Research ArticleIon Channels

mTORC1 controls lysosomal Ca2+ release through the two-pore channel TPC2

See allHide authors and affiliations

Science Signaling  10 Apr 2018:
Vol. 11, Issue 525, eaao5775
DOI: 10.1126/scisignal.aao5775

Channeling lysosomal Ca2+

Inhibition of the multiprotein complex mTORC1 alleviates pulmonary hypertension in animal models. Using mouse pulmonary arterial myocytes and transfected cells, Ogunbayo et al. identified mTORC1 inhibition as a mechanism for activation of the ion channel TPC2 that led to the mobilization of Ca2+ from lysosomes. Furthermore, they showed that TPC2 may be regulated by mTORC1 inhibition and the ligand NAADP through a common pathway. The authors suggest that modulating TPC2 activity could be a promising therapeutic strategy for pulmonary hypertension, which currently lacks effective treatments.

Abstract

Two-pore segment channel 2 (TPC2) is a ubiquitously expressed, lysosomally targeted ion channel that aids in terminating autophagy and is inhibited upon its association with mechanistic target of rapamycin (mTOR). It is controversial whether TPC2 mediates lysosomal Ca2+ release or selectively conducts Na+ and whether the binding of nicotinic acid adenine dinucleotide phosphate (NAADP) or phosphatidylinositol 3,5-bisphosphate [PI(3,5)P2] is required for the activity of this ion channel. We show that TPC2 is required for intracellular Ca2+ signaling in response to NAADP or to mTOR inhibition by rapamycin. In pulmonary arterial myocytes, rapamycin and NAADP evoked global Ca2+ transients that were blocked by depletion of lysosomal Ca2+ stores. Preincubation of cells with high concentrations of rapamycin resulted in desensitization and blocked NAADP-evoked Ca2+ signals. Moreover, rapamycin and NAADP did not evoke discernable Ca2+ transients in myocytes derived from Tpcn2 knockout mice, which showed normal responses to other Ca2+-mobilizing signals. In HEK293 cells stably overexpressing human TPC2, shRNA-mediated knockdown of mTOR blocked rapamycin- and NAADP-evoked Ca2+ signals. Confocal imaging of a genetically encoded Ca2+ indicator fused to TPC2 demonstrated that rapamycin-evoked Ca2+ signals localized to lysosomes and were in close proximity to TPC2. Therefore, inactivation of mTOR may activate TPC2 and consequently lysosomal Ca2+ release.

INTRODUCTION

The two-pore segment channels (TPCs) (1) are nicotinic acid adenine dinucleotide phosphate (NAADP)–gated Ca2+ release channels that are targeted to endolysosomes (27). Consistent with this view, TPC1, TPC2, and TPC3, of which only TPCN1 and TPCN2 genes (2, 8) are present in humans, rats, and mice, are homologous with and represent an evolutionary intermediate between some Ca2+-permeable transient receptor potential (TRP) channels and voltage-gated Ca2+ channels, with an intermediate two-domain structure of subunits that likely assemble as homodimers (2, 8). However, controversy surrounds the capacity of TPC2 to support an NAADP-gated Ca2+ conductance, given that others have demonstrated that human TPC2 expressed in several different mammalian cell types is highly selective for Na+ and carries a Na+ conductance in response to phosphatidylinositol 3,5-bisphosphate [PI(3,5)P2] but not NAADP (7, 9, 10). Countering this notion, cells derived from Tpcn1 and Tpcn2 null mice generate PI(3,5)P2-mediated cation currents but not NAADP-gated Ca2+ currents, and NAADP-dependent cation currents are restored in Tpcn1 and Tpcn2 null cells by overexpression of wild-type TPC2 (11). That said, it is clear that TPC2 mediates a Na+ conductance that is inhibited by mechanistic target of rapamycin (mTOR) (10) and TPC2 contributes to autophagy termination by facilitating mTOR reactivation (12). Activation of mTOR complex 1 (mTORC1) has been implicated in the progression of diseases such as pulmonary hypertension for which current therapies are ineffective (13). Therefore, we sought to determine whether TPC2 mediates lysosomal Ca2+ signals in pulmonary arterial myocytes upon intracellular dialysis of NAADP and PI(3,5)P2, respectively, and the role of mTOR in these processes.

RESULTS

TPC2 deletion in pulmonary arterial myocytes abolishes global Ca2+ waves in response to NAADP but not PI(3,5)P2, inositol 1,4,5-trisphosphate, or cyclic adenosine 5′-diphosphoribose

We examined the action of various Ca2+ mobilizing messengers in pulmonary arterial myocytes prepared from wild-type mice and Tpcn2 knockout mice (2, 12) by using intracellular dialysis from a patch pipette under voltage clamp, with the holding potential set to −40 mV to inactivate voltage-gated Ca2+ channels. Intracellular dialysis of NAADP induced global Ca2+ waves in acutely isolated pulmonary arterial myocytes from wild-type mice (Fig. 1A). By contrast, NAADP did not induce a measurable Ca2+ transient in paired experiments on myocytes derived from Tpcn2 knockout mice (Fig. 1A), as assessed by the Fura-2 F340/F380 ratio. In contrast, the response to intracellular dialysis of either PI(3,5)P2, inositol 1,4,5-trisphosphate [(1,4,5)P3], or cyclic adenosine 5′-diphosphoribose (cADPR) was similar in myocytes derived from Tpcn2 knockout mice when compared to wild-type cells (Fig. 1, B to E).

