtypes. This does not detract from our conclusion that under physiological conditions, where receptors through IP3 initiate SOCE, IP3Rs actively regulate SOCE.

5) The PLA analysis should include both the number and area of spots.

We have included the number of PLA spots in Figure 5—figure supplement 1O. As discussed earlier, there is no difference in the number of PLA spots under +/-Tg treated condition. However, IP3R1 shRNA and IP3R1 shRNA+IP3R1RQ/KQ cells had significantly few PLA spots compared to control shRNA cells.

Reviewer #1 (Recommendations for the authors):

– Are WT and IP3 binding deficient receptors recruited equivalently?

They did not assess this. The rebuttal argument that "there is no reason to believe that IP3R mutants change their localization" does not apply, because they attribute an SOCE-enhancing function to IP3-bound receptors that cannot be filled with the non-binding mutant. So the question is whether the binding of IP3 generates this function by regulating localization.

Response given in the essential review above. Please see Figure 6 in the revised manuscript.

– Does IP3R deficiency or mutant IP3R cause changes in the morphology/anatomy of the MCS?

They did not look at the effect of IP3R KO or KD on ER-PM contacts, stating that MAPPER would not work because it restores junctions on its own. This is certainly true for moderate/high expression levels (I am guessing those are the conditions for what they show in the rebuttal), but my original comment suggested specifically to look at low levels that may not perturb the number of junctions. In fact, MAPPER has been used to monitor changes in ER-PM junctions following store depletion with TG (Chang et al., Cell Rep 2013), and it seems well suited since they show an increased number of STIM-Orai PLA signals following the expression of IP3R. It may be a bit tricky to use properly but would not require as much effort as EM.

A detailed response is given above in the essential review.

– What is the dependence of IP3R occupancy with IP3 for SOCE activation?

Their model predicts that increasing IP3 should increase SOCE even in cells treated with TG to fully deplete stores. This experiment was suggested, but they did not do it. The previous data (referenced in the rebuttal) used a partially depleting dose of CPA, and they conclude that IP3R enhances SOCE without enhancing store depletion. But this is an indirect argument that does not cleanly address the question. This is an important experiment to do because it is extremely unlikely that IP3R are saturated with IP3 at resting levels, and they see a large effect even with this low level of occupancy. Presumably, it would be even larger with higher IP3 levels. This would make the conclusion much more convincing in my view.

Response given in the essential review above. Please see Figure 4 —figure supplement 1A.

– What is the basis of the cell specificity of the IP3R effect in neuronal cells over HEK293 cells and immune cells where no effects on SOCE and CRAC currents were detected when all IP3Rs are deleted?

The new data from HEK cells about differences between neuronal/non-neuronal cells raise new questions. They state "In wild type or HEK-TKO cells, YM-254890 had no effect on thapsigargin-evoked SOCE, but it did inhibit SOCE in HEK cells lacking IP3R1" (lines 218-219). This does not make sense, as TKO cells lack IP3R1. Also, why would the addition of YM (and presumably reduction of basal IP3) reduce SOCE in cells lacking IP3R1, if it acts through IP3R1? These data do not seem self-consistent and do not make sense to me.

Reduced SOCE upon addition of YM is ONLY seen in HEK cells with IP3R1-shRNA (Figure 4H) and NOT in either HEK-WT (Figure 4 – supplement 1E) OR HEK TKO cells (Figure 4 – supplement 1F). We apologise for this confusion. Note that efficiency of knockdown with IP3R1-shRNA is not 100% (Figure 2A). These data are consistent with the idea that addition of YM reduces IP3 formation and thus further reduces ligand bound IP3R1 in IP3R1-shRNA HEK cells.

Reviewer #2 (Recommendations for the authors):

Cannot see the STIM2 WB in the supplemental figures? Figure 2 S1 is cropped and missing panels including the STIM2 WB.

The data for the inhibition of Cch-induced ca2+ release in the presence of YM is not shown in Figure 3 S1 as indicated in the text.

Figure 4D shows a dramatic inhibition of Tg-induced SOCE in the presence of YM in SH cells. This is quite surprising and not addressed, especially since with the ns shRNA the inhibition is much smaller (Figure 4E). It actually looks like SOCE inhibition in WT cells is more dramatic than the residual SOCE remaining after Ip3R1 shRNA treatment. This is quite confusing.

