Mobilization of Lysosomal Calcium Regulates the Externalization of Phosphatidylserine during Apoptosis*

A hallmark of apoptotic cells is the Ca2+-dependent appearance of phosphatidylserine (PS) at the cell surface as a result of its redistribution from the inner-to-outer plasma membrane leaflet. Although endoplasmic reticulum and mitochondrial Ca2+ are known to participate in apoptosis, their role in PS externalization has not been established. In this study, several organelle-specific fluorescent markers and Ca2+-sensitive probes were used to identify the source of Ca2+ critical to PS externalization. By employing Rhod-2AM, fluorescein-labeled high molecular weight dextran, and Calcium Green 1, we provide evidence that lysosomes respond to apoptotic stimuli by releasing their luminal Ca2+ to the cytosol. Cells treated with the cytosolic phospholipase A2 inhibitor, cPLA2α, had no effect on caspase activation but exhibited a significant decrease in lysosomal Ca2+ release and externalization of PS in response to apoptotic stimuli. Similarly, cells depleted of lysosomal Ca2+ underwent programmed cell death yet failed to externalize PS. These data indicate that although Ca2+ release from other intracellular organelles to the cytosol is adequate for apoptosis, the release of Ca2+ from lysosomes is critical for PS externalization.

A hallmark of apoptotic cells is the Ca 2؉ -dependent appearance of phosphatidylserine (PS) at the cell surface as a result of its redistribution from the inner-to-outer plasma membrane leaflet. Although endoplasmic reticulum and mitochondrial Ca 2؉ are known to participate in apoptosis, their role in PS externalization has not been established. In this study, several organelle-specific fluorescent markers and Ca 2؉ -sensitive probes were used to identify the source of Ca 2؉ critical to PS externalization. By employing Rhod-2AM, fluorescein-labeled high molecular weight dextran, and Calcium Green 1, we provide evidence that lysosomes respond to apoptotic stimuli by releasing their luminal Ca 2؉ to the cytosol. Cells treated with the cytosolic phospholipase A 2 inhibitor, cPLA2␣, had no effect on caspase activation but exhibited a significant decrease in lysosomal Ca 2؉ release and externalization of PS in response to apoptotic stimuli. Similarly, cells depleted of lysosomal Ca 2؉ underwent programmed cell death yet failed to externalize PS. These data indicate that although Ca 2؉ release from other intracellular organelles to the cytosol is adequate for apoptosis, the release of Ca 2؉ from lysosomes is critical for PS externalization.
The regulated transbilayer redistribution of phosphatidylserine (PS) 2 from the inner-to-outer bilayer leaflet of cells is a physiologically significant event critical to apoptosis and many other processes (1)(2)(3). Once translocated from the cytoplasmic to the external membrane leaflet, PS provides a platform for the assembly of enzyme complexes that regulate the propagation of the coagulation cascade (4) and serves as a key recognition ligand for the ultimate elimination of the cell from the host by phagocytic reticuloendothelial cells (5)(6)(7). Despite tremendous progress in deciphering apoptotic pathways, however, the mechanism(s) responsible for the appearance of PS at the cell surface remains elusive (8 -10).
Numerous studies have revealed the involvement of several lipid transporters in the regulation of transbilayer lipid distributions in eukaryotic cells (11,12). The concerted action of an inward-directed aminophospholipid translocase and a less specific outward-directed floppase are considered to be responsible for maintaining the dynamic equilibrium distribution of phospholipids between the plasma membrane bilayer leaflets of quiescent cells (13). Activation of a putative protein, coined phospholipid "scramblase," dissipates normal membrane lipid asymmetry by facilitating the bidirectional transbilayer movement of all phospholipid classes irrespective of the composition of the polar head group of the lipid (14,15). The most pronounced effect of perturbed lipid asymmetry is the appearance of PS in the external bilayer leaflet. Although the activation pathways for scramblase activity may differ between different cell types, a key upstream event seems to be a sustained elevation in intracellular [Ca 2ϩ ] that occurs through its release from intracellular stores and by capacitive entry across the plasma membrane.
The major stores of intracellular Ca 2ϩ are the mitochondria (16) and ER (17,18). Although the regulation of Ca 2ϩ transients through these stores clearly represent critical check points in the regulation of apoptosis, the role of other intracellular Ca 2ϩ stores in the externalization of PS has not been determined. In this study, we provide evidence that lysosomes respond to apoptotic stimuli by releasing their luminal Ca 2ϩ to the cytosol. Our data indicate that Ca 2ϩ released from these vesicles is critical to apoptosis-dependent expression of PS at the cell surface.

