IP₃R-driven increases in mitochondrial Ca²⁺ promote neuronal death in NPC disease

Significance

NPC1 is a ubiquitously expressed lysosomal cholesterol transporter whose loss of function results in neurodegenerative NPC1 disease. Here, we report that loss-of-function, knockout, or mutation-causing NPC1 initiates a damaging signaling cascade that alters the expression and nanoscale distribution of IP₃R type 1 that precipitates neuron death. Targeting IP₃R1 or upstream elements of this signaling cascade rescues neuronal death and provides potential therapeutic targets to address IP₃R dysfunction, a feature of NPC1 disease and other neurodegenerative disorders.

Abstract

Ca²⁺ is the most ubiquitous second messenger in neurons whose spatial and temporal elevations are tightly controlled to initiate and orchestrate diverse intracellular signaling cascades. Numerous neuropathologies result from mutations or alterations in Ca²⁺ handling proteins; thus, elucidating molecular pathways that shape Ca²⁺ signaling is imperative. Here, we report that loss-of-function, knockout, or neurodegenerative disease–causing mutations in the lysosomal cholesterol transporter, Niemann-Pick Type C1 (NPC1), initiate a damaging signaling cascade that alters the expression and nanoscale distribution of IP₃R type 1 (IP₃R1) in endoplasmic reticulum membranes. These alterations detrimentally increase G_q-protein coupled receptor–stimulated Ca²⁺ release and spontaneous IP₃R1 Ca²⁺ activity, leading to mitochondrial Ca²⁺ cytotoxicity. Mechanistically, we find that SREBP-dependent increases in Presenilin 1 (PS1) underlie functional and expressional changes in IP₃R1. Accordingly, expression of PS1 mutants recapitulate, while PS1 knockout abrogates Ca²⁺ phenotypes. These data present a signaling axis that links the NPC1 lysosomal cholesterol transporter to the damaging redistribution and activity of IP₃R1 that precipitates cell death in NPC1 disease and suggests that NPC1 is a nanostructural disease.

Calcium (Ca²⁺) is the most-ubiquitous second messenger in neurons. For this reason, spatial and temporal elevations in cytosolic Ca²⁺ signals are tightly controlled, with elevations in cytoplasmic Ca²⁺ used to initiate and orchestrate a host of diverse intracellular signaling cascades fundamentally required for neuronal development, synaptic plasticity, neurotransmission, and neuron fidelity (for review, refer to refs. 1, 2). One of the primary mechanisms to elevate intracellular Ca²⁺ levels is through release of Ca²⁺ from the largest intracellular Ca²⁺ store, the endoplasmic reticulum (ER). Ca²⁺ can be rapidly mobilized from the ER into the cytoplasm through the Inositol 1,4,5-triphosphate receptor (IP₃R), a ligand-gated Ca²⁺ channel in the ER membrane. IP₃Rs are activated when cell surface G_q-protein coupled receptors (G_qPCR) stimulate phospholipase C (PLC) to hydrolyze the plasma membrane (PM) lipid PI(4,5)P₂ into soluble IP₃, the native agonist of the IP₃Rs (3). IP₃Rs are further regulated by phosphorylation, protein–protein interactions, and local Ca²⁺ concentrations (3, 4).

Niemann-Pick Type C1 (NPC1) disease is an autosomal recessive lysosomal storage disorder that arises from mutations in the gene encoding the lysosomal NPC1 protein. This transmembrane cholesterol transporter facilitates the egress of cholesterol from the acidic compartment to other cellular membranes with its loss-of-function resulting in the cellular hallmark of NPC1 disease: an accumulation of lipids within the lysosome and perturbations in cellular cholesterol homeostasis. The underlying mechanisms linking cholesterol dysregulation and lipid accumulation to neurodegeneration are not well understood; however, impaired Ca²⁺ signaling is a consequence of NPC1 disease mutations (5–7). Given that dysfunctional Ca²⁺ signaling is a common feature in neurodegenerative disease and is thought to precede and contribute to neuronal cell death (2, 8), understanding the link between NPC1 function and intracellular Ca²⁺ becomes imperative.

There are three IP₃R isoforms: IP₃R Type 1 (IP₃R1), Type 2, and Type 3, each with its own differential expression across cell types and tissues. IP₃R1 protein is most abundantly expressed in soma and synaptic regions of the brain, including Purkinje cells of the cerebellum (9), the most impacted cell type in NPC1 disease (10). Functionally, IP₃R1 is essential for a variety of intracellular events including G_qPCR signaling (11), synaptic strengthening (12), Ca²⁺-dependent gene transcription (11), and Ca²⁺ transfer at ER–mitochondrial membrane contact sites (MCS) (13). Behaviorally, IP₃R1^−/− mice die in utero or only live for a few weeks while exhibiting severe neurological symptoms such as ataxia and seizures (14), while specific cerebellum loss (15) or disease mutations in the IP₃R1 gene lead to ataxias (16, 17) and Gillespie syndrome (18). Given IP₃R1’s importance in neurological function and development, and our previous reports of decreased PI(4,5)P₂ levels (precursor of IP₃) as well as reductions in ER Ca²⁺ (source of G_qPCR Ca²⁺) in multiple models of NPC disease (19, 20), we hypothesized that IP₃R signaling may be perturbed in NPC1 disease.

Using a combination of murine and patient disease models of NPC1 disease, we show that despite reduced PI(4,5)P₂ (19) and ER Ca²⁺ (20), IP₃-mediated Ca²⁺ release is paradoxically increased. We determine that loss of NPC1 function results in larger and more numerous immobile IP₃Rs clusters near the PM that drive spontaneous Ca²⁺ release events (Ca²⁺ puffs) and prime IP₃R to release more Ca²⁺ during G_qPCR activation. The cellular consequences of enhanced IP₃R clustering and activity are 1) decreased ER Ca²⁺ levels, 2) increased nuclear factor of activated T-cells (NFAT) signaling, and 3) increased mitochondrial Ca²⁺, which leads to neuronal Ca²⁺ cytotoxicity. We propose that the lysosomal NPC1 cholesterol transporter can tune the distribution and activity of IP₃R in health, while its loss of function, such as in NPC1 disease, initiates a pervasive signaling cascade that triggers mitochondrial cytotoxicity.

Results

Loss of NPC1 Function Potentiates G_q-Mediated Ca²⁺ Signaling.