The new HEK293 data with YM to inhibit Gq is also confusing. Figure 4 S1 the panels are mislabeled and the statistical significance in the last panel is not clear. In TKO cells YM doesn't have any effect on SOCE, yet partial loss of IP3R1 (shRNA) with a reduction in IP3 levels decreases SOCE. It is concluded that loss of both IP3R and IP3 production is required to lower SOCE, yet both are lost in TKO cells with no reduction. Is the reduction in SOCE in TKO cells in the absence of YM compared to WT significant? This is not clear from the data in Figure S1F. Please clarify. As it stands conclusions in HEK293 cells are not justified.

Figure 2 S1 the panels are confusing: panel F is not described in the legend and for other panels, there is a mismatch between the Figure legend and figure. Also the full IP3R KO mentioned in the response to reviewers is not clear.

For the PLA analyses, the authors use the area of PLA spots as a measure of STIM-Orai interactions. They should also consider the number of spots as those as well reflect STIM1-Orai1 interactions. A more reliable way to encompass both would be to quantify the percent of cell footprint occupied by spots as a total measure of STIM1-Orai1 interactions.

Please carefully review the supplemental figures labelling and legends as it is currently difficult to follow.

We apologise for these errors. The corrections had been made and revised supplementary figures have been uploaded.

Reviewer #3 (Recommendations for the authors):

This has been a difficult rebuttal to handle because the authors did not respond directly to my queries to explain why they did not consider the experiments suggested. To avoid ambiguities, I would appreciate it if they could clarify the points below and provide a point-by-point response to the initial queries

Location of receptors: I fail to understand the argument that "there is no reason to believe that IPTR mutants changed their localization". The receptors are proposed to facilitate SOCE by stabilizing contact sites (lines 285-286). Altered localization of mutated receptors can thus logically explain the signaling defect and in the absence of experimental evidence, we cannot exclude this possibility. Visualization of IPTR recruitment to STIM/ORAI clusters could be performed without tagging endogenous receptors, by PLA, or by TIRF imaging of re-expressed tagged receptors.

The experiment with tagged receptors is now shown in Figure 6. An explanation for the results are given in the essential review above.

Role of phosphoinositides: The question here was whether altering PiP levels differentially impacts the recruitment of WT and mutated receptors to MCS, but this was not tested. Instead, the authors provide evidence that CCh-induced PIP2 depletion is enhanced in neuronal cells lacking IPTR1. This phenotype is consistent with a tethering function of IPTR1 that would facilitate PIP2 replenishment by stabilizing ER-PM contact sites but does not provide information as to whether phosphoinositides are involved in IPTR1 recruitment. Again, tagged receptors could be used to measure the impact of PIP2 depletion on their recruitment to the TIRF plane. Why was this not attempted?

As evident in Figure 6 of the revised manuscript tagged receptors are present in the TIRF plane of resting SH-SY5Y cells BUT SOCE does not induce further recruitment of overexpressed tagged receptors to the TIRF plane. Probably, this experiment would needs to be done with endogenously tagged WT and mutant receptors. We have not attempted this as yet.

Morphology of MCS: The new data show that MAPPER expression restores SOCE in IPTR1 null cells, confirming that stabilizing contact sites with artificial tethers corrects the signaling defect. Further evidence for a tethering function would require EM which is beyond the scope of this study.

Role of STIM2: A WB showing STIM2 levels is mentioned in the rebuttal as Figure 2 supplement 1M but this Figure is not included in the pdf provided. The key point here is whether the STIM2 levels are similar in WT and KO cells as the low intensity of bands in WB could reflect the poor affinity of the antibody.

STIM2 levels are not changed between WT and IKO cells. Please see the updated Supplementary Figure 2 (Figure 2- supplementary 1G).

Methodological issues:

ER ca2+ levels: The authors did not perform the suggested ER [ca2+] recordings, arguing that "they have not drawn conclusions regarding changes in ER-Ca" and that "there is no reason to believe that loss of IPTR1 affects ER-Ca". Yet in the Ms the Tg-evoked ca2+ release, an indicator of the ER ca2+ content, is repeatedly said to be unaltered or minimally perturbed (lines 99, 128, 132, 138, 141, 199, 229). The problem here is that, as detailed in the initial review, the fura-2 recordings show substantial differences in the amount of ca2+ mobilized from stores by thapsigargin. Please address the queries in point #5 of the first review regarding the re-analysis of the fura-2 data as differences in ER [ca2+] would impact the conclusions of the study.

