d,l-Methadone causes leukemic cell apoptosis via an OPRM1-triggered increase in IP3R-mediated ER Ca2+ release and decrease in Ca2+ efflux, elevating [Ca2+]i

The search continues for improved therapy for acute lymphoblastic leukemia (aLL), the most common malignancy in children. Recently, d,l-methadone was put forth as sensitizer for aLL chemotherapy. However, the specific target of d,l-methadone in leukemic cells and the mechanism by which it induces leukemic cell apoptosis remain to be defined. Here, we demonstrate that d,l-methadone induces leukemic cell apoptosis through activation of the mu1 subtype of opioid receptors (OPRM1). d,l-Methadone evokes IP3R-mediated ER Ca2+ release that is inhibited by OPRM1 loss. In addition, the rate of Ca2+ extrusion following d,l-methadone treatment is reduced, but is accelerated by loss of OPRM1. These d,l-methadone effects cause a lethal rise in [Ca2+]i that is again inhibited by OPRM1 loss, which then prevents d,l-methadone-induced apoptosis that is associated with activation of calpain-1, truncation of Bid, cytochrome C release, and proteolysis of caspase-3/12. Chelating intracellular Ca2+ with BAPTA-AM reverses d,l-methadone-induced apoptosis, establishing a link between the rise in [Ca2+]i and d,l-methadone-induced apoptosis. Altogether, our findings point to OPRM1 as a specific target of d,l-methadone in leukemic cells, and that OPRM1 activation by d,l-methadone disrupts IP3R-mediated ER Ca2+ release and rate of Ca2+ efflux, causing a rise in [Ca2+]i that upregulates the calpain-1-Bid-cytochrome C-caspase-3/12 apoptotic pathway.


Methods.
We confirm that all methods described in this manuscript were carried out in accordance with relevant guidelines and regulations. All experiments followed the University of Calgary's biosafety guidelines.

Determination of surviving cell population. Surviving population of cells (1 × 10 4 cells/well in 96 well
plates) treated with different concentrations of d,l-methadone for 24 h were quantified using Alamar blue assay following the manufacturer's protocol. 2+ ] i , the method described by Grynkiewicz et al. 32 was followed. Briefly, cells (5 × 10 5 ) loaded with 5 μM Fluo-4 AM in Ca 2+ -free KRH buffer at 37 °C for 30 min were washed with Ca 2+ -free KRH buffer. Resting [Ca 2+ ] i was measured every 2 s at 485 nm excitation/530 nm emission using a Shimadzu RF 5301PC spectrofluorometer. F max value was obtained after treatment with 0.02% saponin for 30 s and addition of 2 µM CaCl 2 four times. F min value was measured upon addition of 4  Western blot analysis. Cell lysates were analyzed by 12.5% SDS-PAGE and immunoblotting for PARP1, caspase-3, caspase-12, Bid, calpain 1 and actin. Immunoreactive bands were detected by enhanced chemiluminescence and visualized using the Bio-Rad ChemiDoc Imager at the optimal exposure setup. Ratios of protein bands of interest vs actin were determined after densitometry using the NIH ImageJ 1.61 software.

Measurement of cytosolic cytochrome C level.
Cytosolic and mitochondrial fractions were isolated as described previously 35 . Briefly, cells were harvested by centrifugation at 370×g for 10 min, washed with 10 packed cell volumes of NKM buffer (1 mM Tris-HCl, pH 7.4, 0.13 M NaCl, 5 mM KCl and 7.5 mM MgCl 2 ) and resuspended in 6 packed cell volumes of homogenization buffer (10 mM Tris-HCl, pH 6.7, 10 mM KCl, 0.15 mM MgCl 2 , 1 mM PMSF and 1 mM DTT). Cells were then homogenized using a glass homogenizer (30 strokes), resuspended in 2 M sucrose solution and centrifuged at 1200×g for 5 min. The supernatant was subjected to further centrifugation at 7000×g for 10 min and the resulting supernatant was designated as cytosolic fraction. Pellets containing mitochondria were resuspend in 3 packed cell volumes of mitochondrial suspension buffer (10 mM Tris-HCl, pH 6.7, 0.15 mM MgCl 2 , 0.25 M sucrose, 1 mM PMSF and 1 mM DTT) and centrifuged at 10,000×g for 5 min. Pellets were designated as mitochondrial fraction. The cytosolic and mitochondrial fractions were analyzed by SDS-PAGE and immunoblotting for cytochrome C (cyt C), tubulin and VDAC-1.
