Sphingomyelinase (SMase; sphingomyelin phosphodiesterase, SMPD; EC 3.1.4.12) hydrolyzes sphingomyelin (SM) to phosphocholine and ceramides, which play a central role in sphingolipid metabolism (Hannun and Obeid, 2018). The SMase family can be largely divided into four major subfamilies based on the optimal pH and the requirement of Mg²⁺ ion for enzymatic activity: the acid SMase (A-SMase, encoded by SMPD1), the alkaline SMase (encoded by ENPP7), the Mg²⁺-dependent neutral SMases (nSMase1, nSMase2, nSMase3, and mitochondria-associated [MA] nSMase encoded by SMPD2, SMPD3, SMPD4, and SMPD5, respectively), and the Mg²⁺-independent neutral SMase (iSMase) (Wu et al., 2010a). Despite several purification attempts, iSMase (Lee et al., 2011; Okazaki et al., 1994; Yamaguchi and Suzuki, 1978) has not yet been identified.

Ceramides have various effects on cellular processes, including cell death, differentiation, cell cycle arrest, senescence, autophagy, and insulin resistance (Hannun and Obeid, 2018). In cell death signaling—particularly the intrinsic apoptosis pathway—ceramides enhance the mitochondrial outer membrane permeabilization (MOMP) and the release of cytochrome c from the intermembrane space (IMS) to the cytoplasm, triggering apoptosome formation and caspase-9 activation (Masamune et al., 1996; Obeid et al., 1993; Paris et al., 2001; Wiesner and Dawson, 1996). Mechanistically, ceramides self-assemble in the mitochondrial outer membrane to form large and stable channels with diameters ranging from 5-40 nm, releasing IMS proteins, including cytochrome c, to the cytosol (Colombini, 2010; Ganesan and Colombini, 2010; Samanta et al., 2011). Bacterial SMase targeting the mitochondria-generated ceramides within the mitochondria, released cytochrome c, and induced apoptosis (Birbes et al., 2001). nSMase2 and A-SMase may be involved in stress-induced ceramide production and apoptosis (Galadari et al., 2015). However, these enzymes are not expressed in the mitochondria (Hannun and Obeid, 2018). MA-nSMase has been identified as an enzyme expressed in the mitochondrial outer membrane and endoplasmic reticulum but does not mediate apoptosis (Rajagopalan et al., 2015; Wu et al., 2010b). Thus, identifying the bona fide mitochondrial SMase for MOMP is crucial for understanding the exact mechanism underlying apoptosis.

In this study, we purified a mitochondrial Mg²⁺-independent SMase (mt-iSMase) from rat brain using the Percoll gradient system, a biotinylated SM (biotin-SM) pull-down method, and Mono Q column chromatography. mt-iSMase is localized in the IMS of mitochondria, functions at an optimum pH of 6.5, and is inhibited by GS4869, dithiothreitol (DTT) and cations including Mg²⁺, Mn²⁺, Ni²⁺, Cu²⁺, Zn²⁺, Fe²⁺, and Fe³⁺ ions. This enzyme was the first iSMase isolated from mitochondria. Given the localization of the IMS, we hypothesize that mt-iSMase plays a critical role in MOMP for cytochrome c release and apoptosis.

Materials

[N-methyl-¹⁴C] SM (52 mCi/mmol) was purchased from PerkinElmer (USA). Biotin-SM was purchased from Echelon Biosciences Inc. (USA). Streptavidin-Sepharose High-Performance Beads were purchased from GE Healthcare Bio-Sciences AB (Sweden). Percoll and Mono Q anion exchange columns and Superose 6 and 12 gel filtration columns were purchased from Amersham Biosciences (UK). Anti-nSMase2, anti-SMase1, and anti-SNAP25 mouse monoclonal antibodies and anti-c-Myc antibodies were purchased from Santa Cruz Biotechnology (USA). Anti-Hsp60 mouse monoclonal antibody was purchased from Stressgen Biotechnologies (USA). Anti-VDAC and anti-COX IV rabbit monoclonal antibodies were purchased from Cell Signaling Technology (USA). Ethanol (absolute GR for analysis) was obtained from Merck (Germany). Purified Mouse Anti-cytochrome c antibody was purchased from BD Biosciences (USA). A Cytosol/Mitochondrial Fractionation kit was purchased from Calbiochem (Germany). Mini-PROTEAN^® TGX^TM Gels (4%-15%, 15-well comb, 15 μl) were purchased from Bio-Rad Laboratories (USA). Unless otherwise stated, all reagents used in this study were of the highest purity and were purchased from Sigma-Aldrich (USA). Male Sprague–Dawley rats (5 to 7 weeks old) were purchased from Orient Bio (Korea). All animal experiments were approved by the Chung-Ang University Institutional Animal Care and Use Committee (IACUC No. 201700061).