Changes in ER-ca2+ release vary among individual experiments to an extent. To control for this variation we have ALWAYS performed the control cell type (e.g NS shRNA) with the experimental condition in parallel in every experiment.

PLA experiments: please provide the number of dots for each condition.

This analysis has been added in Figure 5—figure supplement 1O.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

While the revised manuscript is improved, several of the experiments that were requested by reviewers were not feasible or raised further questions, weakening the support for proposed mechanisms. However, the reviewers all agree that the main phenomenon described in the paper – a novel role for IP3R in modulating SOCE independent of their ability to release ca2+ – is exciting and worthy of publishing without further experiments to define an underlying mechanism. They also agree that given the uncertainty as to the mechanism, the authors will need to address the remaining points below clearly in the paper.

1) Are WT and IP3-binding-deficient mutant IP3Rs recruited equivalently to the MCS?

As Reviewer 2 noted, Figure 6A, B shows that the WT IP3R fluorescence in TIRF increases after CPA, but the RQ/KQ mutant does not. This does not appear to support the authors' conclusion that the two receptors are recruited equivalently to ER-PM junctions. The consensus view of the reviewers is that the time course of IP3R intensity in puncta after CPA should be plotted in Figure 6. This result has mechanistic implications, so it is important to show it clearly and discuss its interpretation in the paper.

As suggested we measured the intensity of tagged and overexpressed WT IP3R1, IP3R1RQ/KQ mutants as well as tagged STIM1 before and after CPA treatment within the ROIs containing visible STIM1 puncta. These data are in Figure 6 – supplement 1B-E, described in results on lines 256-264 and discussed in lines 326-336. There is a small increase in intensity of WT IP3Rs after CPA treatment that is not matched by an increase in puncta number. We do not see this change in RQ/KQ. Because the intensity change for IP3Rs is small, we are concerned that it may be an artefact of over-expression. We have stated that this result needs to be verified using alternate methods (lines 262-264).

2) Does IP3R deletion or expression of IP3-binding-deficient mutant IP3R cause changes in the number or dimensions of MCS?

The authors attempted to use MAPPER to monitor the number and size of MCS, but the lowest concentration of MAPPER DNA (200 ng) that produced detectable puncta also rescued SOCE in IKO null cells, suggesting that it altered the number of MCS by itself. It is surprising that 150 ng DNA did not label any MCS, while a slightly higher amount (200 ng) had a large enough effect on MCS stability to fully rescue the SOCE response. A more graded response would be expected, raising questions about whether MAPPER puncta at low transfection levels were overlooked. Nevertheless, difficulties in using MAPPER to track numbers and dimensions of native contact sites has been noted by other groups (including one of the reviewers). EM could be used to address this point, but is beyond the scope of the paper. Unfortunately, this means there is no direct evidence that IP3R increases the number of junctions, a key part of the hypothetical mechanism. For this reason, the reviewers agree that the authors should discuss the limitations of the available tools and propose alternative mechanisms (Discussion, around line 322). One such alternative would be that the IP3R directly interact with STIM1 rather than promoting junction formation. This might explain why the effects of YM in Figure 4 are so rapid (5 min), which may be too short a time for a profound loss of MCS.

We agree and have made the necessary changes in the discussion (lines 331-336).

3) Does additional IP3 (from CCh or caged IP3) increase SOCE in cells with fully depleted stores (e.g., treated with TG)?

(Figure 4 suppl 1A) Raising [IP3] with CCh after TG does increase the SOCE response but the effect is quite small, and as such does not offer strong support for the model. In fact, similar differences in peak Ca from SOCE were described as not significant in other experiments (e.g., Figure 2 supplement 3D). It would seem that if endogenous IP3 levels, which would be expected to only minimally occupy IP3R, are so potent at supporting SOCE (e.g., Figure 2E), raising IP3 significantly should cause a sizable increase in SOCE. It seems remarkable that resting [IP3], which is not enough to open the IP3R, can do so much, and even more so that the RQ mutant, with 10x lower affinity for IP3, rescues more than half the SOCE response (Figure 3E). It is difficult to imagine how this mutant would be binding any significant amount of IP3 in resting cells, unless it is sampling IP3 in a nanodomain close to PLC. Or perhaps as Reviewer 2 suggested, IP3Rs might be less important once STIM-Orai complexes are already formed. In any event, the small size of the CCh effect on SOCE needs to be acknowledged and discussed, as it appears to run counter to expectations given the hypothesis that IP3-bound receptors are needed for the effect.