EXPERIMENTAL PROCEDURES
Cells, Cell Lines, and Reagents-Murine embryonic fibroblasts (MEF) were cultured in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 15% fetal bovine serum, 50 g/ml uridine, 2 mM glutamine, and 100 units/ml leukemia inhibitory factor. All of the fluorescent probes used for Ca 2ϩ measurements and intracellular organelle identification were purchased from Invitrogen. The phospholipase A 2 inhibitor, cPLA2␣ inhibitor (catalog number 525143), was purchased from Calbiochem.
Real-time Ca 2ϩ Measurements-MEF (10 6 ) were incubated with 1 M Fura-2 acetoxymethyl ester (19) (Fura-2AM; Invitrogen) at 37°C for 30 min. The cells were then washed with Tris-buffered Hanks' balanced salt solution and resuspended in the same buffer for fluorescence measurements. Cytosolic Ca 2ϩ levels were measured in real time by assessing the ratio (R) of fluorescence intensities at 510 nm following excitation at 340 and 380 nm. To determine total cell Ca 2ϩ (R max ) and baseline levels (R min ) of fluorescence, Nonidet P-40 (0.2%) and EGTA (2 mM) were added, respectively. Lysosomal Ca 2ϩ was assessed by incubating cells with Rhod-2AM or Calcium Green-1 (CG1) (20) (cell-impermeant hexapotassium salt; 1 M for 1 h at 37°C) followed by confocal microscopy.
Labeling of Intracellular Organelles-Intracellular organelles were identified by incubating MEF on glass coverslips with MitoTracker green (21), ER tracker green (22), NBD-C6-ceramide (23), and LysoTracker Green to label mitochondria, ER, Golgi, and lysosomes, respectively. All the probes (1 M) were incubated with the cells in serum-free Dulbecco's modified Eagle's medium containing 2 mM Ca 2ϩ for 1 h. Noninternalized probe was then removed from the plasma membrane by incubating the cells for an additional 10 min in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. The cells were then washed and resuspended in Hanks' balanced salt solution.
Confocal Microscopy-Fluorescent images of live cells following the various treatments were recorded with an Axioplan 2 Zeiss LSM 510 confocal microscope equipped with an Achroplan ϫ63 water immersion lens. The 488-and 543-nm lines were used and monitored through a 505-530-nm band pass and 560-nm highpass filters, respectively.

Sustained Elevation in Cytosolic Ca 2ϩ Is Required for PS Externalization-Previous results from this laboratory have
shown that incubation of Jurkat cells with thiol-disulfide exchange reagents results in the immediate and sustained release of Ca 2ϩ from intracellular stores to the cytosol and redistribution of PS from the inner-to-outer membrane leaflet of the cell (3). This occurs in the absence of cytochrome c release, caspase activation, and DNA fragmentation, suggesting that independent mechanisms regulate PS externalization and apoptosis. Indeed, although apoptosis is not reversible (once cytochrome c release is initiated), reduction of reagent-induced disulfides with dithiothreitol reactivated aminophospholipid translocase activity and restored membrane lipid asymmetry (3).
Similar to Jurkat cells, PS externalization in MEF can be enforced through the release of Ca 2ϩ from endogenous stores to the cytosol with thiol reagents (3) or by incubation of the cells in the presence of Ca 2ϩ ionophore/Ca 2ϩ (Fig. 1A). Because thapsigargin triggers release of Ca 2ϩ from ER stores, we deter-  mined whether the levels of cytosolic Ca 2ϩ achieved were also sufficient for PS externalization. Fig. 1B shows that the transient increase in cytosolic Ca 2ϩ levels achieved by treating the cells with 50 nM thapsigargin in the absence of extracellular Ca 2ϩ did not result in PS exposure. In the presence of 2 mM extracellular Ca 2ϩ , however, the addition of thapsigargin (50 nM) resulted in a fluorescence spike that did not recover to base levels during the time course of the experiment. Similarly, treatment of the cells with 200 nM of thapsigargin also resulted in a significant spike that recovered to a steady-state level (within ϳ5 min.) similar to that achieved with 50 nm TG in the presence of Ca 2ϩ . Assessment of PS externalization revealed that treatments resulting in sustained steady-state elevations in cytosolic Ca 2ϩ were associated with the appearance of PS at the cell surface (Fig. 1B, inset). It should be noted that FACS analysis within the spike period (ϳ5 min) failed to detect FITCannexin V-labeled cells (not shown). Taken together, these data suggest that PS externalization is dependent on sustained elevations in cytosolic Ca 2ϩ .