Fig. 1 Deletion of TPC2 in pulmonary arterial myocytes blocks Ca2+ transients evoked by NAADP but not those triggered by PI(3,5)P2, I(1,4,5)P3, or cADPR.

(A) Upper panels show a bright-field (BF) image of an acutely isolated mouse pulmonary arterial smooth muscle cell (PASMC) and a series of pseudocolor images of the Fura-2 fluorescence ratio (F340/F380) recorded in the same cell during intracellular dialysis from a patch pipette of 10 nM nicotinic acid adenine dinucleotide phosphate (NAADP). Lower panel shows corresponding record (black) of F340/F380 ratio against time; the time points at which pseudocolor images were acquired are indicated by the numbered lines. WC indicates the beginning of intracellular dialysis upon entering the whole-cell configuration. The red, blue, cyan, and green records show the effect of NAADP in myocytes from Tpcn2 knockout (KO) mice. WT, wild-type. (B to D) As in (A) but showing responses of myocytes from Tpcn2 knockout mice to intracellular dialysis of phosphatidylinositol 3,5-bisphosphate [PI(3,5)P2] (B), inositol 1,4,5-trisphosphate [I(1,4,5)P3] (C), or cyclic adenosine 5′-diphosphoribose (cADPR) (D). (E) Bar chart shows the mean ± SEM for each stimulus across all cells studied [number of cells (n) indicated above bars]. ns, not significant; **P < 0.01, ***P < 0.001, ****P < 0.0001. (F) Upper panel shows a bright-field image of a PASMC acutely isolated from a WT mouse and a series of pseudocolor images of the Fura-2 fluorescence ratio (F340/F380) recorded in the same cell during intracellular dialysis from a patch pipette of 10 nM NAADP. Lower panel shows corresponding record (black) of F340/F380 ratio against time; the time points at which pseudocolor images were acquired are indicated by the numbered lines. WC indicates the beginning of intracellular dialysis upon entering the whole-cell configuration. Red record shows the effect of NAADP after preincubation (≥50 min) of a myocyte with 1 μM bafilomycin-A1 (Baf-A1). (G) As in (F) but with a blue record that shows the effect of NAADP in a different cell after preincubation with the SR Ca2+ adenosine triphosphatase inhibitor thapsigargin (1 μM; ≥40 min). (H) Response of a WT mouse PASMC to extracellular application of 25 μM glycyl-l-phenylalanine-2-naphthylamide (GPN). (I) As in (H) but showing the response of a myocyte from a Tpcn2 knockout mouse. (J and K) As in (H) and (I) but showing responses to 1 μM bafilomycin-A1. (L and M) Bar charts show the mean ± SEM [number of cells (n) indicated above bars] for the experiments shown in (F), (G), (H), and (J) for WT (L) and (I) and (K) for Tpcn2 KO mice (M). *P < 0.05.

Consistent with our previous studies on rat pulmonary arterial myocytes (14), the response of mouse myocytes to NAADP was abolished by depletion of acidic Ca2+ stores with bafilomycin-A1 (Fig. 1F) and attenuated after depletion of sarcoplasmic reticulum (SR) Ca2+ stores by preincubation with the SR Ca2+ adenosine triphosphatase (ATPase) inhibitor thapsigargin (Fig. 1G), suggesting that lysosomal Ca2+ bursts are subsequently amplified by Ca2+-induced Ca2+ release from SR stores. Nevertheless, the failure of NAADP to induce Ca2+ transients in myocytes from Tpcn2 knockout mice was not due to the loss of releasable, acidic Ca2+ stores, because pronounced Ca2+ transients were evoked upon lysis of lysosomes by glycyl-l-phenylalanine-2-naphthylamide [GPN; Fig. 1, H and I (15)] and upon depletion of acidic stores by application of the vacuolar proton pump (V-H+-ATPase) inhibitor bafilomycin-A1 (Fig. 1, J to M). Furthermore and consistent with previous studies (1619), we found that two different voltage-gated Ca2+ channel antagonists, the dihydropyridine nifedipine and the phenylalkylamine verapamil, blocked NAADP-induced Ca2+ signals in pulmonary arterial myocytes (Fig. 2, A and B) and in human embryonic kidney (HEK) 293 cells stably overexpressing hTPC2 (Fig. 2C) to comparable extents (Fig. 2D). These results are consistent with verapamil blocking the Na+ conductance carried by hTPC2 (9). From these data, we conclude that irrespective of the continued presence of releasable, acidic Ca2+ stores, TPC1, functional ryanodine receptors, and I(1,4,5)P3 receptors, TPC2 is required for the induction of global Ca2+ signals by NAADP but by contrast is not necessary for Ca2+ signaling by PI(3,5)P2, I(1,4,5)P3, or cADPR, at least in the cell types studied here. Consistent with this view, double-blind paired analysis showed that intracellular dialysis of NAADP induced Ca2+ transients in HEK293 cells that stably overexpressed enhanced green fluorescent protein (EGFP)–tagged hTPC2 but not in those expressing its inactive mutant (N653K) that does not support Na+ conductance (fig. S1, A and B). As previously reported (9), we found by direct electrophysiological measurements of isolated lysosomes that hTPC2 conferred a PI(3,5)P2-gated, NAADP-insensitive Na+ conductance. However, in the presence of high luminal and “cytoplasmic” Ca2+ [60 mM in pipette, 200 μM in bath, which reproduces previously used conditions (11)], TPC2 supported an NAADP- and PI(3,5)P2-sensitive Ca2+ conductance in our hands (fig. S2, A to C and E), consistent with previous reports on the capacity of TPC2 to support a small (7), NAADP-sensitive Ca2+ conductance (11).