We have previously reported that both ER-Ca²⁺ stores and PM PI(4,5)P₂ levels are significantly reduced in NPC1 disease (19, 20). Given that IP₃R-mediated Ca²⁺ release requires both a steep Ca²⁺ gradient across the ER and IP₃’s precursor, PI(4,5)P_2; (Fig. 1A), we tested the hypothesis that cytosolic Ca²⁺ signals evoked from IP₃Rs following activation of a G_qPCR were perturbed in NPC1 loss-of-function cells. First, we compared G_qPCR responses from control and U18666A-treated (U18; specific inhibitor of NPC1) (21) tsA201 cells loaded with the cytosolic Ca²⁺ indicator Fluo-4 and treated with histamine, an agonist for the G_q-coupled histamine H₁ receptor, in a Ca²⁺-free solution to eliminate extracellular Ca²⁺ contribution. Analysis of the resulting Ca²⁺ responses revealed that overnight inhibition of NPC1 significantly increased the amount of Ca²⁺ released into the cytoplasm (Fig. 1 B and C). Similar results were observed from NPC1^−/− cells (Fig. 1 D and E) and patient fibroblasts harboring the most prevalent patient mutation (NPC1^I1061T; Fig. 1 F and G). To confirm that NPC1-dependent increases in G_qPCR-mediated Ca²⁺ release were conserved in neurons, we performed similar experiments in two murine models of NPC1 disease: 1) NPC1 knockout (NPC1^−/−) neurons and 2) Npc1^{tm(I1061T)Dso} (hereafter referred to as NPC1^I1061T) neurons, which recapitulate the pathological features of the most prevalent human disease allele (22). Similar to cell models, both NPC1^−/− (Fig. 1 H and I) and NPC1^I1061T (Fig. 1 J and K) neurons exhibited significantly larger elevations in Ca²⁺ following G_qPCR stimulation as compared to wild type (WT) control neurons. Thus, across all models of NPC1 disease, including primary neurons, IP₃R-mediated Ca²⁺ release was significantly increased.

Fig. 1.

G_qPCR Ca²⁺ release is enhanced in NPC1 disease. (A) G_qPCR signaling pathway. (B) Average time series of Fluo-4 loaded control (black) and U18-treated (red) tsA201 cells treated with UTP (100 µM). (C) Quantitative analysis of the UTP-evoked Ca²⁺ area under the curve (AUC). (D and E) Same as B and C, only NPC1^−/− cells. (F and G) Same as B and C, only control and NPC1^I1061T patient cells and treated with histamine (100 µM). (H and I) Same as B and C, only NPC1^−/− neurons treated with Oxo-M. (J and K). Same as B and C, only NPC1^I1061T neurons treated Oxo-M. All the data are expressed as mean ± SEM from individual cells. Statistical analysis was an unpaired t test. ***P < 0.001; ****P < 0.0001.

NPC1 Inhibition or Disease Mutation Alters IP₃R1 Expression and Distribution.

To determine the molecular mechanism(s) underlying increased G_qPCR-induced Ca²⁺ release, we first measured mRNA levels from each of the PM elements along the G_qPCR signaling cascade from receptors to PLC enzymes (SI Appendix, Fig. S1 A and B). Quantification of gene transcripts from control and NPC1^I1061T patient cells revealed that several varieties of receptor, Gα_q, and PLC isoforms were unaltered following loss of NPC1 function. The sole exception was PLCβ2, which had elevated messenger ribonucleic acid (mRNA) levels (SI Appendix, Fig. S1B) but unaltered protein levels (SI Appendix, Fig. S1C) in NPC1^I1061T cells relative to control. Based on these analyses, the observation that three different G_qPCR agonists (histamine, purinergic, and muscarinic) all resulted in enhanced Ca²⁺ release (Fig. 1 B–K), and our previous determination that PM PI(4,5)P₂ is decreased in NPC1 disease (19) suggest that the molecular element(s) responsible for enhancing G_qPCR signaling in NPC1 disease cells do not lie at the PM.

We next examined whether elements downstream of PI(4,5)P₂ are involved in potentiating G_qPCR Ca²⁺ release in NPC1 cells. To begin, we analyzed IP₃R1, as it is 1) the most common isoform in the brain and 2) abundantly expressed in Purkinje cells of the cerebellum (9), the most vulnerable neuronal cell type in NPC (23). Western blot analysis determined that NPC1^I1061T patient cells had a 2.6-fold increase in protein levels compared to control (Fig. 2A). To test whether this expressional change in IP₃R1 protein levels translates to a functional role in potentiating G_qPCR Ca²⁺ release, we treated IP₃R1 knockout (IP₃R1^−/−) human embryonic kidney 293 (HEK) cells with U18 and evoked IP₃R-mediated Ca²⁺ release using the purinergic receptor agonist, UTP (100 μM) (24). Unlike overnight U18 treatment in control cells (Fig. 1B), we found that incubating IP₃R1^−/− cells with U18 did not alter G_qPCR Ca²⁺ responses (Fig. 2 B and C). These data support a model wherein increased IP₃R1 protein levels facilitate augmented IP₃R-mediated signaling following NPC1 dysfunction.

Fig. 2.

IP₃R1 expression and distribution is remodeled in NPC1 disease. (A, Top) representative Western blot of IP₃R1 in control and NPC1^I1061T disease fibroblasts. (Bottom) Quantification of IP₃R1 protein expression, normalized to β-Actin. (B) Average time series of Fluo-4 loaded control (black) and IP₃R1^−/− (red) HEK cells, treated with UTP. (C) Quantitative analysis of the UTP-evoked Ca²⁺ area under the curve (AUC). (D, Top) Schematic of TIRF imaging. (Bottom) Representative TIRF images from eGFP-IP₃R HeLa cells, treated with vehicle (Left) or U18 (Right). (E, Left) Quantification of average IP₃R puncta area. (Right) IP₃R density. (F) Representative TIRF images from control (Left) and NPC1^I1061T (Right) patient fibroblasts fixed and stained for IP₃R1. (G) Quantification of IP₃R1 cluster area (Left) and density (Right) from patient fibroblasts. (H) Representative images from cell body (Left) and neurite (Right) regions of control (Top) and U18-treated (Bottom) hippocampal neurons fixed and stained for IP₃R1 and MAP2. (I) Quantification of IP₃R1 intensity. (J) Schematic of TIRF image in neuron. (K) Representative images from control (Left) and U18-treated (Right) cortical neurons fixed and stained for IP₃R1. Insets show zoomed regions containing IP₃R1 clusters. (Right) Quantification of IP₃R1 cluster area. (L) Representative cerebellar sections from wild-type (Top) and NPC1^I1061T brains fixed and stained for calbindin and IP₃R1. All the data are expressed as mean ± SEM from individual cells. Statistical analyses were unpaired t tests. ns: not significant; ***P < 0.001; ****P < 0.0001.