Apoptosis-dependent Increase in Cytosolic Ca 2ϩ Levels Is Associated with PS Exposure-The data presented above suggest a direct relationship between the levels of cytosolic Ca 2ϩ and the propensity of cells to externalize PS. To determine whether PS externalization during apoptosis is also related to cytosolic Ca 2ϩ levels, apoptosis was initiated in MEF with STS. At the indicated times, the cells were washed and labeled with Fura-2AM or Rhod-2AM, and cytosolic Ca 2ϩ levels were estimated by fluorescence spectroscopy and confocal microscopy, respectively (Fig. 2). Cells incubated with STS demonstrated a progressive increase in cytosolic Ca 2ϩ levels and a concomitant increase in the fraction of annexin V-positive, PS-expressing, cells. Consistent with these observations, the cytosol of Rhod-2-labeled control cells did not fluoresce, indicating very low preapoptosis Ca 2ϩ levels (Fig. 2B). Upon apoptosis, however, a dramatic increase in the intensity of cytosolic fluorescence was observed (Fig. 2C), indicating a significant increase in the concentration of cytosolic Ca 2ϩ . Taken together, with the TG results shown in Fig. 1, these data suggest that the propensity of cells to externalize PS is directly dependent on sustained increases in cytosolic [Ca 2ϩ ].
Rhod-2AM Labeling of Endosomes and Lysosomes-In contrast to the typical mitochondrial staining pattern obtained after incubation of the cells with Rhod-2AM at 4°C (Figs. 2B and 3A), cells exhibited prominent staining of relatively large vacuole-like organelles when incubated with the probe at 37°C (Fig.  3B). To identify these organelles, the cells were labeled with known organelle tracking dyes. Double labeling of the cells with MitoTracker green (mitochondria label), NBD-C6-ceramide (Golgi and ER label), or ER tracker green (endoplasmic reticulum label) together with Rhod-2AM at 37°C indicated that these organelles were distinct from mitochondria, Golgi, and ER, respectively (Fig. 3, C-E).
Because the morphology and size of the Rhod-2-labeled organelles were consistent with endocytotic vesicles, we determined whether these structures were pinocytotic Ca 2ϩ -containing vesicles by incubating cells with Rhod-2AM in the presence of the aqueous space marker FITC-labeled dextran or the lysosome tracer, LysoTracker Green. Both the probes co-localized to the Rhod-2AM-labeled organelles, indicating that these structures were indeed endosomes (Fig. 3F) and/or lysosomes A-E, MEF were labeled with Rhod-2AM and MitoTracker green at 4°C (A) or with Rhod-2 AM alone at 37°C (B) for 1 h and subsequently with MitoTracker green at 37°C for 1 h (C), Golgi probe NBD-C6-ceramide at 37°C for 1 h (D), or ER probe, ER tracker green at 37°C for 1 h (E). F, shows co-localization of Rhod-2 and FITC-dextran labeling at 37°C for 1 h (pinocytotic vesicle staining). Left, red channel (Rhod-2); middle, green channel (FITC-dextran); right, red/green overlap. G shows co-localization of Rhod-2 and LysoTracker Green (lysosomal staining). Left, red channel (Rhod-2); middle, green channel (LysoTracker Green); right, red/green overlap. (Fig. 3G) that acquired their luminal Ca 2ϩ from endocytosed fluid-phase vesicles that sequestered exogenous Ca 2ϩ from the tissue culture medium.
Role of Endocytotic Vesicle Ca 2ϩ in PS Externalization-To determine the fate of vesicular Ca 2ϩ in apoptosis-induced PS externalization, endosomes were specifically labeled by incubating MEF in the presence of the cell-impermeable Ca 2ϩ probe, CG1, and the aqueous space marker dextran red. Confocal microscopy of control and apoptotic cells showed that both probes redistributed in an independent manner (Fig. 4). In control cells, the vesicles adopted a preferential perinuclear distribution, and essentially all of them were labeled with both fluorescent probes (Fig. 4A). In apoptotic cells, however, the CG1 fluorescence appeared to be cytosolic, suggesting that Ca 2ϩ entrapped within the vesicles redistributed to the cytosol in an apoptosis-dependent manner (Fig. 4B). The average intensity of CG1 fluorescence in the cytosol of apoptotic cells was ϳ1.7-fold higher than in control cells (Fig. 4, center panels).