Ratios of cyt C vs tubulin or VDAC1 band intensities were determined after densitometry using NIH ImageJ 1.61. Standard deviations of the calculated ratios from three independent sets of experiments were determined.
Statistical analysis. Student's t-test (unpaired, two-tailed) was used. Significance was set at p < 0.05.

Results
d,l-Methadone induces leukemic cell apoptosis through activation of OPRM1. An opioid receptor has been implicated in d,l-methadone-induced leukemic cell apoptosis 5 , but the specific identity of this opioid receptor was not determined. Since we previously found that presence or absence of OPRM1 determines the fate of aLL cells following l-asparaginase treatment, i.e., presence leads to apoptosis while absence leads to survival or resistance 31 , we sought to examine the possibility that OPRM1 is targeted by d,l-methadone to induce leukemic cell apoptosis. To do so, we utilized POETIC2 cells (*) infected with retrovirus carrying a pRS empty vector (*+pRS) or pRS-shOPRM1 (*+pRS-shOPRM1) as model systems. POETIC2 cells are continuously growing leukemia cells established from a 14-year-old patient diagnosed with pre-B aLL 31   d,l-Methadone evokes OPRM1-mediated Ca 2+ release from the ER through the IP3R Ca 2+ channel. We next sought to determine whether the d,l-methadone-induced rise in [Ca 2+ ] i in leukemic cells is due to Ca 2+ release from the ER. To do so, cells loaded with an ER Ca 2+ probe 38 , Mag-Fluo-4 AM, then treated with d,l-methadone were analyzed for ER Ca 2+ release by spectrofluorometric Ca 2+ imaging. As shown in Fig. 3A, d,l-methadone induced Ca 2+ release from the ER but *+pRS-shOPRM1 cells showed reduced ER Ca 2+ release compared to *+pRS cells, indicating inhibition of Ca 2+ release by loss of OPRM1 and, therefore, d,l-methadoneinduced ER Ca 2+ release is mediated by OPRM1. We then tested whether d,l-methadone-induced ER Ca 2+ release occurs through the IP3R and/or the ryanodine receptor (RyR), both of which form Ca 2+ channels in the ER 39 . For this experiment, Mag-Fluo-4 AM-loaded and d,l-methadone-stimulated cells were treated with xestospongin C (XeC, a potent IP3R inhibitor: Fig. 3B) 40 or tetracaine 41 (Tet, a potent RyR inhibitor: Fig. 3C) and analyzed for ER Ca 2+ release. As shown in Fig. 3, XeC ( Fig. 3B) but not Tet (Fig. 3C) inhibited d,l-methadoneinduced ER Ca 2+ release, indicating that such Ca 2+ release is mediated by IP3R and not by RyR.