Isolation of intact mitochondria

The synaptosome and mitochondria were isolated from rat brains by centrifugation using the Percoll gradient method according to a previously described protocol with minor modifications (Sims and Anderson, 2008). The brain was removed immediately (<1 min) from the euthanized adult male rats, rinsed twice with 15 ml isolation buffer (320 mM sucrose, 1 mM EDTA, 5 mM Tris-HCl [pH 7.4], 0.25 mM DTT), and cooled. The forebrain was separated from the cooled brain and minced into small pieces using a sterile razor. After weighing (approximately 1.1-1.2 g), the minced whole forebrain was homogenized with nine volumes of isolation buffer (approximately 10-11 ml) on ice using 10 up-and-down strokes with the Potter–Elvehjem tissue grinder. The homogenate was centrifuged at 1,300 × g for 3 min at 4°C. The supernatant (S1.3) was collected, and the pellet was re-homogenized with the same volume of fresh isolation buffer as that used for whole forebrain homogenization. The pellet homogenate was centrifuged at 1,300 × g for 3 min at 4°C, and the supernatant was collected and pooled. The pooled S1.3 was centrifuged at 21,000 × g for 10 min at 4°C, and the pellet (P21) was retained. P21 was then resuspended in a 15% Percoll solution. Percoll gradient solutions for mitochondrial isolation were prepared by diluting Percoll to various densities with an isolation buffer, adjusting the pH to 7.5, and slowly layering above the 23% and 40% Percoll solutions. This material was centrifuged at 30,700 × g for 5 min at 4°C in a fixed-angle rotor. The material accumulating at the top of the gradient was carefully removed. The material accumulating around 15%-23% Percoll solution (the synaptosome-rich fraction) was collected using a Pasteur pipette, and the material accumulating around the 23%-40% Percoll solution (the mitochondrial fraction) was collected and transferred to a fresh tube. Isolation buffer that was thrice the volume of the mitochondrial fraction was slowly added; the fraction was gently suspended by stirring with a pipette tip (no pipetting). The resuspended mitochondrial fraction was centrifuged at 16,700 × g for 10 min at 4°C, and the supernatant was carefully discarded. Then, 0.5 ml of BSA (10 mg/ml) was slowly added to the soft pellet, and the mixture volume was made up to 3 ml with the isolation buffer. The mixture was slowly resuspended by stirring with the pipette tip and centrifuged at 6,900 × g for 10 min at 4°C. The pellet was used to isolate intact mitochondria and was resuspended in the isolation buffer or other homogenization buffers depending on the conditions required for further experiments.

mt-iSMase purification using column chromatography

Intact mitochondria were resuspended in cold buffer A (25 mM Tris-HCl [pH 7.5], 1 mM EDTA) containing 0.1% Triton X-100 and disrupted using a sonicator (Sonic & Materials, USA) on ice. The debris was removed by centrifugation at 6,900 × g for 10 min. The solubilized mitochondria were incubated at 4°C overnight, followed by centrifugation at 100,000 × g for 1 h at 4°C. The resulting supernatant was pooled and applied to a Mono Q anion exchange fast protein liquid chromatography (FPLC) column (5.0 cm × 5.0 mm; Pharmacia LKB, Sweden) that was previously equilibrated with Buffer A containing 0.1% Triton X-100. The proteins bound to the column were eluted at a flow rate of 0.5 ml/min using a 30 ml linear gradient of Buffer A containing 1 M NaCl and 0.01% Triton X-100. Aliquots (10 μl) of the applied sample (8 ml), flow-through (8 ml), and each fraction (1 ml) were assayed for Mg²⁺-independent SMase activity. The active fractions or the flow-through were collected, concentrated using Centricon^®, and applied to a Superose 6 gel filtration column that was pre-equilibrated with Buffer A containing 0.25 M NaCl and 0.01% Triton-X 100. Aliquots (10 μl) of the applied sample (1 ml) and each fraction (0.5 ml) were assayed for Mg²⁺-independent SMase activity.