Although these data raise the possibility that vesicular Ca 2ϩ plays a role in apoptosis-dependent PS externalization, it is not clear whether this is an obligatory source of Ca 2ϩ . To unequivocally determine whether vesicular Ca 2ϩ is required for PS externalization, MEF were incubated in Ca 2ϩ -free medium that was substituted with 2 mM Mg 2ϩ during the time course of the experiment. The rationale for this experimental design was to generate cells that contained Ca 2ϩ -free endocytotic vesicles and trigger apoptosis and to determine the propensity of the apoptotic cells to express PS on their outer membrane leaflet. MEF were labeled with Rhod-2AM at 37°C for 1 h, and the distribution of fluorescence was monitored before and after triggering apoptosis. The data presented in Fig. 5 indicate that, in contrast to control cells incubated in the presence of extracellular Ca 2ϩ (Fig. 5C), the Rhod-2-containing vesicles in cells incubated in the absence of Ca 2ϩ did not fluoresce, suggesting that these vesicles were indeed Ca 2ϩ -free (Fig. 5A). Upon apoptosis ( Fig. 5E and analysis of DNA fragmentation, not shown), however, both populations exhibited cytosolic Rhod-2 fluorescence (Fig. 5, B  and D). Because the Mg 2ϩ -incubated cells did not fluoresce before the initiation of apoptosis (Fig. 5A), the source of Rhod-2 fluorescence was likely from increased cytosolic Ca 2ϩ that originated from organelles other than endocytotic vesicles. Importantly, although the Mg 2ϩ -incubated cells were apoptotic, as assessed by activation of caspases (Fig. 5E), PS externalization was inhibited (Fig. 5B). On the other hand, the Ca 2ϩ -incubated cells redistributed both Ca 2ϩ from the endocytotic vesicles to the cytosol (Fig. 5, C versus D, red fluorescence) and PS from the inner-toouter membrane leaflet (annexin V-positive; Fig. 5D). As expected, assessment of cytosolic Ca 2ϩ with Fura2 revealed that lysosomes contributed a significant amount of imported exogenous Ca 2ϩ to the cytosol upon apoptosis (Fig. 5F).
cPLA2␣ Inhibits Release of Lysosomal Ca 2ϩ and PS Externalization-Apoptosis-dependent increases in cytosolic Ca 2ϩ (24,25) are known to activate cytosolic phospholipase A 2 (PLA2) (26), which in turn can lead to lysosomal membrane destabilization (27,28). These results raise the possibility that PLA2 plays a role in regulating the release of lysosomal Ca 2ϩ . To investigate this possibility, cells containing Rhod-2AM-labeled lysosomes were triggered into apoptosis in the presence and absence of the cytosolic phospholipase A 2 inhibitor, cPLA2␣. These cells were then assessed for PS externalization, caspase activation, and cytosolic Ca 2ϩ . Fig. 6 shows that upon apoptosis, cells incubated in the absence of the inhibitor exhibited intense cytosolic Rhod-2 fluorescence, became annexin V-positive, and generated PS-expressing apoptotic blebs ( Fig.  6B and inset). Cells incubated with cPLA2␣, however, appeared to contain less intense cytosolic Rhod-2 fluorescence, did not generate apoptotic blebs, and contained a significantly lower fraction of annexin V-positive cells. To confirm that cPLA2␣ indeed inhibited the release of lysosomal Ca 2ϩ , duplicate cultures were assessed for cytosolic Ca 2ϩ with Fura-2AM. Fig. 6F shows that in the presence of cPLA2␣, cytosolic Ca 2ϩ levels were reduced by ϳ40%. Importantly, cPLA2␣ had no effect on apoptosis as assessed by caspase activation (Fig. 6E and DNA fragmentation, not shown).