The rate of Ca 2+ extrusion following d,l-methadone treatment is slower in *+pRS cells compared to *+pRS-shOPRM1 cells. Our next step was to explore the possibility that OPRM1 loss also affects extracellular Ca 2+ influx and intracellular Ca 2+ extrusion following treatment with d,l-methadone. To do so, internal Ca 2+ stores in Fura-2 AM-loaded cells were initially emptied by treatment with TBHQ in Ca 2+ -free/ EGTA-containing buffer. Cells were then treated with d,l-methadone in the presence of TBHQ followed by www.nature.com/scientificreports/ 2 mM (Fig. 4A) or 500 mM Ca 2+ (Supplementary Fig. 3), which was added to the external buffer to initiate Ca 2+ entry. Ca 2+ extrusion was also measured in the presence of TBHQ following a switch to Ca 2+ -free/EGTA-containing buffer. The initial treatment with TBHQ, which causes the release of internal Ca 2+ stores, allows measurement of internal Ca 2+ store capacity, and as shown in Fig. 4A, B, OPRM1-depleted *+pRS-shOPRM1 cells have reduced capacity compared to *+pRS cells. In the presence of d,l-methadone and upon external Ca 2+ addition, there was no difference in Ca 2+ entry between *+pRS and *+pRS-shOPRM1 cells (Fig. 4C) as well as in their rates of Ca 2+ influx as measured by T 1/2 of influx (Fig. 4D). However, upon switching to Ca 2+ -free/EGTA-containing buffer and withdrawal of d,l-methadone but continued presence of TBHQ, the rate of Ca 2+ extrusion was faster (i.e., reduced T 1/2 of efflux) in *+pRS-shOPRM1 cells compared to *+pRS cells (Fig. 4E), indicating that OPRM1 regulates the rate of Ca 2+ extrusion. Thus, the integrated Ca 2+ signals, which correspond to the calculated area under the curve [i.e., from the beginning to the end (back to baseline) of Ca 2+ signal] was reduced in OPRM1deficient *+pRS-shOPRM1 cells compared to control *+pRS cells (Fig. 4F).

d,l-Methadone-induced leukemic cell apoptosis is linked to increased [Ca 2+
] i . Since we found that d,l-methadone, which causes leukemic cell apoptosis, triggers IP3R-mediated ER Ca 2+ release and delays the rate of Ca 2+ extrusion, causing increased [Ca 2+ ] i , we next wished to establish a link between the d,l-methadoneassociated rise in [Ca 2+ ] i and d,l-methadone-induced apoptosis. Leukemic cells were treated with d,l-methadone in the presence or absence of the Ca 2+ chelator, BAPTA-AM, then subjected to flow cytometry analysis after staining with PI and FITC-Annexin V. As shown in Fig. 5, d,l-methadone induced *+pRS cell apoptosis that was inhibited by BAPTA-AM, indicating that d,l-methadone-induced apoptosis is linked to increased [Ca 2+ ] i . Consistent with our results above, loss of OPRM1 in *+pRS-shOPRM1 cells inhibited d,l-methadone-induced apoptosis, which was unaffected by BAPTA-AM. d,l-methadone induces leukemic cell apoptosis by upregulating the Ca 2+ -mediated calpain-1-Bid-cytochrome C-caspase-3/12 apoptotic pathway. Previously, d,l-methadone-induced leukemic cell apoptosis was linked to activation of caspase-9 and -3 as well as downregulation of BCL 6 . zVAD. fmk, the broad-spectrum inhibitor of caspases, almost completely inhibits d,l-methadone-induced leukemic cell apoptosis, indicating that caspases (together with BCL activation) is the main route for apoptosis. Here, we sought to analyze the expression of some of the components of the Ca 2+ -mediated apoptotic pathway in *+pRS leukemic cells treated with d,l-methadone. As shown in Fig. 6A, levels of activated calpain-1, t-Bid, caspase-3 and caspase-12 were elevated in d,l-methadone-treated *+pRS cells, and these were reversed by treatment with BAPTA-AM. A similar pattern was observed in the caspase-3-mediated PARP1 cleavage product. Inhibition of calpain-1 by calpeptin reduced d,l-methadone-induced leukemic cell apoptosis ( Supplementary Fig. 4), indicat- www.nature.com/scientificreports/ ing the involvement of calpain-1 in