Fig. 2 Nifedipine and verapamil block NAADP-evoked Ca2+ transients.

(A) Upper panels show a bright-field image of an acutely isolated rat PASMC and a series of pseudocolor images of the Fura-2 fluorescence ratio (F340/F380) recorded in the same cell during intracellular dialysis from a patch pipette of 10 nM NAADP. Lower panel shows corresponding record (black) of F340/F380 ratio against time; the time points at which pseudocolor images were acquired are indicated by the numbered lines. WC indicates the beginning of intracellular dialysis upon entering the whole-cell configuration. The magenta and brown records show the effect of NAADP in myocytes preincubated with 10 μM nifedipine and 10 μM verapamil, respectively. (B) As in (A) but for an acutely isolated WT mouse PASMC. (C) As in (A) but for a human embryonic kidney (HEK) 293 cell stably overexpressing hTPC2. (D) Bar chart shows the mean ± SEM for each stimulus across all cells studied [number of cells (n) indicated above bars]. *P < 0.05, **P < 0.01, ***P < 0.001.

Rapamycin induces Ca2+ signals through acidic stores in an mTOR- and TPC2-dependent manner

Extracellular application of rapamycin induced concentration-dependent Ca2+ signals in hTPC2-expressing HEK293 cells (Fig. 3, A to E and H); by contrast, rapamycin induced small Ca2+ transients in wild-type HEK293 cells (Fig. 3I), which have low TPC2 abundance (2). Rapamycin triggered Ca2+ transients in hTPC2-expressing HEK293 cells at as low a concentration as 100 nM (Fig. 3A). At the higher concentrations tested, rapamycin-evoked Ca2+ transients were multiphasic, characterized by an initial large transient followed by Ca2+ oscillations that were variable in number and magnitude, and all were superimposed upon what appeared to be a slower, more uniform, and prolonged Ca2+ signal (Fig. 3I). However, the concentration-response curve for rapamycin-induced Ca2+ signals was bell-shaped, peaking at 30 μM and exhibiting progressive levels of “desensitization” between 100 and 300 μM, at which concentrations rapamycin-induced desensitization occurred without a detectable Ca2+ signal being evoked (Fig. 3H). These results are consistent with our previous reports on the nature of the response of these cells (20) and the response of pulmonary arterial myocytes (14) to NAADP. Although rapamycin inhibits mTORC1 at low concentrations, we found that at the high concentration that gave the peak response in the Ca2+ assay (30 μM), the drug decreased mTORC1 activity in HEK293 cells faster than at the lower concentrations (300 nM), as indicated by S6K phosphorylation (fig. S3, A and B). Therefore, 30 μM rapamycin was used in all subsequent experiments to ensure robust and reproducible responses.

Fig. 3 Rapamycin induces Ca2+ transients in pulmonary arterial myocytes and hTPC2-expressing HEK293 cells that are blocked by bafilomycin-A1 and nifedipine.

(A to H) Example records (black) showing the concentration-response relationship as changes in Fura-2 fluorescence ratio (F340/F380) against time during extracellular application of indicated concentrations of rapamycin (0.1 to 300 μM) onto different HEK293 cells that stably expressed human TPC2. (I) Upper panel shows a bright-field image of a HEK293 cell stably overexpressing hTPC2 and a series of pseudocolor images of F340/F380 recorded in the same cell during extracellular application of 30 μM rapamycin. Lower panel shows corresponding record (black) of F340/F380 ratio against time; the time points at which pseudocolor images were acquired are indicated by the numbered lines. Green record shows the effect of dimethyl sulfoxide (DMSO) (vehicle control) in a different cell, and a cyan record shows the effect of rapamycin on a WT HEK293 cell. (J) As in (I) but with a red record showing the effect of rapamycin after preincubation (≥50 min) of an hTPC2-expressing HEK293 cell with 1 μM bafilomycin-A1. (K) As in (I) but with an orange record showing the effect of rapamycin after preincubation (≥50 min) of a cell with 1 μM thapsigargin. (L) As in (I) but with a magenta record showing the effect of rapamycin after preincubation (≥50 min) of a cell with 10 μM nifedipine. (M to P) As in (I) to (L) but for acutely isolated rat pulmonary arterial myocytes, except that (O) shows an additional record in gold of the effect of rapamycin after preincubation (≥50 min) of a myocyte with 100 μM ryanodine.