SMase activity assay

The substrate [N-methyl-¹⁴C] SM (labeled with ¹⁴C on the choline moiety) was dried under a nitrogen stream and resuspended in ethanol. The standard reaction mixture (100 μl) for assaying Mg²⁺-independent SMase activity contained 5 mM EDTA, 2.5 μM [N-methyl-¹⁴C] SM (approximately 30,000 cpm), 2 mM sodium deoxycholate, and 100 mM Tris–HCl (pH 7.5) or 200 mM imidazole-HCl (pH 6.5). For the Mg²⁺-dependent SMase assay, 10 mM MgCl2 was used instead of 5 mM EDTA. The reactions were performed at 37°C for 10-30 min and stopped by adding 320 μl of chloroform/methanol (1:1 volume) and 30 μl of 2 M HCl, according to the method described earlier (Bligh and Dyer, 1959). After vortexing and centrifugation, 200 μl of the clear aqueous phase was added to a 2.5 ml scintillation solution (Insta-Gel Plus; PerkinElmer), and the radioactivity was determined using a PerkinElmer Tri-Carb Liquid Scintillation Counter.

To evaluate the enzyme characteristics at various concentrations of the nSMase inhibitor GW4869, the active fractions of Superose 6 or whole brain lysate were incubated in the absence or presence of GW4869, as described previously (Marchesini et al., 2003). After pre-incubation of the enzyme sources at 37°C in the absence or presence of GW4869, the substrate ([¹⁴C] SM) was added, and SM hydrolysis was determined as described earlier.

pH characterization

pH characterization was performed using a standard reaction system with modifications. The following buffers were applied to the standard reaction to adjust the pH: 100 mM acetate–HCl (pH 4.0-5.5), 100 mM MES–HCl (pH 5.7-6.3), 200 mM imidazole–HCl (pH 6.5), 100 mM Tris–HCl (pH 7.0-8.5), and 100 mM glycine–NaOH (pH 9.0-10.0).

Characterization for mt-iSMase for cations

To characterize calcium dependency and remove other cations, an aliquot (0.5 ml) of each active fraction was applied to a PD-10 desalting column (Pharmacia LKB) that was pre-equilibrated with 25 mM Tris–HCl (pH 7.5) containing 0.01% Triton X-100. Mg²⁺-independent SMase activity of the desalted fractions was assayed at various Ca²⁺ concentrations. The absolute concentration of free Ca²⁺ was calculated using an equation based on the stability constant of the EGTA/CaCl2 system. Likewise, Mg²⁺-independent SMase activity was assayed in the presence of various cationic concentrations (Mg²⁺, Ni²⁺, Mn²⁺, Cu²⁺, Zn²⁺, Fe²⁺, and Fe³⁺). After pre-incubation at 37°C in the absence or presence of the indicated cationic concentrations, the substrate ([¹⁴C] SM) was added, and SM hydrolysis was determined as described earlier.

Enzyme kinetics assay

To determine the kinetics of mt-iSMase, the active fraction of Superose 6 was incubated with the indicated concentrations of SM in a standard reaction system. The concentration of [¹⁴C] SM was fixed at 2.5 μM, and the concentration of non-radioactive SM (cold SM) was increased from 0 to 150 μM. As described earlier, SM hydrolysis was determined as counts per minute and converted to specific activity (nmol/mg/h) and total SM hydrolysis. Vmax and Km were calculated from the kinetic data using SigmaPlot software.