d,l-methadone-induced leukemic cell apoptosis. As expected, translocation of t-Bid from the cytosol to mitochondria was observed (Fig. 6A, left and right bottom panels). Since activation of these elements of the Ca 2+ -mediated apoptotic pathway induces mitochondrial cytochrome c (cyt C) release into the cytosol, which leads to the activation of downstream caspases and subsequent apoptosis 42,43 , we also examined cytosolic and mitochondrial cyt C levels in *+pRS cells following treatment with d,l-methadone. As shown in Fig. 6B, d,l-methadone caused an increase in cyt C level in the cytosol and corresponding decrease in mitochondria, which was reversed by BAPTA-AM treatment. We also assessed mPTP opening in d,l-methadone-induced leukemic cell apoptosis by calcein-AM staining followed by treatment with CoCl 2 . Calcein-AM is a cell permeable fluorophore that diffuses and gets trapped in all subcellular compartments, including mitochondria. Treatment with cobalt (Co 2+ ) quenches calcein fluorescence in all subcellular compartments except the mitochondrial matrix which is enclosed by a Co 2+ impermeable inner mitochondrial membrane when mPTP is closed. Thus, the ability of Co 2+ to quench mitochondrial calcein fluorescence only when mPTP is open allows determination of open vs closed status of mPTP in the cell. Upon treatment with CoCl 2 , d,l-methadone treatment did not alter calcein fluorescence intensity, indicating that increased mPTP opening is not involved in d,l-methadone-induced apoptosis ( Supplementary Fig. 5). Altogether, our findings support our view that d,l-methadone induces OPRM1-regulated apoptosis by increasing IP3R-mediated ER Ca 2+ release and slowing down intracellular Ca 2+ extrusion, causing a rise in [Ca 2+ ] i and upregulating the calpain-1-Bid-cytochrome C-caspase-3/12 apoptotic pathway.  4 . In xenograft-derived aLL cells, it was reported that d,l-methadone induces apoptosis through activation of opioid GPCRs and subsequent downregulation of cAMP and activation of caspase-9 and -3 5 (Fig. 7A). This was deduced from the observed expression of opioid receptors in leukemic cells and the inhibition of d,l-methadone-induced  www.nature.com/scientificreports/ apoptosis by 3-isobutyl-1-methylxanthine (IBMX), PTX, and naloxone. IBMX is a cAMP phosphodiesterase inhibitor that causes cAMP upregulation; PTX inhibits the inhibitory α subunit (G αi ) of heterotrimeric G αβγ -proteins; and naloxone blocks opioid receptor activation. The indirect approach to ascertain opioid receptors as targets for d,l-methadone may, however, be questioned as the use of opioid receptor expression may be seen as a weak supporting evidence, and the use of naloxone appears to be challenged by reports that naloxone also inhibits toll-like receptor 4 (TLR4) signaling, a prominent regulator of immune cell function 44 .
The current study utilized a stringent knockdown system to evaluate and identify OPRM1 as a specific d,lmethadone target that mediates apoptosis in leukemic cells. While it is believed that opioid receptors signal through G αi/o subunits to inhibit adenylyl cyclase activity, reducing cAMP production and protein kinase A activity 22 , opioids elicit IP3R-mediated Ca 2+ release from the ER via PLC 45,46 . In leukocytes 22,27 , pituitary 28 and neuroblastoma cells 26 , G βγ subunit activation of PLCβ and stimulation of IP3R are associated with opioid-induced   29 . These findings together with our previous discovery that aLL cell expression of OPRM1 leads to apoptosis following l-asparaginase treatment while lack of OPRM1 expression leads to survival or resistance to l-asparaginase 31 , prompted us to investigate the possibility that OPRM1 is targeted by d,l-methadone to induce leukemic cell apoptosis, and that OPRM1 regulation of [Ca 2+ ] i is a critical component of d,l-methadone-induced apoptotic pathway in leukemic cells.