The spatial distribution of IP₃Rs influences ER Ca²⁺ release, inactivation, wave propagation, and oscillations (3, 25–27). To determine whether increased protein levels alter the spatial organization of IP₃R1, we took several complementary fluorescent approaches to visualize the endogenous distribution of IP₃R1 in isolated cells, including neurons, and intact brain slices. To begin, we used HeLa cells with endogenous IP₃R1 tagged with a monomeric eGFP (eGFP-IP₃R cells) (28). eGFP-IP₃R cells allow us to quantify the endogenous distribution of the IP₃R1 without influence from overexpression or off-target binding of immunofluorescent antibodies. Total internal reflection fluorescence (TIRF) microscopy, to visualize IP₃R1 close to the PM, revealed that U18 treatment increased both the area and number of IP₃R1 clusters close to the PM (Fig. 2 D and E). To independently confirm that loss of NPC1 function alters IP₃R1 distribution near the PM, we performed immunofluorescence experiments on patient fibroblasts, neurons, and brain slices from the NPC1^I1061T murine model using a validated IP₃R1 antibody (SI Appendix, Fig. S2 A–D). TIRF imaging from patient fibroblasts determined that NPC1 disease mutations also increase IP₃R1 cluster size and number (Fig. 2 F and G). Similarly, IP₃R1 cluster intensity and area was increased at both the soma and neurites of isolated hippocampal (Fig. 2 H and I) and cortical neurons (Fig. 2 J and K). Finally, to ensure that 1) IP₃R1 clustering occurs in intact brain regions and 2) increased IP₃R1 cluster size occurs in brain regions affected by NPC1 disease, cerebellar slices from age-matched wild-type and symptomatic NPC1^I1061T animals were fixed and double immunostained for IP₃R1 and the Purkinje cell marker calbindin. In agreement with our single-cell data, IP₃R1 clusters within the Purkinje cell soma was significantly larger in NPC1 disease cerebellar sections than controls (Fig. 2L). Collectively, these data present strong evidence that loss of NPC1 function increases expression and distribution of IP₃R1 to enhance cluster size and increase G_qPCR-mediated Ca²⁺ release.

Nanoscale Distribution of IP₃R1 Is Altered in NPC1 Disease.

Having demonstrated that IP₃R1 clustering and distribution is altered near the PM in NPC1 disease using diffraction-limited approaches, we next wanted to precisely map and quantify the nanoscale organization of IP₃R1 at the PM using superresolution single-molecule localization microscopy in TIRF mode (superresolution_TIRF; resolution is approximately 20 nm) (29). This is important, because it allows us to investigate the nanoscale molecular architecture within and between IP₃R1 clusters to determine whether they are altered in NPC1 disease cells.

Quantitative analysis of superresolution_TIRF localization maps from control and U18-treated eGFP-IP₃R cells revealed that similar to diffraction-limited images, NPC1 inhibition resulted in an increase in both IP₃R cluster size and density (Fig. 3 A and B). Similar results were also obtained from neurons (Fig. 3 C and D). During analysis of localization maps, we noticed that single IP₃R1 clusters visualized using conventional TIRF microscopy actually consisted of many discrete puncta when imaged with superresolution microscopy (SI Appendix, Fig. S3 A–E). Cells lacking NPC1 function displayed a higher density of puncta in these discrete localized areas (SI Appendix, Fig. S3 D–F). Further, nearest-neighbor analysis measuring the distance from both a centroid and perimeter (SI Appendix, Fig. S3C) of these discrete puncta indicated that puncta were more closely associated with one another in U18-treated cells (SI Appendix, Fig. S3 G–J). Together, these data present evidence that loss of NPC1 function not only increases IP₃R puncta size but also increases the density of IP₃R1 puncta both within and between a cluster in NPC1 disease.

Fig. 3.

Nanoscale distribution of IP₃R1 is remodeled in NPC1 disease. (A, Left) Representative TIRF images of Vehicle (Top) and U18-treated (Bottom) eGFP-IP₃R HeLa cells fixed and immunostained with an anti-GFP antibody. (Right) Superresolution localization map from the same cell. Zoomed regions are from white solid rectangles. (B) Quantification of density and area of eGFP-IP₃R1. (C and D) Same as A and B, only cortical neurons fixed and stained for IP₃R1. All the data are expressed as mean ± SEM from individual cells. Statistical analyses were unpaired t tests. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

IP₃R1 Cluster Mobility and Activity Are Increased in NPC1 Disease.

Given that IP₃R1 cluster size is increased in NPC1 disease, we wanted to determine whether the mobility of IP₃R1 in the ER was altered. To this end, we used fluorescent recovery after photobleaching to quantify the mobility of eGFP-tagged IP₃R1 clusters both near the PM (SI Appendix, Fig. S4 A–C) and within the cell interior (SI Appendix, Fig. S4 D–F). Fitting of fluorescence recovery kinetics after photοbleaching revealed that the mobile fraction of IP₃R1 in U18- treated eGFP-IP₃R cells was significantly reduced compared to vehicle controls both near the PM (SI Appendix, Fig. S4C) and near the middle of the cell (SI Appendix, Fig. S4F).

Recently, it has been reported that immobile IP₃R1 clusters localized close to ER–PM junctions are optimally positioned to respond to receptor-generated IP₃ to initiate Ca²⁺ signals (28). Further, immobile IP₃R clusters have been observed to be the site of spontaneous localized Ca²⁺ release from the ER, termed “Ca²⁺ puffs” (28), with the activity of puffs further influenced by the number of IP₃R within the originating cluster (25, 30). Given that IP₃R1 clusters are larger, contain more discreet IP₃R1 puncta within a cluster, and are more immobile following NPC1 dysfunction, we wanted to determine whether spontaneous Ca²⁺ puffs were altered. First, we loaded eGFP-IP₃R cells with the Ca²⁺ dye Calbryte590 and simultaneously measured IP₃R1 distribution and spontaneous Ca²⁺ activity. This approach allows us to measure and correlate Ca²⁺ activity in the same region as IP₃R1 (Fig. 4A). We define this signal simply as Ca²⁺_IP3R1, because 1) the Ca²⁺ signal originates within the same location as IP₃R1, 2) is sensitive to the IP₃R antagonist Xestospongin C (SI Appendix, Fig. S5 A and C), and 3) spontaneous Ca²⁺ signals occur in the absence of extracellular Ca²⁺ (SI Appendix, Fig. S5B). We do acknowledge the possibility that these Ca²⁺ signals may not only result from IP₃R1 but represent a convolved signal from other Ca²⁺ sources. That being said, analysis of spontaneous Ca²⁺_IP3R1 signals determined that the amplitude, duration, and number of spontaneous Ca²⁺ events were all significantly elevated in U18-treated cells relative to control (Fig. 4 B and C), consisted with previous reports of increased spontaneous activity with larger IP₃R clusters.