Immunoprecipitation

The mitochondrial and synaptosomal fractions isolated from the rat brains using the Percoll gradient isolation method were used for immunoprecipitation. Each organelle fraction and anti-Hsp60 or anti-c-Myc antibody were mixed and vortexed mildly at 4°C for 2 h. After incubation, 50 μl of Protein A Sepharose bead was added to the immunoprecipitation mixture and vortexed mildly at 4°C overnight. The immunoprecipitated pellet was washed with five volumes of buffer A and resuspended in one volume of buffer A. Aliquots from the immunoprecipitated pellet or supernatant were assayed for SMase activity in the presence or absence of MgCl2.

Western blotting analysis

The subfractions or purified fractions were mixed with an aliquot of Laemmli buffer (0.125 M Tris–HCl [pH 6.8], 4% sodium dodecyl sulfate [SDS], 20% glycerol, 10 mM mercaptoethanol, and 0.002% bromophenol blue). After boiling for 5 min, the samples were cooled to room temperature, separated using a 15% SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to a nitrocellulose membrane (Schleicher & Schuell, USA) at 30 mA constant current. The membrane was blocked for 1 h at room temperature with 5% nonfat milk in Tris-buffered saline (20 mM Tris–HCl [pH 7.4], 137 mM NaCl, and 2.7 mM KCl). The blocked nitrocellulose membranes were incubated with rabbit antibodies against nSMase2, SNAP25 (diluted 1:2,000; Santa Cruz Biotechnology), Hsp60 (diluted 1:2,500; Stressgen), and cytochrome c (diluted 1:1,000; BD Biosciences) at 4°C with constant shaking. Antibody-binding sites were detected using an alkaline phosphatase-conjugated IgG antibody (Santa Cruz Biotechnology) and a chromogenic substrate (1-Step^TM NBT/BCIP; Pierce, USA).

Mitochondrial subfractionation

The mitochondrial subfractionation method was modified from that reported in a previous study (Rardin et al., 2009). The isolated intact mitochondrial pellet (2 mg) was resuspended with 1 ml of hypotonic buffer (10 mM KCl, 2 mM HEPES [pH 7.2]) and incubated with mild vortexing at 4°C for 20 min. The incubated sample was added to 0.33 ml of shrink buffer (1.8 mM sucrose, 2 mM ATP, 2 mM MgSO4, and 2 mM HEPES [pH 7.2]) and mixed gently. The mitochondrial mixture was sonicated gently (3% amp, 15 s) on ice. Disrupted mitochondrial outer membrane (MOM) and IMS were separated from non-disrupted mitoplast (mitochondrial inner membrane (MIM) and matrix) by centrifugation at 12,000 × g for 10 min at 4°C. The mitoplast pellet (MP) was completely disrupted by a sonicator with 0.1% Triton X-100. Each membrane fraction (MOM and MIM) was separated from the soluble fraction (IMS and matrix) by centrifugation at 100,000 × g for 1 h at 4°C. iSMase activity in each sub-fraction was measured in equal volumes (each 1/200 of total volume) for 30 min, and equal amounts of each sub-mitochondrial fraction were immunoblotted with the following markers for each fraction: VDAC for MOM, cytochrome c for IMS, COX IV for MIM, and Hsp60 for the matrix.