Our studies revealed that resting [Ca 2+ ] i in control *+pRS leukemic cells is increased compared to OPRM1depleted *+pRS-shOPRM1 cells. We established that increased [Ca 2+ ] i in *+pRS cells is linked to the stimulation of apoptosis by d,l-methadone, and that the rise in [Ca 2+ ] i is due to increased IP3R-mediated ER Ca 2+ release and reduced rate of intracellular Ca 2+ extrusion. These conclusions were based on the analysis of cells treated with the Mag-Fluo-4 AM Ca 2+ probe 38 , which distinguishes the release of Ca 2+ from the ER, and with XeC, which inhibits Ca 2+ release via the IP3R Ca 2+ channel 40 , as well as by introduction and withdrawal of extracellular Ca 2+ in the presence of TBHQ, which empties internal Ca 2+ stores 37 , allowing measurement of intracellular Ca 2+ extrusion. These observed Ca 2+ dynamics in *+pRS cells following treatment with d,l-methadone are reversed by loss of Figure 7. Proposed mechanism for OPRM1-mediated d,l-methadone-induced apoptosis in leukemic cells. (A) Previously, stimulation of an unidentified type of opioid receptor (#) by d,l-methadone was shown to induce leukemic cell death via G αi , which blocks adenylyl cyclase activity that in turn reduces [cAMP] i , activating caspase-9 and -3 5 . (B) In the current study, we identify the OPRM1 opioid receptor as a specific d,l-methadone target in leukemic cells. Since activation of opioid receptors has been shown to cause G βγmediated rise in [Ca 2+ ] i via PLC 22,[26][27][28] , we propose that d,l-methadone activation of OPRM1 in leukemic cells causes G βγ -mediated stimulation of PLC, which then triggers a rise in [Ca 2+ ] i through increased IP3R-mediated ER Ca 2+ release and reduced rate of Ca 2+ efflux, causing upregulation of the Ca 2+ -mediated calpain-1-Bid-cyt C-caspase-3/12 apoptotic pathway and subsequent apoptosis. In fact, we demonstrate that OPRM1-regulated Ca 2+ -mediated apoptosis induced by d,l-methadone in leukemic cells is triggered by activation of the Ca 2+ -dependent calpain-1 and associated proteolytic events mediated by cysteine proteases. Our findings are consistent with the model illustrated in Fig. 7B where Ca 2+ -activated calpain-1 targets Bid to generate its truncated form, t-Bid, which translocates to mitochondria where it binds and neutralizes the anti-apoptotic Bcl-2. This allows BAK and BAX to form pores in the outer mitochondrial membrane, causing cyt C release, activation of caspases such as caspase-12 47 and caspase-3 48 , and subsequent apoptosis 42,43,49 . Caspase-3 cleaves PARP1, which drives apoptosis in leukemic cells 50 . Our analysis show that d,l-methadone triggers OPRM1-regulated apoptosis by inducing cleavage of calpain-1, Bid, procaspase-3, procaspase-12 and PARP1, indicating an [Ca 2+ ] i -mediated apoptosis. This was substantiated by the reversal of apoptosis by the Ca 2+ chelator, BAPTA-AM.
In summary, we identify OPRM1 as a novel and specific d,l-methadone target in leukemic cells. Our observations indicate that d,l-methadone activates OPRM1, which triggers an increase in IP3R-mediated ER Ca 2+ release and decrease in the rate of Ca 2+ efflux, causing a rise in [Ca 2+ ] i that upregulates the Ca 2+ -mediated calpain-1-Bid-cyt C-caspase-3/12 apoptotic pathway. We note that although loss of OPRM1 in *+pRS-shOPRM1 cells almost completely inhibited d,l-methadone-induced apoptosis, indicating that d,l-methadone primarily targets OPRM1 to induce leukemic cell death, the fact that BAPTA-AM only partially inhibits d,l-methadoneinduced *+pRS cell apoptosis suggests the co-existence of at least one other branch of the OPRM1-mediated d,l-methadone-induced apoptotic pathway. While still under our separate investigation, we presume that the previously reported d,l-methadone-induced apoptotic pathway that downregulates cAMP is in fact a branch of the OPRM1 apoptotic pathway. Certainly, the mode of action of d,l-methadone requires further investigation if it is to be clinically deliberated as a relevant adjuvant or sensitizer for aLL chemotherapy.