The key point here is that neuronal cells have a certain SOCE capacity and to reach this they require ligand-bound IP3Rs. Our data do not support an infinite increase in SOCE in presence of higher and higher levels of IP3. In Figure 4 suppl 1A, the small change in SOCE by Cch addition is observed after maximal store depletion and in the presence of WT IP3R1. Under these conditions SOCE is already close to its maximal capacity. The excess IP3 generated in this condition thus results in a minimal change in SOCE. We have modified the text in lines 204 and 206 to reflect this better. As shown in Author response image 4 partial store depletion by a lower concentration of Tg results in reduced SOCE which can be further increased by addition of Cch. We did not include these data in the manuscript because reviewer 2 wanted us to try the experiment with complete store depletion. In Figure 3E SOCE is induced in the absence of WT IP3R1 and is thus very low. The RQ mutant is able to bring SOCE back to ~50% of its maximal value. We agree that it might well do this by the mechanism suggested by the reviewer (see lines 305-306).

Author response image 4
Download asset Open asset

4) How can the cell specificity of the IP3R effect on SOCE in neuronal cells over HEK293 and immune cells be explained?

The authors have provided a plausible explanation; however, there is no evidence that normal SOCE seen in TKO HEK cells "probably" arises from adaptive changes within the SOCE pathway (l. 310-311). "Possibly" would be more justified here (also considering that the response in TKO HEK cells appears somewhat reduced in Figure 2 Suppl 3D,E).

We agree and have changed “probably” to “possibly”.

5) The PLA analysis should include both the number and area of spots.

The new data in Figure 5 suppl 1 show the number of PLA spots, but the quantification in panel O is confusing. The number of spots in the bar graph seems much lower than the number of spots visible in the PLA images of Figure 5 or the STIM/Orai puncta in Figure 5 suppl 1E-N. The authors should explain this apparent discrepancy (or choose more representative images to display).

We have redone the quantification of PLA spots by counting them manually in Figure 5 supplement 1. We have also replaced images of PLA in Figure 5 with more representative images. In the previous version the PLA spots were counted through an automated mechanism which appeared to take two closely lying spots as one.

Reviewer #2 (Recommendations for the authors):

For the new experiments shown in Figure 6 to address the recruitment of IP3R to MCS (ie the TIRF plane) the authors conclude that there is no difference in the recruitment of the WT vs RQ/KQ mutants. However, the intensity data for the two cells shows suggests otherwise: whereas there is a clear increase in the TIRF ROI for WT it is not apparent in the RQ/KQ mutant. Quantification of IP3R intensity in the TIRF plane on a per cell or ROI basis would quantitatively answer this question. It is clear that the number of IP3R puncta is not different between WT and the mutant (Figure 6C), but the intensity change is not clear given the way the data is presented. Need to show a time course of normalize intensity changes for WT and the mutant IP3R following CPA. This is important as it would argue for modulation of ER-PM MCS and as such provide a potential mechanism.

Please see our response to point 1 above.

Figure 4—figure supplement 1A. Please elaborate on the finding of the relatively small increase with CCh compared to the significant decrease in SOCE following knockdown of IP3R (Figure 1D). This is an important finding as throughout the manuscript it is argued that ligand binding is critical for IP3R to support SOCE. Why the differential then with knockdown versus engaging the receptors after establishment of the STIM1-Orai1 complex? Are IP3Rs less important once the STIM1-Orai1 complexes are fully formed? Does TG treatment induce IP3 production?

Please see our response to point 3 above. There is no reason to believe that Tg induces IP3 production.

https://doi.org/10.7554/eLife.80447.sa2

Article and author information

Author details

  1. Pragnya Chakraborty

    1. National Centre for Biological Sciences, Tata Institute of Fundamental Research, Bangalore, India
    2. SASTRA University, Thanjavur, India
    Contribution
    Conceptualization, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing – original draft, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5916-5534

  2. Bipan Kumar Deb

    National Centre for Biological Sciences, Tata Institute of Fundamental Research, Bangalore, India
    Present address
    Department of Molecular and Cell Biology, University of California, Berkeley, United States
    Contribution
    Formal analysis, Validation, Methodology
    Competing interests
    No competing interests declared

  3. Vikas Arige

    Department of Pharmacology and Physiology, University of Rochester, Rochester, United States
    Contribution
    Methodology
    Competing interests
    No competing interests declared