Pull-down of mt-iSMase using biotin-SM

Isolated mitochondrial lysates or purified fractions were applied to a pull-down assay using the biotinylated substrate biotin-SM (Echelon-Inc., USA). Each enzyme-enriched fraction was mixed with 10 μM of biotin-SM in a 200 μl standard assay system (100 mM Tris–HCl [pH 7.5], 5 mM EDTA, and 2 mM SDC). The mixture was incubated for 1 h at 4°C with constant shaking. After incubation, 50 μl of Streptavidin–Sepharose beads in a standard assay buffer system was added to the mixture. The precipitation mixture was mixed thoroughly and incubated for 30 min at 4°C with constant shaking. After additional incubation, the supernatant was discarded after centrifugation at 700 × g for 1 min at 4°C. The precipitated mixture was washed four times with 10 volumes of phosphate-buffered saline (PBS) and once with 10 volumes of the standard assay buffer system and resuspended in the standard assay buffer system. To solubilize mt-iSMase from the pull-down pellets, 0.005% Triton X-100 was added to the resuspended pull-down pellets. The mixture was thoroughly mixed and incubated for 1 h at 4°C with constant shaking. After incubation, the supernatant was collected by centrifugation at 700 × g for 1 min at 4°C. Aliquots from the pull-down pellets or supernatants of each process were assayed for SMase activity in the presence or absence of magnesium, subjected to SDS-PAGE, and visualized using immunoblotting or silver staining. nSMase2 was detected using an anti-nSMase2 monoclonal antibody (Santa Cruz Biotechnology). All steps in the pull-down assay were performed in a temperature controlled room (3°C-5°C) or on ice.

Transmission electron microscopy

To determine the purity of the mitochondria isolated using the Percoll gradient method, the isolated mitochondrial pellets were washed twice by centrifugation with an isolation buffer at 6,900 × g for 10 min at 4°C. After the supernatant was discarded, the intact mitochondrial pellets were fixed with 2.5% glutaraldehyde in PBS and incubated overnight at 4°C. The mitochondrial specimen was dehydrated in a graded ethanol series of 60%, 70%, 80%, 90%, and 100% for 30 min each. The dehydrated specimen was substituted with propylene oxide and infiltrated with an Epon mixture. Ultrathin sections (60 nm) were prepared using an ultramicrotome (Leica, Germany) and placed on a copper grid for electrostaining. The prepared specimens were viewed and interpreted using the Tecnai^TM F20 (FEI, USA).

Two-dimensional electrophoresis (2-DE) and Sypro Ruby staining

The buffer in the active fractions obtained from Mono Q anionic exchange column chromatography was changed to a reswelling solution containing 7 M urea, 2 M thiourea, 2% of 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS), 1% DTT, and 1% Pharmalyte (pH 3.5-10; Amersham Pharmacia Biotech., UK). The dry strips (pH 4-10 NL, 24 cm; Genomine, Korea) were re-swelled with the sample solution for 12-16 h at 20°C. Isoelectric focusing of the re-swelled strips was performed using a Multuphore II system (Amersham Biosciences) under the following running condition: 150 V for 1 h, 1,000 V for 1 h, 2,000 V for 1 h, 3500 V for 26 h, and a final 96 kVh. The electrofocused strips were equilibrated in an equilibration buffer (50 mM Tris–Cl [pH 6.8], 6 M urea, 2% SDS, and 30% glycerol) containing 1% DTT for 10 min and incubated in an equilibration buffer containing 2.5% iodoacetamide for 10 min. The equilibrated strips were loaded onto SDS-PAGE gels (20 cm × 24 cm, 10%-16%) and separated using a Hoefer DALT 2D system (Amersham Bioscience) at 20°C. After electrophoresis, the gel was fixed in a fixing solution (40% ethanol and 10% acetic acid) for 1 h and rehydrated with 5% ethanol and 5% acetic acid for 30 min (rehydration was repeated thrice). The rehydrated gel was washed with double distilled water (DDW) and stained using the silver staining method and Sypro ruby (Invitrogen, USA) for 1 h. The gel was visualized and digitized using DIVERSITY (Syngene, India) with a Cy3 emission filter. The scanned images were analyzed using the PDQuest software (ver. 7.1; Bio-Rad Laboratories).

Peptide mass fingerprinting (PMF)

For protein identification through PMF, protein spots were excised, digested with trypsin (Promega, USA), mixed with α-cyano-4-hydroxycinnamic acid in 50% acetonitrile/0.1% TFA, and subjected to MALDI-TOF analysis (Microflex LRF 20; Bruker Dalton, USA) as described by Fernandez et al. (1998). The spectra were collected from 300 shots per spectrum over an m/z range of 600-3,000 and calibrated using a two-point internal calibration with trypsin autodigestion peaks (m/z 842.5099 and 2211.1046). The peak list was generated using Flex Analysis 3.0. The thresholds for peak picking were as follows: 500 for the minimum resolution of monoisotopic mass and 5 for S/N. The search program MASCOT, developed by Matrix Science (http://www.matrixscience.com/), was used for protein identification using PMF. The following parameters were used for the database search: trypsin as the cleaving enzyme, a maximum of one missed cleavage, iodoacetamide (Cys) as a complete modification, oxidation (Met) as a partial modification, monoisotopic masses, and a mass tolerance of ± 0.1 Da. Probability scoring was used as the PMF acceptance criterion.