Fig. 4.

Spontaneous intracellular IP₃R1 Ca²⁺ signaling is enhanced in NPC loss-of-function cells. (A) Representative maximum intensity projections of eGFP-IP₃R1 (Left) and Calbryte590 (Right) from control (Top) and U18-treated cells. Zoomed regions are taken from white rectangles. (B) Representative time series of spontaneous Ca²⁺ puffs in control cells (Top) and U18-treated (Bottom, red lines) cells. (Right) Zoomed time regions taken from dashed black rectangle. (C) Quantification of amplitude, duration, and number of spontaneous Ca²⁺ events. (D, Left) Representative maximum intensity projections from control fibroblasts loaded with Fluo-4. Zoomed regions are taken from solid yellow squares. (Right) Representative time series of spontaneous Ca²⁺ activity. (E) Same as D, only NPC1^I1061T patient fibroblasts. (F) Frequency distribution of the amplitude of Ca²⁺ events from healthy fibroblasts (black) and NPC1^I1061T disease fibroblasts (red). (G) Same as F, except showing the full-width half-max. (H) Same as F, except showing the duration of Ca²⁺ events. (I) Same as F, except showing Ca²⁺ puff mass. (J) Representative time series of Ca²⁺ events in control cells (Left) or IP₃R1^−/− cells (Right) loaded with Fluo-4 and treated with either U18 (red) or vehicle control (black). (K) Quantitative analysis of Ca²⁺ puff area under the curve from the cell types and conditions in J. (L) Same as K, except analyzing Ca²⁺ puff peak fluorescence from base. All the data are expressed as mean ± SEM from individual cells. Statistical analyses were unpaired t tests except K and L which were two-way ANOVAs. ns: not significant; *P < 0.05; **P < 0.01; ****P < 0.0001.

To test whether disease mutations exhibit altered Ca²⁺_IP3R1, we loaded control and NPC1^I1061T patient fibroblasts with Fluo-4 and measured spontaneous Ca²⁺ activity. Similar to U18-treated cells, NPC1^I1061T cells also had increased amplitude, duration, and mass (volume) of Ca²⁺ puffs compared to healthy controls (Fig. 4 D–I). To further investigate a role for IP₃R1 in facilitating these increases in Ca²⁺ activity, we treated control and IP₃R1^−/− cells with U18 or vehicle control and quantified spontaneous Ca²⁺ events. We found that treatment with U18 recapitulated the Ca²⁺ puff phenotype observed in eGFP-IP₃R cells and NPC1^I1061T fibroblasts, while IP₃R1^−/− cells were insensitive to the U18 (Fig. 4 J and K). Peak fluorescence of Ca²⁺ puffs was also quantified showing no significant difference between vehicle and U18 treatment in both cell types but a significant reduction in IP₃R1^−/− cells (Fig. 4L). These results indicate that the altered nanoscale distribution of IP₃R1 contribute to enhanced Ca²⁺ puff activity in NPC1 disease.

Presenilin 1 Alters IP₃R1 Distribution and Function in NPC1 Disease.

Having established that G_qPCR-mediated Ca²⁺ signaling, IP₃R1 distribution, and IP₃R1 activity are all severely altered in NPC1 disease, we next wanted to identify the underlying mechanism(s) responsible for these changes. Recently, we have reported that increases in full-length, holoprotein, Presenilin 1 (PS1) expression is involved in mediating decreases in ER Ca²⁺ in NPC1 disease (20). Further, Alzheimer’s disease mutations in PS1 (PS1^D257A) that shift expression mostly to the holoprotein form have previously been reported to interact and produce exaggerated gain of function alterations to the biophysical gating properties of IP₃Rs (31, 32). To interrogate a role for PS1 in tuning IP₃R1 distribution/function, we first transfected HeLa cells with HA-PS1 (HA-PS1-WT) or HA-tagged PS1^D257A and treated overnight with either vehicle control or U18 (10 µM; Fig. 5 A and B). TIRF imaging revealed that PS1-WT formed discrete clusters close to the PM, with cluster area significantly increased following overnight treatment with U18 to inhibit NPC1 function (Fig. 5 A and B). Interestingly, the PS1^D257A mutation also clustered close to the PM and displayed a similar mean cluster size as the PS1-WT group treated with U18 (Fig. 5 A and B). Treatment of PS1^D257A-expressing cells with U18 did not further increase PS1 cluster area (Fig. 5B). These data suggest that loss of NPC1 function or Alzheimer’s disease mutations increase the size of PS1 clusters near the PM. To test whether these increases correlate with changes in IP₃R1 distribution, we expressed the aforementioned constructs in eGFP-IP₃R1 HeLa cells and calculated Pearson’s correlation coefficients. Image analysis revealed that U18 treatment significantly increases the colocalization of PS1 to IP₃R1 (compare Fig. 5 C and D), while the PS1^D257A mutant also colocalizes more with IP₃R compared to wild-type PS1 even in the absence of NPC1 inhibition (Fig. 5E; analysis in Fig. 5F). These data are consistent with published observations that PS1 physically interacts with IP₃R1 and suggests there is a strong correlation between PS1 cluster size and IP₃R1 distribution. To test whether the full-length form of PS1, which is increased in NPC1 disease (20), alters IP₃R1, we overexpressed PS1^D257A in eGFP-IP₃R1 HeLa cells. Aligned with a key role for the holoprotein in mediating the distribution/function of IP₃R1, expression of PS1^D257A increased both IP₃R1 clusters (Fig. 5 G and H) and G_qPCR-mediated Ca²⁺ release (Fig. 5 I and J) to a similar extent as disease mutations or inhibition of NPC1. Altogether, these results suggest that enhanced PS1 cluster size may be responsible for altered nanoscale distribution/function of IP₃R1 following NPC1 loss of function.

Fig. 5.