  4. Thasneem Musthafa

    National Centre for Biological Sciences, Tata Institute of Fundamental Research, Bangalore, India
    Contribution
    Methodology
    Competing interests
    No competing interests declared

  5. Sundeep Malik

    Department of Pharmacology and Physiology, University of Rochester, Rochester, United States
    Contribution
    Methodology
    Competing interests
    No competing interests declared

  6. David I Yule

    Department of Pharmacology and Physiology, University of Rochester, Rochester, United States
    Contribution
    Conceptualization, Funding acquisition
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-6743-0668

  7. Colin W Taylor

    Department of Pharmacology, University of Cambridge, Cambridge, United Kingdom
    Contribution
    Conceptualization, Resources, Supervision, Funding acquisition, Writing – original draft, Writing – review and editing
    For correspondence
    cwt1000@cam.ac.uk
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-7771-1044

  8. Gaiti Hasan

    National Centre for Biological Sciences, Tata Institute of Fundamental Research, Bangalore, India
    Contribution
    Conceptualization, Resources, Supervision, Funding acquisition, Writing – original draft, Writing – review and editing
    For correspondence
    gaiti@ncbs.res.in
    Competing interests
    Reviewing editor, eLife
    Additional information
    Lead Contact
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-7194-383X

Funding

Department of Science and Technology, Ministry of Science and Technology, India (DST/INSPIRE Fellowship/2017/IF170360)

  • Pragnya Chakraborty

Department of Biotechnology, Ministry of Science and Technology, India (BT/PR6371/COE/34/19/2013)

  • Gaiti Hasan

Tata Institute of Fundamental Research (NCBS)

  • Gaiti Hasan

Wellcome Trust (101844)

  • Colin W Taylor

Biotechnology and Biological Sciences Research Council (BB/T012986/1)

  • Colin W Taylor

National Institutes of Health (DE014756)

  • David I Yule

Tata Institute of Fundamental Research (TIFR core support)

  • Gaiti Hasan

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. For the purpose of Open Access, the authors have applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission.

Acknowledgements

This research was supported by grants to GH from the Dept. of Biotechnology, Govt. of India (BT/PR6371/COE/34/19/2013) and NCBS-TIFR core support, to CWT from the Wellcome Trust (101844) and Biotechnology and Biological Sciences Research Council (BB/T012986/1) and to DIY from the NIH (NIDCR, DE014756). PC is supported by a DST-INSPIRE fellowship (DST/INSPIRE Fellowship/2017/IF170360) and she received an Infosys-NCBS travel award to visit CWT’s lab at Cambridge. We are grateful to Renjitha Gopurappilly (NCBS, TIFR) for the derivation of human neural precursor cells. We acknowledge use of the Central Imaging and Flow Cytometry Facility (CIFF), Stem Cell Culture Facility and Biosafety level-2 laboratory facility at NCBS, TIFR.

Senior Editor

  1. Kenton J Swartz, National Institute of Neurological Disorders and Stroke, National Institutes of Health, United States

Reviewing Editor

  1. Richard S Lewis, Stanford University School of Medicine, United States

Reviewers

  1. Khaled Machaca, Weill Cornell Medicine Qatar, Qatar
  2. Nicolas Demaurex, Department of Cell Physiology and Metabolism, University of Geneva, Switzerland

Version history

  1. Preprint posted: April 13, 2022 (view preprint)
  2. Received: May 20, 2022
  3. Accepted: July 18, 2023
  4. Accepted Manuscript published: July 19, 2023 (version 1)
  5. Version of Record published: August 7, 2023 (version 2)

Copyright

© 2023, Chakraborty et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 739
    Page views
  • 178
    Downloads
  • 1
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Pragnya Chakraborty
  2. Bipan Kumar Deb
  3. Vikas Arige
  4. Thasneem Musthafa
  5. Sundeep Malik
  6. David I Yule
  7. Colin W Taylor
  8. Gaiti Hasan
(2023)
Regulation of store-operated Ca2+ entry by IP3 receptors independent of their ability to release Ca2+
eLife 12:e80447.
https://doi.org/10.7554/eLife.80447

Categories and tags

Research organism

Further reading

    1. Cell Biology

    PM2.5 leads to adverse pregnancy outcomes by inducing trophoblast oxidative stress and mitochondrial apoptosis via KLF9/CYP1A1 transcriptional axis