Statistical analysis

A two-sided t-test was used for single pairwise comparisons. One-way ANOVA with Tukey’s post-hoc test was used for multiple comparisons. Values are expressed as the mean ± SEM. All statistical analyses were performed using the Prism 8 software (GraphPad Software, USA).

Activity of mitochondrial mt-iSMase isolated from rat brain

We isolated mitochondria from rat brains using the method developed by Sims and Anderson (2008) with minor modifications (Fig. 1A). We harvested intact mitochondria with ≥95% purity as determined using transmission electron microscopy (Supplementary Fig. S1A). This result was confirmed by the detection of mitochondrial proteins (cytochrome c and Hsp60) using western blotting analysis (Fig. 1B). nSMase1 was discarded with the supernatant after centrifugation at 21,000 × g (S21), and nSMase2 was removed with the pellet after centrifugation at 21,000 × g (P21) using the Percoll gradient isolation (Fig. 1B). The two types of nSMases were not detected in the mitochondrial fractions. The activity of Mg²⁺-dependent SMases was predominant in the fractions collected from whole brain tissue through P21 (Fig. 1C). SMase in the mitochondrial fraction did not require Mg²⁺ ions for activation (i.e., mt-iSMase). mt-iSMase activity was absent in the synaptosomal fraction where nSMase2 was predominant (Fig. 1D, Supplementary Fig. S1B). We obtained similar results using a synaptosome isolation method (Supplementary Figs. S1C and S1D), confirming that mt-iSMase activity was present in the mitochondrial fraction isolated from the rat brain. mt-iSMase activity in mitochondrial fraction decreased to 61% after treatment with 10 μM of the nSMase inhibitor GW4869 (Fig. 1E). GW4869 inhibited Mg²⁺-dependent nSMase activity in S1.3 fraction (Fig. 1F) but did not inhibit enzymatic activity in the mitochondrial fraction (Fig. 1E), implying that GW4869-insensitive and Mg²⁺-dependent nSMase(s) may exist in the mitochondria.

Interestingly, the optimum pH for mt-iSMase activity was 6.5 (Fig. 2A). Maximum activity was observed when buffered with 50 mM imidazole (Supplementary Fig. S2). As previously reported, Mg²⁺-dependent SMase activity in S1.3 had was optimal at pH of 7.5 (Figs. 2B and 2C). Given that the pH of IMS is reported to be 6.88 (Porcelli et al., 2005), mt-iSMase appears to be suitable for enzyme reactions in IMS.

Next, we investigated where mt-iSMase sublocalizes in the mitochondria by performing mitochondrial subfractionation. Mitochondrial subfractionation was verified by immunoblotting with a marker antibody for each fraction (Fig. 3A, upper panel). We analyzed mt-iSMase activity in each fraction such as the MOM, IMS, inner membrane, and the matrix of mitochondria and observed mt-iSMase activity primarily in the IMS (Fig. 3A, lower panel). A previous study reported that nSMase2 directly interacts with Hsp60 via negative regulation (Ahn et al., 2013); hence, we examined whether mt-iSMase binds to Hsp60. Mitochondrial and synaptosomal fractions were immunoprecipitated using anti-Hsp60 and anti-c-Myc antibodies (Fig. 3B). As shown in Fig. 3C, mt-iSMase activity was not detected in the immunoprecipitated pellets of either the mitochondrial or synaptosomal fractions, whereas Mg²⁺-dependent SMase activity was detected when co-precipitated with anti-Hsp60 antibody in the synaptosomal fraction but not in the mitochondrial fraction. These results imply that mt-iSMase from rat brain is localized in the mitochondrial IMS and does not interact with Hsp60.