Inhibition of NPC1 increases PS1 clusters and colocalization with IP₃R1. (A) Representative TIRF images of HeLa cells expressing HA-PS1, HA-PS1 + treated with U18, or HA-PS1^D257A. (B) Quantification of PS1 area normalized to footprint. (C, Top) TIRF images of eGFP-IP₃R HeLa cells transfected with HA-PS1-WT. (Bottom) Zoomed regions from solid white squares (Left) and line scan showing distribution of signals (Right). (D) Same as C, only treated with U18. (E) Same as C, only transfected with HA-PS1^D257A. (F) Quantitative analysis showing average Pearson’s correlation coefficients from C–E. (G) Representative TIRF images from Control (Top) and HA-PS1^D257A-expressing eGFP-IP₃R HeLa cells. (H) Quantification of IP₃R1 puncta intensity and size in control and HA-PS1^D257A-expressing cells. (I) Average time series of Fluo-4 loaded control (black), HA-PS1-expressing (red), and HA-PS1^D257A-expressing (orange) tsA201 cells treated with UTP (100 µM). (J) Quantitative analysis of the UTP-evoked Ca²⁺ area under the curve. All the data are expressed as mean ± SEM from individual cells. Statistical analyses were unpaired t tests except (B) and (F) which were two-way and one-way ANOVAs, respectively. ns: not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

A key question remained whether Ca²⁺ potentially coming through PS1 (33) can contribute to increases in IP₃R1, or whether only physical interactions (31) between the two proteins are important. To begin answering this question, we treated eGFP-IP₃R1 cells with the Sacro/ER Ca²⁺-ATPase (SERCA) inhibitor Thapsigargin or the Ca²⁺ chelating agent EGTA in a 0-mM, Ca²⁺-containing extracellular solution (SI Appendix, Fig. S6 A–D) and quantified IP₃R1 clusters near the PM over a period of 60 mins. Under each of these cytoplasmic Ca²⁺-altering conditions, IP₃R1 cluster size significantly decreased near the PM, independently of alterations in the cell footprint (SI Appendix, Fig. S6 E–M). Thus, although modulating cytoplasmic Ca²⁺ concentrations does alter IP₃R1 clustering near the PM, it is in the opposing direction compared to loss of NPC1 function.

Next, to test a direct role for PS1 driving IP₃R1 alterations in NPC1 disease, we knocked-down or knocked-out PS1 in different cells and measured IP₃R1 distribution, activity, and G_qPCR-mediated Ca²⁺ release. First, using small interfering RNA (siRNA) directed against PS1 in eGFP-IP₃R1 HeLa cells, which decreases PS1 protein levels by 75% (20), we determined that IP₃R1 cluster areas were refractory to U18 treatment (Fig. 6 A and B). Second, aligned with its ability to normalize IP₃R1 clusters back to control sizes, PS1 knock-down in NPC1^I1061T patient cells significantly reduced spontaneous Ca²⁺_IP3R1 puffs back to control levels (Fig. 6C). Third, NPC1^I1061T patient cells treated with siRNA PS1 had similar G_qPCR-mediated Ca²⁺ release as control (Fig. 6 D and E). As a final approach, we measured G_qPCR-mediated Ca²⁺ release from control (PS1^+/+) and PS1 knockout (PS1^−/−) mouse embryonic fibroblasts and determined that overnight U18 incubation increased G_qPCR Ca²⁺ release in PS1^+/+ cells (Fig. 6 F and H) but had no effect on PS1^−/− cells relative to vehicle controls (Fig. 6 G and H). Transfecting PS1^−/− cells with PS1-WT reestablished U-18-mediated increases in G_qPCR Ca²⁺ release (SI Appendix, Fig. S7 A and B). Collectively, these data support the hypothesis that PS1-dependent increases in IP₃R1 cluster size drive puff activity and G_qPCR-mediated Ca²⁺ release in NPC1 disease.

Fig. 6.

Knocking out PS1 or IP₃R1 rescues Ca²⁺ defects in NPC1 disease cells. (A) Representative TIRF images from control, U18-treated, and U18-treated + siRNA PS1 eGFP-IP₃R1 cells. Zoomed areas are taken from the solid white squares. (B) Quantification of eGFP-IP₃R1 cluster size. (C) Quantification of spontaneous Ca²⁺ activity. (D) Average time series of histamine responses from Fluo-4 loaded control (black), NPC1^I1061T(red), and NPC1^I1061T + siRNA PS1 (blue) fibroblasts. (E) Quantitative analysis of the histamine-evoked Ca²⁺ area under the curve (A.U.C). (F) Average time series of UTP responses from Fluo-4 loaded PSEN^+/+ MEF cells treated with vehicle control (black) or U18 (red). (G) Same as F, only PS1^−/− cells. (H) Quantification of UTP-evoked Fluo-4 signals. (I) Schematic of genetically encoded ER-GCaMP3 Ca²⁺ indicator. (J) Quantification of normalized ER-GCaMP3 intensity in patient fibroblasts. (K) Same as L, only tsA201 cells treated with U18. (L) Same as K, only tsA201 cells expressing PS1^D257A. (M) Same as K, only PS1^−/− cells. (N) Same as K, only IP₃R1^−/− cells. (O) Schematic of IP₃R1–PS1 hypothesis in NPC1 disease. (P) Average time series of thapsigargin responses from Fluo-4 loaded control (black line) and NPC1^−/− (red line) cells. (Q) Quantification of area under thapsigargin responses. (R) Same as P, except IP₃R1^−/− cells. (S) Same as Q. All the data are expressed as mean ± SEM from individual cells. B, C, and E were one-way ANOVAs; H was a two-way ANOVA; and J–N, Q, and S were unpaired t tests. ns: not significant; *P < 0.05; **P < 0.01; ****P < 0.0001.

Previously, we have reported that PS1 levels play a crucial role in decreasing ER Ca²⁺ (Ca²⁺_ER) to drive synaptic remodeling of Purkinje neurons in the cerebellum (20). Given the published role that PS1 plays in enhancing the biophysical release properties of IP₃R1 (32, 31, 34) and the increases in spontaneous Ca²⁺_IP3R1 puffs following loss of NPC1 function reported herein, we asked whether IP₃R1 is downstream of PS1 and involved in decreasing Ca²⁺_ER in NPC1 disease. To begin, we expressed a genetically encoded Ca²⁺ indicator targeted to the cytoplasmic leaflet of the ER membrane (Fig. 6I; ER-GCaMP3) in patient fibroblasts, U18-treated tsA cells, or PS1^D257A-expressing tsA cells to ask whether we could measure steady-state Ca²⁺_ER leak. Aligned with published data (20), both NPC1^I1061T patient cells (Fig. 6J) and U18-treated tsA201 cells (Fig. 6K) exhibited significantly higher resting ER-GCaMP3 intensity values, consistent with increased Ca²⁺_ER leak following loss of NPC1 function. To probe a role for PS1–IP₃R1 interactions facilitating enhanced Ca²⁺_ER leak in NPC1 disease, we performed similar experiments in PS1^−/− and IP₃R1^−/− cell lines and discovered that ER-GCaMP3 intensity values were refractory to U18 treatment under both conditions. These data suggest that PS1 and IP₃R1 are both key players in increasing Ca²⁺_ER leak in NPC1 disease.