DISCUSSION
Apoptosis is a highly regulated process that involves multiple upstream checkpoints that control a variety of sequentially activated pathways. For example, the release of Ca 2ϩ from the ER and its subsequent uptake into mitochondria initiate organelle remodeling that results in loss of mitochondrial membrane potential (16 -18, 29), opening of the permeability transition pore, and release of factors that trigger apoptosis. Clearly, Ca 2ϩ homeostasis and alterations in amplitude and its spatial and temporal patterns play a central role in regulating cell death. Ca 2ϩ has also been shown to play a critical role in the movement of PS from the inner-to-outer bilayer leaflet of the cells. Unlike the complex Ca 2ϩ transients involved in apoptosis, PS externalization only requires temporal elevations in cytosolic [Ca 2ϩ ]. Because PS externalization is one of the earliest consequences of apoptosis that precedes cytochrome c release and caspase activation, we initiated an investigation into the source of Ca 2ϩ that might regulate the redistribution of PS between bilayer leaflets.
The data presented in Fig. 1 indicate that PS externalization is associated with elevation of cytosolic Ca 2ϩ . However, the TG data indicate that PS externalization requires not only elevated cytosolic Ca 2ϩ levels but that these levels need to persist to initiate the transbilayer redistribution of PS. Similarly, treatment of cells with apoptosis-inducing agents resulted in concomitant increases in both cytosolic Ca 2ϩ and cell surface PS (Fig. 2).
Preliminary experiments to determine optimal labeling of MEF mitochondria with Rhod-2AM revealed that organelle specificity was lost when the probe was incubated with the cells at 37°C (Fig. 3A). Organelle-specific probes showed that the Rhod-2 colocalized with both endosomes and lysosomes (Fig. 3,  F and G). Because the fluorescence of Rhod-2 is directly proportional to [Ca 2ϩ ], these organelles likely contained relatively high [Ca 2ϩ ]. Indeed, endocytotic pathways can provide a significant amount of cellular calcium because uptake of the ion takes place through pinocytosis of the extracellular medium (30). Using rhodamine-labeled dextran in combination with the cellimpermeable Ca 2ϩ probe, CG1, we showed that upon apoptosis, both probes redistributed independently (Fig. 4). Although the rhodamine-dextran was retained within the vesicles, the redistribution of CG1 fluorescence indicated that both Ca 2ϩ and CG1 were transferred from the vesicles to the cytosol.  Taken together, these data are consistent with endocytosed Ca 2ϩ being released from the lumen of lysosomes and used to mediate membrane events that regulate PS externalization.
To directly determine the potential contribution of vesicular Ca 2ϩ to PS externalization, the propensity of MEF containing Ca 2ϩ -free and Ca 2ϩ -containing vesicles to undergo cell death and externalize PS was determined. This key experiment demonstrated that although the Ca 2ϩ present in intracellular organelles (ER and mitochondria) is sufficient for apoptosis, the spatial redistribution of vesicular Ca 2ϩ to the cytosol was critical for PS externalization (Fig. 5).
Because cytosolic PLA2 is known to affect lysosome stability (27,28), it is possible that activated PLA2 promotes the release of lysosomal Ca 2ϩ . To test this, MEF were triggered into apoptosis in the absence and presence of the PLA2 inhibitor cPLA2␣. The data presented in Fig. 6 indicate that inhibition of PLA2 reduced the level of cytosolic Ca 2ϩ by ϳ40% but completely inhibited PS externalization. Because cytosolic Ca 2ϩ levels were not completely inhibited (Fig. 6F) and the Rhod-2labeled lysosomes remained intensely fluorescent (Fig. 6D), it is possible that the increase observed was due to release from other organelles. Because PLA2 is activated with Ca 2ϩ , these data raise the possibility that Ca 2ϩ released from other organelles during apoptosis activates the enzyme, which in turn destabilizes the lysosomes to release additional Ca 2ϩ to levels that are sufficient for PS externalization.
In summary, the data presented here indicate that although apoptosis and PS externalization are both regulated through a common cytosolic Ca 2ϩ -dependent pathways, PS externalization requires higher cytosolic [Ca 2ϩ ] that is only achievable through the contribution of imported vesicular Ca 2ϩ , which is released to the cytosol in an apoptosis-dependent manner. Although these studies addressed the source of Ca 2ϩ , the actual mechanism by which increased cytosolic Ca 2ϩ redistributes PS between bilayer leaflets remains elusive. Our data do, however, establish the concept that a vesicular Ca 2ϩ -dependent signaling mechanism is primarily responsible for the regulation of PS externalization during apoptosis.