Purification of mt-iSMase from rat brain mitochondria using column chromatography

The mitochondrial lysate was disrupted using 0.1% Triton X-100 in buffer A, incubated with vortexing, and centrifuged at 100,000 × g to extract mt-iSMase. Almost all mt-iSMase activity was observed in the supernatant (T-S100), which was used as the extracted fraction for subsequent purification (Figs. 4A and 4B). The extracted fraction was applied to a Mono Q anion-exchange FPLC column to purify and identify mt-iSMase from the mitochondria. In total, 37% of mt-iSMase activity was detected in the fraction eluted with the elution buffer, and the specific activity increased 1.8-fold (Fig. 4C, Table 1). The active fractions were pooled, concentrated, and applied to Superose 6 gel filtration. mt-iSMase was eluted as a single peak with an estimated molecular mass of approximately 65 kDa (Fig. 4D) using molecular mass standard proteins (data not shown). The chromatographic process resulted in a 251-fold purified mt-iSMase from the rat brain and 1,300 × g of the supernatant, as summarized in Table 1.

Pull-down purification of mt-iSMase using biotin-SM

To efficiently purify mt-iSMase, we established an SMase purification method using biotin-SM. We validated this method by first purifying nSMase2. Triton X-100 extracts containing nSMase2 from the rat brains (PS100) were obtained as described in the Materials and Methods section (Supplementary Fig. S3A). nSMase2 was competitively inhibited by biotin-SM and biotinylated ceramide (biotin-Cer), although this activity was most efficiently attenuated by cold SM (Supplementary Fig. S3B), implying that biotin-SM bound to nSMase2 as a substrate. Next, PS100 was incubated with biotin-SM, biotin-Cer, and cold SM at 4°C. The mixtures were pulled down using Streptavidin–Sepharose beads. The pellets pulled down with biotin-SM exhibited nSMase activity, which was not observed in the pellets pulled down with cold SM or in the bead-only control (Supplementary Fig. S3C). Moreover, this activity disappeared in the supernatant pulled down with biotin-SM. The nSMase2 band was detected in the pull-down pellets with biotin-SM by immunoblotting with an anti-nSMase2 antibody (Supplementary Fig. S3D). Interestingly, a small amount of nSMase2 was precipitated by pull-down using biotin-Cer (Supplementary Figs. S3C and S3D). Thus, we confirmed that this method could be used to purify SMase.

We applied this pull-down system to purify mt-iSMase. We first deduced that mt-iSMase was inactive at 4°C for least for 2 h (Supplementary Fig. S4A), implying that mt-iSMase should be incubated with biotin-SM at 4°C for less than 2 h. When the mitochondrial fraction was incubated with biotin-SM, biotin-Cer, and cold SM and pulled down with Streptavidin–Sepharose beads, mt-iSMase activity was observed only in the pellets of the biotin-SM group and decreased only in the supernatant of the same group (Fig. 5A). In addition, the pull-down of mt-iSMase by biotin-SM was reduced by cold SM treatment (Supplementary Fig. S4B), confirming that biotin-SM competes with cold SM to bind to mt-iSMase. Interestingly, biotin-Cer failed to pull down mt-iSMase (Fig. 5A), unlike in the case of nSMase2 (Supplementary Figs. S3C and S3D). Although a large number of mitochondrial proteins was removed by the pull-down process, many proteins remained in the pull-down pellets of biotin-SM, and the band for mt-iSMase was difficult to identify on the SDS-PAGE gel (Fig. 5B). Therefore, we used Triton X-100 to elute mt-iSMase from the pull-down beads using biotin-SM. The pull-down pellet was incubated with 0.001%, 0.005%, and 0.01% Triton X-100 for 1 h, followed by centrifugation. At 0.005% Triton X-100, most of the active mt-iSMase was extracted from the pull-down pellet (Figs. 5C and 5D), implying that mt-iSMase dissociated SM from the active site in the presence of Triton X-100. This step increased the specific activity by up to 6-fold (Table 2). The fraction eluted from the pull-down pellet was further subjected to a Mono Q anion-exchange FPLC (Figs. 6A and 6B). This resulted in an 18-fold increase in the specific activity (Table 2). The purification process resulted in a 6,130.5-fold purified mt-iSMase and yielded 6.8% of the S1.3 fraction, as summarized in Table 2.