The response to 30 μM rapamycin (Fig. 3I) was abolished by previous depletion of acidic Ca2+ stores with bafilomycin-A1 (Fig. 3J) and attenuated by preincubation of cells with thapsigargin (Fig. 3K). Thapsigargin blocked the fast oscillating Ca2+ transients but not the slow progressive rise in Ca2+ on which these were superimposed, suggesting that the Ca2+ oscillations likely result from Ca2+-induced Ca2+ release from the endoplasmic reticulum triggered by previous Ca2+ release from acidic Ca2+ stores. Furthermore, all signals were abolished by preincubation of cells with nifedipine (Fig. 3L). In pulmonary arterial myocytes, rapamycin induced Ca2+ transients that were mostly biphasic, characterized by a single rapid, global Ca2+ transient that was once again superimposed on an underlying slower transient (Fig. 3M). Occasionally, the initial, fast Ca2+ transient was followed by multiple smaller transients that were all superimposed on the slower underlying Ca2+ signal as for HEK293 cells, but such Ca2+ oscillations were not a common feature of the response to rapamycin for pulmonary arterial myocytes. In the myocytes, rapamycin-evoked Ca2+ transients were blocked by previous depletion of acidic Ca2+ stores with bafilomycin-A1 (Fig. 3N), attenuated by preincubation with thapsigargin (Fig. 3O) or ryanodine (Fig. 3O), and abolished by preincubation with nifedipine (Fig. 3P). These results were similar to those obtained in HEK293 cells stably overexpressing hTPC2 and to that previously reported for NAADP-evoked Ca2+ signals in pulmonary arterial myocytes (14). Because of the more complex nature of rapamycin-induced Ca2+ signals, from this point forward, we analyzed the responses for mTOR inhibition by measuring the “area under the curve” (AUC), which may provide a more accurate assessment of outcomes (Fig. 4, A and B). Consistent with previous findings of others for TPC2 (19), rapamycin- and NAADP-induced Ca2+ transients were also blocked by tetrandrine (Fig. 4, C and D), which has previously been shown to activate autophagy (21) and block TPC2 (22). Moreover, similar to rapamycin, the mTOR inhibitor torin-2 also evoked biphasic Ca2+ transients in HEK293 cells stably overexpressing hTPC2 and in pulmonary arterial myocytes (Fig. 4, C and D). These Ca2+ transients exhibited characteristics similar to those induced by NAADP (fig. S4, A to D). Increases in cytoplasmic Ca2+ concentration were also induced by torin-1, but these were sustained and did not return to baseline during the course of these experiments (fig. S5, A to C).

Fig. 4 Comparison of the effects of rapamycin and torin-2 in pulmonary arterial smooth muscle cells acutely isolated from rats or mice, WT HEK293 cells and HEK293 cells stably overexpressing hTPC2.

(A) Bar chart showing the concentration-response relationship (mean ± SEM) for the peak change in Fura-2 fluorescence ratio (ΔF340/F380) during extracellular application of rapamycin onto HEK293 cells that stably expressed hTPC2. (B) As in (A) but for area under the curve. (C) Bar chart compares the mean ± SEM for the peak change in Fura-2 fluorescence ratio during extracellular application of rapamycin onto WT HEK293 cells, HEK293 cells that stably expressed hTPC2 and pulmonary arterial myocytes under all conditions studied (see key). (D) As in (C) but for area under the curve [number of cells (n) indicated above bars]. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Neither rapamycin nor torin-2 induced measurable Ca2+ transients in myocytes derived from Tpcn2 knockout mice (Fig. 5, A to C). Moreover, short hairpin RNA (shRNA) knockdown of mTOR in HEK293 cells that stably overexpressed hTPC2 attenuated Ca2+ transients induced by rapamycin and NAADP (Fig. 5, D and G). These data suggest that rapamycin induces Ca2+ transients in a TPC2-dependent manner by inhibiting mTOR, and TPC2 may therefore be the point of convergence for the induction of Ca2+ signals by NAADP and mTOR inhibition. This notion was confirmed by cross-desensitization of signaling between mTOR inhibition and NAADP. NAADP-induced Ca2+ signals were abolished by preincubation with rapamycin or torin-2 both in pulmonary arterial smooth muscle cells (PASMCs) (Fig. 6, A to D) and in HEK293 cells that stably overexpressed hTPC2 (Fig. 6, E and H). However, the effects of previous intracellular dialysis of high, desensitizing concentrations of NAADP on rapamycin-induced Ca2+ signals were more variable but generally appeared to have little effect. In rat pulmonary arterial myocytes, after intracellular dialysis of NAADP, Ca2+ transients evoked by rapamycin were marginally attenuated only when assessed by AUC, and these differences were not significant (Fig. 6, C and D). Moreover, in hTPC2-HEK293 cells, previous intracellular dialysis of desensitizing concentrations of NAADP did not affect Ca2+ transients (Fig. 6, G and H). This contrary outcome may be due to disparities with respect to the amount of time in which pulmonary arterial myocytes and hTPC2-expressing HEK293 cells could be held in the whole-cell configuration and the time dependence of the process of desensitization by high concentrations of NAADP. Alternatively, NAADP may self-desensitize this macromolecular signaling complex at a point upstream of mTOR.

Fig. 5 Ca2+ transients induced by mTOR inhibitors in pulmonary arterial myocytes are blocked by deletion of TPC2 and by shRNA knockdown of mTOR.