The consequence of increased Ca²⁺_ER leak is a decrease in free Ca²⁺_ER concentrations in NPC1 loss-of-function cells (20); thus, we tested the effect of knocking out IP₃R1 (Fig. 6O). We loaded NPC1^−/− cells and IP₃R1^−/− cells with the cytosolic Ca²⁺ dye, Fluo-4, and applied the SERCA inhibitor thapsigargin in a 0-mM, Ca²⁺-containing extracellular solution. Treatment with thapsigargin gave the expected increase in cytoplasmic fluo-4 intensity from control cells, with NPC1^−/− cells having a smaller increase in intensity (20) (Fig. 6 P and Q). Conducting similar experiments in IP3R1^−/− cells treated with vehicle control or U18 did not result in any significantly differences in fluo-4 intensity following Thapsigargin treatment and provide evidence that increased PS1–IP₃R1 interactions drives decreased Ca²⁺_ER in NPC1 disease (Fig. 6 R and S).

Inhibition of SREBP Prevents Increases in IP₃R1 Cluster Size and G_qPCR Ca²⁺ Following Loss of NPC1 Function.

Loss of NPC1 function decreases lysosomal to ER cholesterol transport, resulting in depletion of ER membrane cholesterol and subsequent activation of the sterol regulatory element binding protein (SREBP) pathway, leading to increased gene expression of SREBP-dependent targets (35). We have previously reported that increased SREBP activity in NPC1 loss-of-function cells up-regulates expression of several Ca²⁺-handling proteins, including PS1, with inhibition of the SREBP pathway reversing differential protein expression along with aberrant ER Ca²⁺ and SOCE phenotypes (20). Thus, we wanted to determine whether the SREBP pathway is involved in the remodeling of IP₃Rs and potentiated G_qPCR-mediated Ca²⁺ release. To begin, we conducted qPCR experiments on NPC1^I1061T cells to determine whether mRNA levels of each IP₃R isoform are altered. Consistent with Western blot data, mRNA levels for IP₃R1, but not IP₃R2 or IP₃R3, were increased in NPC1^I1061T cells (Fig. 7A). Treating cells overnight with the SREBP inhibitor, PF-429242 (500 nM), abolished increases in IP₃R1 levels from NPC1^I1061T cells without affecting steady-state mRNA levels of IP₃R2 or IP₃R3. Furnished with the information that PF-429242 can decrease IP₃R1 mRNA levels, we next asked whether PF-429242 treatment rescues IP₃R1 cluster size in NPC1 loss-of-function cells. To answer this question, we treated eGFP-IP₃R1 cells with U18, or U18 and PF-429242, before quantifying IP₃R1 clusters using TIRF microscopy (Fig. 7B). Similar to the observed decreases in mRNA levels, treatment with PF-429242 decreased IP₃R1 cluster size back to control levels (Fig. 7C). Next, we measured G_qPCR-mediated Ca²⁺ release to test whether PF-429242-mediated decreases in IP₃R1 mRNA and cluster size would also decrease Ca²⁺ release. Analysis of intracellular Fluo-4 intensity following application of histamine determined that PF-429242 reduces the G_qPCR evoked Ca²⁺ signal in NPC1 fibroblasts back to healthy fibroblast control levels. Finally, we conducted G_qPCR experiments in SCAP^−/− cells that lack a functional SREBP pathway (36) and found that U18 treatment resulted in a slight decrease in Ca²⁺ release following receptor activation (Fig. 7 F and G). Therefore, like other Ca²⁺-handling proteins that are increased in NPC1 disease, IP₃R1 can be normalized back to control levels following inhibition of SREBP.

Fig. 7.

Inhibition or loss of SREBP pathway activity rescues expression and distribution of IP₃R1 and Ca²⁺ phenotypes in NPC1 disease. (A) Quantification of ITPR1, ITPR2, and ITPR3 mRNA levels in control (black), control + PF-429242 (gray), NPC1^I1061T (red), and NPC1^I1061T + PF-429242 (blue) patient fibroblasts. (B) Representative TIRF images form eGFP-IP₃R1 HeLa cells treated with U18 (Left) or U18 and PF-429242. (C) Quantification of IP₃R1 cluster size. (D) Average time series of histamine responses from Fluo-4 loaded control (black line), NPC1^I1061T (red line) cells, and NPC1^I1061T + PF429242 fibroblasts. (E) Quantification of area under the histamine responses. (F) Average time series of UTP responses from Fluo-4 loaded control (black line) or U18-treated (red line) SCAP^−/− cells. (G) Quantification of area under the UTP responses. All the data are expressed as mean ± SEM from individual cells. Statistical significance was calculated using the following test: A, two-way ANOVA; C and G, unpaired t tests; and E, one-way ANOVA. ns: not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

IP₃R1 Drives Ca²⁺-Dependent NFAT Translocation and Cytotoxic Elevations in Mitochondrial Ca²⁺ Following Loss of NPC1 Function.

Finally, we wanted to begin elucidating the cellular consequences of aberrant nanoscale distribution and activity of the IP₃R in NPC1 disease. Because alterations in cytoplasmic Ca²⁺ and IP₃R-mediated Ca²⁺ release are fundamentally required for many cellular processes, we focused on two of these pathways: NFAT signaling (37) and mitochondrial Ca²⁺ (13) (Ca²⁺_Mito) to ask whether they are altered in NPC1 disease.