    Shuxian Li, Lingbing Li ... Xietong Wang
    Research Article

    Epidemiological studies have demonstrated that fine particulate matter (PM2.5) is associated with adverse obstetric and postnatal metabolic health outcomes, but the mechanism remains unclear. This study aimed to investigate the toxicological pathways by which PM2.5 damaged placental trophoblasts in vivo and in vitro. We confirmed that PM2.5 induced adverse gestational outcomes such as increased fetal mortality rates, decreased fetal number and weight, damaged placental structure, and increased apoptosis of trophoblasts. Additionally, PM2.5 induced dysfunction of the trophoblast cell line HTR8/SVneo, including in its proliferation, apoptosis, invasion, migration and angiogenesis. Moreover, we comprehensively analyzed the transcriptional landscape of HTR8/SVneo cells exposed to PM2.5 through RNA-Seq and observed that PM2.5 triggered overexpression of pathways involved in oxidative stress and mitochondrial apoptosis to damage HTR8/SVneo cell biological functions through CYP1A1. Mechanistically, PM2.5 stimulated KLF9, a transcription factor identified as binding to CYP1A1 promoter region, which further modulated the CYP1A1-driven downstream phenotypes. Together, this study demonstrated that the KLF9/CYP1A1 axis played a crucial role in the toxic progression of PM2.5 induced adverse pregnancy outcomes, suggesting adverse effects of environmental pollution on pregnant females and putative targeted therapeutic strategies.

    1. Cell Biology
    2. Microbiology and Infectious Disease

    PerTurboID, A targeted in situ method reveals the impact of kinase deletion on its local protein environment in the cytoadhesion complex of malaria causing parasites

    Heledd Davies, Hugo Belda ... Moritz Treeck
    Tools and Resources

    Reverse genetics is key to understanding protein function, but the mechanistic connection between a gene of interest and the observed phenotype is not always clear. Here we describe the use of proximity labeling using TurboID and site-specific quantification of biotinylated peptides to measure changes to the local protein environment of selected targets upon perturbation. We apply this technique, which we call PerTurboID, to understand how the P. falciparum exported kinase, FIKK4.1, regulates the function of the major virulence factor of the malaria causing parasite, PfEMP1. We generated independent TurboID fusions of 2 proteins that are predicted substrates of FIKK4.1 in a FIKK4.1 conditional KO parasite line. Comparing the abundance of site-specific biotinylated peptides between wildtype and kinase deletion lines reveals the differential accessibility of proteins to biotinylation, indicating changes to localization, protein-protein interactions, or protein structure which are mediated by FIKK4.1 activity. We further show that FIKK4.1 is likely the only FIKK kinase that controls surface levels of PfEMP1, but not other surface antigens, on the infected red blood cell under standard culture conditions. We believe PerTurboID is broadly applicable to study the impact of genetic or environmental perturbation on a selected cellular niche.

    1. Cell Biology

    CRB3 navigates Rab11 trafficking vesicles to promote γTuRC assembly during ciliogenesis

    Bo Wang, Zheyong Liang ... Peijun Liu
    Research Article

    The primary cilium plays important roles in regulating cell differentiation, signal transduction, and tissue organization. Dysfunction of the primary cilium can lead to ciliopathies and cancer. The formation and organization of the primary cilium are highly associated with cell polarity proteins, such as the apical polarity protein CRB3. However, the molecular mechanisms by which CRB3 regulates ciliogenesis and the location of CRB3 remain unknown. Here, we show that CRB3, as a navigator, regulates vesicle trafficking in γ-tubulin ring complex (γTuRC) assembly during ciliogenesis and cilium-related Hh and Wnt signaling pathways in tumorigenesis. Crb3 knockout mice display severe defects of the primary cilium in the mammary ductal lumen and renal tubule, while mammary epithelial-specific Crb3 knockout mice exhibit the promotion of ductal epithelial hyperplasia and tumorigenesis. CRB3 is essential for lumen formation and ciliary assembly in the mammary epithelium. We demonstrate that CRB3 localizes to the basal body and that CRB3 trafficking is mediated by Rab11-positive endosomes. Significantly, CRB3 interacts with Rab11 to navigate GCP6/Rab11 trafficking vesicles to CEP290, resulting in intact γTuRC assembly. In addition, CRB3-depleted cells are unresponsive to the activation of the Hh signaling pathway, while CRB3 regulates the Wnt signaling pathway. Therefore, our studies reveal the molecular mechanisms by which CRB3 recognizes Rab11-positive endosomes to facilitate ciliogenesis and regulates cilium-related signaling pathways in tumorigenesis.