(A) Upper panel shows a bright-field image of a WT mouse PASMC and a series of pseudocolor images of the Fura-2 fluorescence ratio (F340/F380) recorded in the same cell during extracellular application of 30 μM rapamycin. Lower panel shows corresponding record (black) of F340/F380 ratio against time; the time points at which pseudocolor images were acquired are indicated by the numbered lines. Pink record shows the response to 30 μM rapamycin of a myocyte isolated from a Tpcn2 knockout (Tpcn2 KO) mouse. (B and C) Bar charts show the mean ± SEM [number of cells (n) indicated above bars] for the peak change in F340/F380 during the first transient recorded (B) and the area under the curve during the response of each cell to rapamycin and torin-2 (C). ***P < 0.001, ****P < 0.0001. (D) Upper panel shows a bright-field image of a HEK293 cell overexpressing hTPC2 48 hours after transfection of scrambled short hairpin RNA (shRNA) and a series of pseudocolor images of the Fura-2 fluorescence ratio (F340/F380) recorded in the same cell during extracellular application of 30 μM rapamycin. Lower panel shows corresponding record (black) of F340/F380 ratio against time; the time points at which pseudocolor images were acquired are indicated by the numbered lines. Green and pink records show the response to DMSO and 30 μM rapamycin, respectively, of two different hTPC2-expressing HEK293 cells that were transfected with shRNA against mechanistic target of rapamycin (mTOR). (E) As in (D) but for different cells during intracellular dialysis with 10 nM NAADP, as performed in Fig. 1; red record for a cell 48 hours after transfection of scrambled shRNA and blue record for a cell transfected with shRNA against mTOR. (F and G) Bar chart shows the mean ± SEM [number of cells (n) indicated above bars] for the experiments shown in (D) and (E), for the peak change in F340/F380 ratio induced by NAADP and the peak change attained during the first transient recorded after rapamycin (C) and the area under the curve during the response to rapamycin (D). *P < 0.05, ***P < 0.001, ****P < 0.0001.

Fig. 6 Ca2+ transients evoked by mTOR inhibition and NAADP exhibit cross-desensitization in pulmonary arterial myocytes and hTPC2-expressing HEK293 cells.

(A) Upper panel shows a bright-field image of a rat PASMC and a series of pseudocolor images of the Fura-2 fluorescence ratio (F340/F380) recorded in the same cell during intracellular dialysis of 10 nM NAADP. Lower panel shows corresponding record (black) of F340/F380 ratio against time; the time points at which pseudocolor images were acquired are indicated by the numbered lines. WC indicates the beginning of intracellular dialysis upon entering the whole-cell configuration. Blue and brown records show, respectively, the response to 10 nM NAADP of myocytes preincubated (30 min) with 30 μM rapamycin or 300 nM torin-2. (B) As in (A) but for rapamycin in the absence of (red) and 2 min after intracellular dialysis of 100 μM NAADP (pink). (C and D) Bar charts show the mean ± SEM for the experiments shown in (A) and (B) for the peak change induced by 10 nM NAADP and the peak change attained during the first transient recorded after 30 μM rapamycin (C) and the area under the curve during the response to rapamycin (D). (E to H) As in (A) to (D) but for HEK293 cells stably overexpressing hTPC2 [number of cells (n) indicated above bars]. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Rapamycin evokes bafilomycin-A1– and nifedipine-sensitive Ca2+ signals proximal to TPC2 in HEK293 cells

We next used HEK293 cells that stably overexpressed GCaMP5-hTPC2, in which the genetically encoded Ca2+ indicator GCaMP5 was fused to the cytoplasmic N terminus of TPC2, allowing the detection of Ca2+ signals arising near the lysosome-localized TPC2. Confocal imaging showed that GCaMP5-hTPC2 specifically labeled LysoTracker Red–positive vesicles, which formed either static clusters of variable density or smaller motile units (Fig. 7A). Application of rapamycin induced slowly developing increases in fluorescence that arose and returned to baseline with a time course of 60 to 120 s (Fig. 7B). These followed a similar time course to the slower, thapsigargin-insensitive signals recorded by way of Fura-2 fluorescence ratio (Fig. 3K). These findings suggest that the sampling interval (0.2 Hz) for confocal experiments did not provide the temporal resolution necessary to reveal those fast Ca2+ transients superimposed on the slower, thapsigargin-insensitive components of rapamycin-induced signals described above. Consistent with this view, these signals were blocked by previous depletion of acidic stores by bafilomycin-A1 and by preincubation with nifedipine but were not affected by preincubation of cells with thapsigargin (Fig. 7, C to E). Furthermore, nifedipine reduced basal fluctuations in GCaMP5-hTPC2 fluorescence relative to control, providing indirect support for the view that this L-type voltage-gated Ca2+ channel antagonist directly inhibits lysosomal Ca2+ efflux, whether through TPC2 or a distinct pathway governed by activation of TPC2.

Fig. 7 Rapamycin induces bafilomycin- and nifedipine-sensitive Ca2+ signals proximal to lysosomes in HEK293 cells that stably overexpress GCaMP5-TPC2.

(A) Deconvolved confocal images show a three-dimensional reconstruction (from left to right) of a single HEK293 cell stably expressing GCaMP5-hTPC2 (green), the distribution of LysoTracker-Red (red) labeling within the same cell, and a merged image depicting regions of colocalization (yellow). (B) Upper panels show confocal images of a HEK293 cell overexpressing GCaMP5-hTPC2 (green) during extracellular application of 30 μM rapamycin. Lower panel shows corresponding record (black) of F/F0 ratio against time; the time points at which confocal images were acquired are indicated by the numbered lines. Green record shows the effect of DMSO (vehicle control) in a different cell. (C) As in (B) but showing the response of different cells to 30 μM rapamycin after preincubation with 1 μM thapsigargin (black), 1 μM bafilomycin-A1 (red), and 10 μM nifedipine (pink). (D and E) Bar charts show the mean ± SEM for the peak change in F/F0 (D) and the area under the curve during the response to rapamycin (E) [number of cells (n) indicated above bars]. *P < 0.05, **P < 0.01. A.U., arbitrary units; TG, thapsigargin.