NFAT proteins are a family of transcription factors who are phosphorylated and reside in the cytoplasm. Upon IP₃R stimulation to simultaneously elevate intracellular Ca²⁺ concentrations and deplete Ca²⁺_ER, thereby activating store-operated Ca²⁺ entry, NFATs are dephosphorylated by the Ca²⁺/calmodulin-dependent serine phosphatase, calcineurin, and translocate to the nucleus to become active (Fig. 8A). First, we tested whether steady-state NFAT distribution is different between control and NPC1^I1061T cells; analysis of fluorescence intensity ratios between the cytoplasm and nucleus revealed that NPC1^I1061T cells had a small but significant increase in the proportion of NFAT in the nucleus (Fig. 8 B and C). A similar distribution was observed in U18-treated cells (Fig. 8D). To test whether IP₃R1 is involved in mediating steady-state increases in nuclear NFAT in NPC1 loss-of-function cells, we performed similar experiments in IP₃R1^−/− cells. Quantification of NFAT cytoplasmic/nuclear intensities from U18-treated IP₃R1^−/− disclosed that NFAT no longer preferentially accumulated in the nucleus at steady-state (Fig. 8D). Next, we tested whether G_qPCR-stimulated NFAT translocation was also different between control and NPC1^I1061T patient cells. Time series analysis revealed that 100-s histamine application, a stimulus that increases cytoplasmic Ca²⁺ (Fig. 1F), increased NFAT more quickly in the nucleus of NPC1^I1061T cells relative to control (Fig. 8E). Thus, there is a correlation between the amount of G_qPCR-Ca²⁺ release and NFAT translocation. To probe the involvement of IP₃R1, we repeated experiments in tsA201 cells and IP₃R1^−/− cells treated with U18. Similar to NPC1^I1061T patient cells, treatment of tsA201 cells with U18 significantly increased NFAT translocation to the nucleus (Fig. 8F), an increase that was completely absent in IP₃R1^−/− cells (Fig. 8G). These experiments provide evidence that NPC1-dependent increases in IP₃R1 increase NFAT signaling in NPC1 disease.

Fig. 8.

Knockout of IP₃R1 rescues Ca²⁺-dependent gene transcription and Ca²⁺_Mito cytotoxicity in NPC1 loss-of-function cells. (A) Schematic of NFAT translocation to the nucleus. (B) Representative images from control (Top) and NPC1^I1061T (Bottom) fibroblasts expressing NFAT. (C) Quantification of NFAT intensity in the nucleus relative to the cytoplasm in control and NPC1^I1061T fibroblasts. (D) Same analysis as C, only using control and IP₃R1^−/− cells. (E) Averaged time series of nuclear NFAT intensity following the addition of Histamine in control (black line) and NPC1^I1061T (red line) fibroblasts. (F) Same as E, only UTP application in control and U18-treated tsA201 cells. (G) Same as F, only using IP₃R1^−/− cells. (H) Schematic of Ca²⁺ transfer at ER–mitochondria MCS. (I) Representative images of Mito-RCaMP in control (Top) and IP₃R1^−/− cells (Bottom) treated with vehicle control (Left) or U18 (Right). (J) Quantification of normalized Mito-RCaMP intensity in control and IP₃R1^−/− cells. (K) Quantification of live cell/dead cell ratio in control and IP₃R1^−/− cells treated with vehicle control or U18. All the data are expressed as mean ± SEM from individual cells. Statistical significance was calculated using the following tests: C, unpaired t test and D, J, and K, two-way ANOVA. ns: not significant; *P < 0.05; **P < 0.01; ****P < 0.0001.

As a final test to determine the cellular consequences of altered IP₃R1 signaling in NPC1 disease, we investigated Ca²⁺_Mito and cell viability. Ca²⁺ represents a key rheostat that couples mitochondria biogenetics to cellular demand and is frequently altered in pathophysiological conditions, including neurodegeneration. Intimate interactions between VDAC1 on mitochondria membranes and IP₃R1 on ER membranes provide the platform for constitutive IP₃R-mediated Ca²⁺ release to maintain cellular bioenergetics to meet cellular demands (38) (Fig. 8H). Given the importance of IP₃R1 in regulating Ca²⁺_Mito, we expressed a genetically encoded Ca²⁺ indicator targeted to mitochondria (Mito-RCaMPh1) to ask whether Ca²⁺_Mito is altered following loss of NPC1 function. Quantification of Mito-RCaMPh1 intensities between cells treated with a vehicle or U18 rvealed that loss of NPC1 function significantly increased Ca²⁺_Mito (Fig. 8 I and J). Identical experiments conducted on IP₃R1^−/− cells revealed reduced Ca²⁺_Mito levels compared to control with IP₃R1^−/− cells now being refractory to U18 treatment (Fig. 8J). In parallel, we asked whether increases in Ca²⁺_Mito, which can be cytotoxic in neurodegenerative conditions, are accompanied by changes in cell health. Consistent with this hypothesis, cell viability assays determined U18-dependent increases in Ca²⁺_Mito correlated with increased cell death (Fig. 8K), an observation that was completely absent in IP₃R1^−/− cells. Collectively, these data suggest that NPC1-dependent increases in IP₃R1 distribution/activity increase Ca²⁺_Mito cytotoxicity in NPC1 disease and presents the IP₃R as a potential target to slow neurodegeneration.

Discussion

Dysfunction of IP₃Rs is a common feature across many neurodegenerative diseases (39–41); thus, determining molecular details regarding their regulation and control is imperative for human health. Here, we discover a signaling axis that links the NPC1 lysosomal cholesterol transporter to the nanoscale distribution and activity of IP₃R1 (SI Appendix, Fig. S8). We show that loss of NPC1 function results in augmented clustering of IP₃R1 near the PM, increased frequency and size of spontaneous Ca²⁺ puffs from IP₃Rs, and enhanced G_qPCR-mediated intracellular Ca²⁺ signaling. Upstream of changes in IP₃R1 appears to be PS1, with overexpression of PS1^D257A phenocopying the observed changes in NPC1 disease cells. Mechanistically linking cholesterol egress from the lysosome to IP₃R1-dependent alterations in Ca²⁺ is the SREBP pathway, with inhibition of this transcription pathway restoring mRNA levels, IP₃R1 cluster size, and G_qPCR signaling, respectively. Functionally, altered IP₃R1 distribution/activity leads to increased steady-state and stimulated Ca²⁺-dependent NFAT signaling and Ca²⁺_mito toxicity.

G_qPCRs are pivotal purveyors of physiological information, transducing fluctuations in hormones, growth factors, and neurotransmitters across the PM to stimulate intracellular second messengers that initiate downstream signaling cascades. In this study, we determine that the PM-localized elements of G_qPCR signaling cascades, namely receptor, G_q, and PLC, are unaltered or, in the case of PI(4,5)P₂, decreased (19) following loss of NPC1 function. These observations, coupled with decreased Ca²⁺_ER levels (20), made it surprising to find that both receptor-activated and spontaneous IP₃R1-mediated Ca²⁺ responses are amplified in NPC1 disease cells. This counterintuitive discovery can be reconciled with information that IP₃R1 protein abundance, nanoscale distribution, mobility near the PM, and activity are all enhanced in NPC1 disease. The degree of spatial, temporal, and allosteric plasticity exhibited by IP₃R1 has been previously reported (25, 31, 42); we add another modulatory mechanism for IP₃R1 and G_qPCR-mediated Ca²⁺ release through NPC1 abundance and/or cholesterol egress across lysosomal–ER MCS. This positions NPC1 and lysosomal–ER MCS as potential targets to tune G_qPCR signaling with the goal of altering global or local Ca²⁺ release to modify neuronal bioenergetics, Ca²⁺-dependent transcription, or synaptic plasticity in health or disease.