DISCUSSION

Using fully differentiated, acutely isolated pulmonary arterial myocytes to examine the role of TPC2 in lysosomal Ca2+ signaling (2325) in combination with HEK293 cell lines stably overexpressing human TPC2, we showed that TPC2 was required for global Ca2+ transients in response to intracellular dialysis of NAADP. Briefly, NAADP failed to induce global Ca2+ transients in pulmonary arterial myocytes from Tpcn2 knockout mice although both lysosomes and the SR retained replete and releasable Ca2+ stores, mobilization of the former being triggered by GPN and bafilomycin-A1 and the latter by intracellular dialysis of I(1,4,5)P3 and cADPR. TPC1 (2), ryanodine receptors, and I(1,4,5)P3 receptors remain available in pulmonary arterial myocytes (14, 20, 2426) from Tpcn2 knockout mice, thereby demonstrating that TPC2 is required for the induction of global Ca2+ waves by NAADP in this cell type (24, 25).

These findings recapitulate our previous observations on HEK293 cells, which suggested that overexpression of hTPC2 is required to support NAADP-dependent global Ca2+ signals (2, 20) and that recombinant expression of ryanodine receptors is required for cADPR-induced Ca2+ transients (20). These findings are consistent with the observation that HEK293 cells primarily rely on I(1,4,5)P3 receptors to mediate global Ca2+ signals, while exhibiting little or no endogenous functional expression of either TPCs (2) or ryanodine receptors (27). On the other hand, these findings argue against the proposal that PI(3,5)P2-gated Ca2+ signals arise in a TPC2-dependent manner in these cells, regardless of whether TPC2 itself supports a Na+-specific conductance (9) or both Ca2+ and Na+ conductances (11). This view was strengthened by our studies on HEK293 cells and pulmonary arterial myocytes, in which we found that PI(3,5)P2 not only evoked Ca2+ oscillations in HEK293 cells stably overexpressing hTPC2 and pulmonary arterial myocytes derived from wild-type mice, but that PI(3,5)P2 also did so in pulmonary arterial myocytes from Tpcn2 null mice and in wild-type HEK293 cells, in which endogenous expression of TPC2 is too low to support NAADP-dependent global Ca2+ signals (2, 20). These outcomes are therefore consistent with the retention of PI(3,5)P2-gated cation currents in cells from Tpcn1 and Tpcn2 double-knockout mice (11). Accordingly, others have shown that PI(3,5)P2 can gate various channel families other than TPCs, including TRPML1 and ryanodine receptors (2830). We cannot, however, rule out the possibility that PI(3,5)P2 availability is required to support TPC2 activity.

We found that rapamycin and another mTOR inhibitor, torin-2, also induced Ca2+ signals by mobilizing bafilomycin-A1–sensitive, acidic stores in a manner that was markedly attenuated by shRNA knockdown of mTOR and abolished by Tpcn2 deletion. These data suggest that mTOR inhibitors induce Ca2+ release from lysosomes through TPC2 in an mTOR-dependent manner rather than by directly binding to TPC2, although we cannot rule out the possibility that some mTOR inhibitors may also directly activate TPC2; torin-1 appeared to have a distinct mode of activating TPC2 (fig. S5, A to D). Moreover, using HEK293 cells stably overexpressing GCaMP5-hTPC2, we showed that rapamycin increased lysosomal Ca2+ flux proximal to hTPC2 itself. This finding is not inconsistent with the findings of others that suggest that mTOR is an endogenous inhibitor of PI(3,5)P2-gated Na+ currents carried by TPC2 (10). Although small when compared to their estimated Na+ permeability, it appears that endogenous TPC2 may support a Ca2+ conductance sufficient to trigger global Ca2+ signals due to amplification of lysosomal Ca2+ bursts at lysosome-SR nanojunctions (2325), as has been suggested by computer simulations (23) and electrophysiological investigations on the ion selectivity conferred by the filter sequence of the human TPC2 channel pore (7) and whole-endolysosomal recordings for human TPC2 carried out with Ca2+ as the major permeant ion by others (7) and us (fig. S2, A to D). This notion is supported by our demonstration of cross-desensitization between mTOR inhibitor and NAADP with respect to their capacity to evoke Ca2+ signals both in pulmonary arterial myocytes and in HEK293 cells stably overexpressing hTPC2. This observation is consistent with the indirect activation of TPC2 by NAADP through associated proteins (31, 32) and may confirm the primary physiological role of TPC2, in that mTOR-dependent Ca2+ mobilization may modulate the luminal pH of lysosomes (33, 34) and thus autophagic flux (12, 35).

We conclude that inactivation of mTOR is a major endogenous pathway for initiating lysosomal Ca2+ release in HEK293 cells stably overexpressing hTPC2 and in pulmonary arterial myocytes. Thus, the mTORC1-TPC2 complex may act as the endogenous gatekeeper of lysosomal Ca2+ flux, which may in turn affect autophagy termination and mTOR reactivation (12). Because mTORC1-dependent myocyte proliferation promotes the progression of pulmonary hypertension (13), for which current therapies are ineffective, TPC2 may be a potential therapeutic target in idiopathic pulmonary hypertension and pulmonary hypertension secondary to lysosomal dysfunction associated with Pompe and Gaucher diseases (36, 37). In this respect, our demonstration that nifedipine attenuates basal, NAADP- and rapamycin-evoked lysosomal Ca2+ release in a TPC2-dependent manner is notable, because this Ca2+ channel antagonist is used in the treatment of pulmonary hypertension (38, 39).

MATERIALS AND METHODS

Preparation of PASMCs and HEK293 cell lines stably expressing human TPC2

Experiments were performed in accordance with the United Kingdom Animals (Scientific Procedures) Act 1986. Single arterial smooth muscle cells were isolated from second-order branches (external diameter, ~1 mm) of the pulmonary artery. Briefly, the arteries were dissected from lungs of male wild-type (C57/Bl6) and Tpcn2 knockout mice (28 to 33 g) (2), or Wistar rats (250 to 300 g), and single cells were acutely isolated from pulmonary arterial smooth muscle from mouse or rat as previously described (24) and placed in low-Ca2+ solution of the following composition: 135 mM NaCl, 5 mM KCl, 1 mM MgCl2, 0.5 mM NaH2PO4, 0.5 mM KH2PO, 15 mM NaHCO3, 0.16 mM CaCl2, 0.5 mM EDTA, 10 mM glucose, 10 mM taurine, and 10 mM Hepes (pH 7.4). For cell isolation, dissected arteries were transferred into fresh low-Ca2+ solution containing papain (1 mg/ml), dithiothreitol (0.8 mg/ml), and bovine serum albumin (0.7 mg/ml) and incubated for 10 min at 37°C and gently triturated using a fire-polished glass pipette to obtain dispersed PASMCs. Acutely isolated PASMCs were then plated in a 35-mm cell-culture dish (FluoroDish, World Precision Instruments Inc.) and used within 24 hours. The HEK293 cell line stably expressing human TPC2 was also developed and cultured as previously described (2). Site-directed mutagenesis was performed to generate a TPC2 N653K mutant that lacked a Na+ conductance using the QuikChange Site-Directed Mutagenesis Kit (Stratagene). The mutation was confirmed by DNA sequencing.

Ca2+ imaging

Cells were incubated for 30 min with 5 μM Fura-2-AM in nominally Ca2+-free physiological salt solution (PSS) of the following composition: 130 mM NaCl, 5.2 mM KCl, 1 mM MgCl2, 10 mM glucose, and 10 mM Hepes (pH 7.45) in an experimental chamber that was then placed on a Leica DM IRB/E inverted microscope after washing with Ca2+ containing 1.7 mM CaCl2, Fura-2–free PSS for at least 30 min before experimentation. PSS was of the following composition: 130 mM NaCl, 5.2 mM KCl, 1 mM MgCl2, 1.7 mM CaCl2, 10 mM glucose, and 10 mM Hepes (pH 7.45). Cytoplasmic Ca2+ concentration was reported by Fura-2 fluorescence ratio (F340/F380 excitation; emission, 510 nm). Emitted fluorescence was recorded at 22°C with a sampling frequency of 0.5 Hz using a Hamamatsu 4880 charge-coupled device camera and a Zeiss Fluar 40× [1.3 numerical aperture (NA)] oil immersion objective. Background subtraction was performed online. Analysis was completed using OpenLab imaging software (PerkinElmer). NAADP (10 nM) was applied intracellularly into single cells in the whole-cell configuration of the patch-clamp technique (voltage clamp mode; holding potential = −40 mV). The pipette solution contained 140 mM KCl, 10 mM Hepes, 1 mM MgCl2, and 5 μM Fura-2 (free acid) (pH 7.4) nominally Ca2+-free (~100 nM). The seal resistance was ≥3 gigohms throughout each experiment. Series and pipette resistance were ≤10 and ≤3 megohms, respectively, as measured by an AxoPatch 200B Amplifier (Axon Instruments).

Confocal microscopy

The GCaMP5 fluorescence ratio (expressed as F/F0; F0 = GCaMP5 fluorescence intensity at 0 s; Fx = fluorescence intensity at time x for given region of interest) was recorded at 22°C with a sampling frequency of 0.5 Hz, using a Nikon A1R+ confocal system and a Nikon Eclipse Ti inverted microscope with a Nikon Apo 40× λS DIC N2 (1.25 NA) water immersion objective (Nikon Instruments Europe BV). Experiments were processed with ImageJ software (Rasband WS. ImageJ, U.S. National Institutes of Health, Bethesda, Maryland, USA, imagej.nih.gov/ij/, 1997—2012).