We detail an important role for PS1 in regulating IP₃R1 distribution and activity following loss of NPC1 function. Under normal physiological conditions, PS1 undergoes autoproteolytic cleavage, resulting in two fragments (N and C terminals) that associate with other proteins to create the γ-secretase complex. Several mutations in the gene encoding PS1 can abolish this autoproteolytic activity, keeping PS1 in an uncleaved and full-length state, which is purported to form Ca²⁺ leak channels at the ER membrane (33). Contrasting the hypothesis that PS1 is a Ca²⁺ leak channel, Cheung et al. have reported that both Familial Alzheimer’s Disease–linked PS1 mutations, which lack the ER Ca²⁺ leak activity, as well as the catalytically dead PS1 mutants that maintain their full uncleaved length bind and enhance the gating properties of IP₃Rs (31), suggesting the mechanism of Ca²⁺_ER leak involves physical interactions between PS1 and IP₃R1. We find that PS1 and IP₃Rs have little colocalization after overexpression of wild-type PS1; however, inhibition of NPC1 significantly increased their localization with one another. Additionally, overexpression of the catalytically dead PS1^D257A mutation that results in increased full-length protein form reorganizes IP₃R1 nanoscale distribution, spontaneous puff activity, and G_qPCR-mediated Ca²⁺ release, independent of NPC1 inhibition. Finally, IP₃R1^−/− cells are refractory to U18 with no NPC1-dependent Ca²⁺ phenotypes. This leads us to suggest that it is the physical interactions between PS1 and IP₃R1, coupled with elevations in cytoplasmic Ca²⁺ (20), that modify the clustering properties of IP₃R1 to enhance their cooperative gating and thereby their release probability in NPC disease. If this model is correct, it would position IP₃R1-mediated Ca²⁺ release as the distal element underlying enhanced Ca²⁺_ER leak in NPC1 disease. Questions remain to fully determine the molecular choreography of PS1–IP₃R1 complex, including the following: 1) What is the optimal stoichiometry of PS1–IP₃R1 interactions? 2) Do decreases in IP₃R1 mobility act as a diffusion trap for PS1 to bind, aggregate, and influence IP₃R1? 3) Does decreased ER cholesterol in NPC1 cells, as evidenced by enhanced SREBP processing (20), reduce cholesterol-dependent partitioning of PS1 and IP₃R1 to allow their increased clustering/interaction? and 4) Are there other accessory proteins that exert regulatory influence over this signaling complex?

The SREBP transcription factor appears to link changes in NPC1 function to alterations in IP₃R1 and G_qPCR signaling. Previously, we have reported that the SREBP pathway is involved in mediating expressional changes in the Ca²⁺-handling proteins SERCA, STIM1, ORAI1, and PS1 in NPC1^I1061T cells; therefore, it is perhaps not surprising that the IP₃R1 is also influenced by SREBP. That being said, it is still unknown whether regulation of IP₃R1 by SREBP is through direct or indirect transcriptional regulation. Further, a key question that remains unanswered is whether alterations in intracellular Ca²⁺ signaling are purely deleterious sequences (see paragraph below) or are part of a molecular mechanism(s) attempting to correct the cellular phenotypes of NPC1 disease. There are clear links between elevations in intracellular Ca²⁺ levels and endocytic fusion and transport events; further, decreasing Ca²⁺_ER has been reported to redistribute the intracellular pool of unesterified cholesterol (43) and rescue key cellular phenotypes of NPC1 disease (6), while knocking out INPP5A (44) (IP₃ 5-phosphatase) triggers depletion of cellular cholesterol. Thus, changes in Ca²⁺_ER signaling may be a key rheostat for maintenance and control of cellular cholesterol homeostasis.

Ca²⁺-handling proteins have been linked to a host of pathologies, including several neurodegenerative disorders (33, 34). MCS between the ER and mitochondrial are regions of intimate membrane apposition that facilitate transport of Ca²⁺ from the ER to the mitochondria. Following G_qPCR activation or spontaneous IP₃R opening, Ca²⁺ is taken up into mitochondria to participate in the control of bioenergetics by activating oxidative metabolism, mitochondrial respiration, and ATP synthesis. Notably, dysregulation of Ca²⁺_Mito are potent triggers of necrosis, apoptosis, and autophagy (for review, refer to ref. 45). Here, we present evidence that NPC1-dependent increases in IP₃R1 distribution and activity increase Ca²⁺_Mito to enhance cell death in NPC1 disease models. Thus, although increased Ca²⁺ release from the ER may be initiated as a mechanism to alter cholesterol homeostasis or redistribute toxic metabolites associated with NPC1 disease mutations, the prolonged elevations in cytoplasmic and Ca²⁺_Mito could ultimately be a central driver of neuronal cell death in NPC disease. This presents IP₃R1 and MAMs as potential therapeutic targets to correct Ca²⁺ phenotypes in order to slow Purkinje neuron neurodegeneration and NPC disease progression.

In conclusion, we characterize a link between the lysosomal NPC1 cholesterol transporter and changes in IP₃R-dependent Ca²⁺ signaling. We find that differential expressional and localization of IP₃R1 enhances both spontaneous and G_qPCR-mediated Ca²⁺ release to promote mitochondrial cytotoxicity. This work furthers our understanding of the regulatory factors that tune IP₃R signaling, most importantly PS1, whose form and expression are altered in NPC1 disease. By elucidating these regulatory pathways, we have provided potential therapeutic targets to address IP₃R dysfunction, a feature of NPC1 disease and other neurodegenerative disorders.

Materials and Methods

Detailed methods can be found in SI Appendix. Briefly, NPC1^I1061T patient cells, NPC1^−/− cells, and tsA201 cells treated overnight with U18666A and male and female NPC1^I1061T mice served as models of NPC disease and were comparted to control cells in this study. Protein levels were determined using immunoblot approaches. Ca²⁺ imaging was performed using an Andor W1 spinning disk system. Immunofluorescence TIRF, superresolution, and confocal images were acquired using a Leica single-molecule localization system or a Zeiss AiryScan confocal microscope.

Data Availability

All study data are included in the article and/or SI Appendix.