Analysis of mTOR activities in response to rapamycin treatment by immunoblotting

HEK293 cells were seeded in six-well plates at 40 to 50% confluence and cultured overnight to reach ~70% confluence. Rapamycin was diluted to final concentrations of 0.3 and 30 μM in complete culture medium [Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum] and warmed to 37°C before being added to the cells. At desired times (1 to 60 min) after the addition of rapamycin, the medium was aspirated, and 200 μl of 1× SDS sampling buffer [50 mM tris-HCl, 2% SDS, 10% glycerol, 1% β-mercaptoethanol, 12.5 mM EDTA, and 0.02% bromophenol blue (pH 6.8)] was immediately added to cause cell lysis. Cell lysates were then collected into 1.5-ml Eppendorf tubes and sonicated for 15 s before boiling at 99°C for 5 min. Lysates were resolved in 7.5% tris-glycine SDS–polyacrylamide gel electrophoresis gel and analyzed using mouse anti–phospho-S6K (Thr389) (1:1000; Cell Signaling Technology, catalog no. 9206) and rabbit anti-S6K (1:1000; Cell Signaling Technology, catalog no. 9202) primary antibodies. DyLight 800 goat anti-mouse (1:5000; Invitrogen, catalog no. SA5-10176) and DyLight 680 goat anti-rabbit (1:5000; Invitrogen, catalog no. 21109) secondary antibodies were used to reveal signals, and blots were scanned with fluorescent immunoblot instruments and software by LI-COR Odyssey software images.

Whole-endolysosome patch-clamp experiments

Endolysosomal patch-clamp recordings were performed in isolated enlarged endolysosomes using a modified patch-clamp method (9, 11, 40). HEK293 cells stably expressing TPC2-EGFP were treated with 1 μM vacuolin-1 for 12 to 36 hours. Electrophysiological recordings were performed using an EPC10 USB acquisition system (HEKA). PATCHMASTER software (HEKA) was used to record and analyze data. Recording solutions followed either that by Ruas et al. (11) or that by Wang et al. (9), which differ mainly in that the pipette solution was either Ca2+-rich (Na+-free) or Na+-rich, respectively. The Ca2+-rich pipette solution contained 70 mM K-methanesulfonate (MSA), 60 mM Ca-MSA, 1 mM MgCl2, and 10 mM Hepes (pH adjusted with MSA to 4.6 and mannitol used to adjust osmolarity); the bath solution contained 130 mM K-MSA, 0.2 mM Ca-MSA, and 10 mM Hepes (pH adjusted with KOH to 7.2). The Na+-rich pipette solution had 145 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM Hepes, 10 mM MES, and 10 mM glucose (pH adjusted with NaOH to 4.6); the bath solution contained 140 mM K-gluconate, 4 mM NaCl, 1 mM EGTA, 2 mM MgCl2, 0.39 mM CaCl2, and 20 mM Hepes (pH adjusted with KOH to 7.2; free [Ca2+] = 100 nM). Briefly, an isolated pipette was used to slice the cell membrane and release endolysosomes labeled with EGFP fluorescence. Only one enlarged endolysosome vacuole was recorded from each coverslip. Then, a freshly polished recording pipette was used to form a gigaseal to the vacuole. ZAP pulses (fast pulses of −400 to −1000 mV in 50- to 100-ms durations) were used to establish the whole-endolysosome configuration. Capacitance transients were compensated automatically. The holding potential was set at 70 mV, and inside-out recording mode was used for data acquisition in which the inward currents were defined as currents flowing out of lysosomal lumen into the cytosol. Voltage ramps from −150 to +150 mV within 200 ms were applied every second. A 20-ms step to −150 mV and a 20-ms step to +150 mV were included at the beginning and end of each ramp, respectively. Currents at −150 mV were used for further analysis.

Data presentation and statistical analysis

Data are presented as means ± SEM for n experiments. Comparisons between groups were by Kruskal-Wallis test with Dunn’s multiple comparison test and nonparametric unpaired t test. Probability values less than 0.05 were considered to be statistically significant.

Drugs and chemicals

Unless otherwise stated, all compounds were from Sigma-Aldrich.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/11/525/eaao5775/DC1

Fig. S1. Blind experiments on active and null hTPC2 constructs demonstrate robustness of intracellular dialysis technique.

Fig. S2. Na+ and Ca2+ currents mediated by endolysosomal TPC2 in response to NAADP and PI(3,5)P2.

Fig. S3. High and low concentrations of rapamycin suppress mTORC1 activities in HEK293 cells at different rates.

Fig. S4. Torin-2 induces increases in intracellular Ca2+ in HEK293 cells stably overexpressing hTPC2 and in rat pulmonary arterial myocytes.

Fig. S5. Torin-1 induces low-magnitude, sustained increases in intracellular Ca2+ in HEK293 cells stably overexpressing hTPC2 and in rat pulmonary arterial myocytes.

REFERENCES AND NOTES

Acknowledgments: We would like to thank P. Skehel for his help and advice. Funding: This work was primarily funded by a British Heart Foundation Programme Grant (29885 to A.M.E.), and by NIH RO1 grants (GM 092759 to M.X.Z. and AR 070752 to J.M.). J.D. also received support from the China Scholarship Council (201508060127). Author contributions: O.A.O. and A.M.E. carried out calcium imaging of pulmonary arterial myocytes and HEK293 cells. O.A.O., J.D., and A.M.E. carried out confocal microscopy. O.A.O., J.X., and M.X.Z. carried out shRNA knockdown of mTOR. J.X. generated the mutant constructs and performed immunoblotting. Q.W. and X.F. carried out whole-endolysosomal recordings. J.X. and M.X.Z. developed cell lines expressing human TPC2 and TPC2-GCaMP5. A.M.E., M.X.Z., and J.M. designed experiments. A.M.E. wrote the manuscript and all authors commented on the manuscript. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Stable cell lines described in this paper will be distributed to qualified researchers in academic institutions using an unmodified version of the material transfer agreement approved by the University of Texas Health Science Center at Houston.
View Abstract

Navigate This Article