MALDI-TOF analysis of protein spots correlated with mt-iSMase activity

To identify the purified mt-iSMase, we separated the proteins in the fractions collected using Mono Q FPLC column chromatography using SDS-PAGE (Fig. 6B). Among the several protein bands that had approximately similar molecular masses to that determined using Superose 6 gel filtration, the enzymatic activity paralleled the intensity of the band that was of approximately 65 kDa (Fig. 4D). The proteins around pI 4.5-5.0 were further separated into three spots (which we named “spot I”) using 2-DE and SDS-PAGE (Fig. 6C). Spot I was digested with trypsin and identified using MALDI-TOF. The entire experimental process from purification to MALDI-TOF MS analysis was performed twice independently. However, we did not obtain similar results in either experiment (data not shown).

Characterization of purified mt-iSMase

The characteristics of mt-iSMase were determined from the active fractions obtained using Superose 6 gel filtration. First, the effects of several cations were examined. mt-iSMase activity was not affected by Ca²⁺ addition (Fig. 7A). However, mt-iSMase activity was reduced to approximately 55% at 10 mM Mg²⁺, 40% at 10 mM Mn²⁺, less than 27% at 10 mM Ni²⁺ or 10 mM Cu²⁺, and less than 32% at 1 mM of Zn²⁺, Fe²⁺, and Fe³⁺ (Figs. 7B and 7C). ATP increased the activity of mt-iSMase by 1.7-fold at 5 mM but inhibited it to 27% at 10 mM (Fig. 7C). DTT decreased the mt-iSMase activity in a dose-dependent manner (Fig. 7D), implying that mt-iSMase might be in an oxidized form for optimal function. Finally, the kinetic properties of mt-iSMase were determined. mt-iSMase followed the classical Michaelis–Menten kinetics, displaying an apparent Km of 37.0 ± 5.6 μM and a Vmax of 35.0 ± 1.7 nmol/min/mg of protein (Fig. 7E).

SMase produces ceramides, which are related to apoptosis, senescence, and differentiation (Hannun and Obeid, 2018). In particular, mitochondrial ceramides play a critical role in MOMP, leading to the release of cytochrome c from the IMS into the cytosol (Birbes et al., 2001; Novgorodov et al., 2005; Siskind et al., 2006). Although MA-nSMase was identified as the sole mammalian mitochondrial SMase and is expressed in the mitochondrial outer membrane toward the cytosol, it may not be involved in apoptosis (Rajagopalan et al., 2015). Thus, identifying an apoptosis-related mitochondrial SMase is important for understanding sphingolipid biology and developing therapeutic strategies for related diseases.

In this study, we first isolated pure and intact mitochondria from rat brain using the Percoll density gradient method to purify mitochondrial SMases (Sims and Anderson, 2008). We then purified an mt-iSMase present in the mitochondrial IMS using a pull-down method with biotin-SM and sequential column chromatography. The 2-DE spots that correlated with mt-iSMase activity were not repeatedly identified as single proteins by MALDI-TOF MS analysis. However, we found that mt-iSMase has several unique biochemical properties compared with that of the previously reported forms of SMase. mt-iSMase does not require divalent cations, is optimally activated at pH 6.5, and is located in the mitochondrial IMS. Similarly, A-SMase also does not require Mg²⁺ ions for enzymatic activation but showed maximum activity at pH 5.0 and is expressed either in the lysosomes or is secreted (Brady et al., 1966). nSMase1, nSMase2, sSMase3, and MA-nSMase require Mg²⁺ or Mn²⁺ ions for SM hydrolysis, show maximum activity at pH 7.5, and are membrane-associated (Marchesini et al., 2003; Wu et al., 2010b).

Two other research groups have attempted to identify mammalian iSMases. In 1978, Yamaguchi and Suzuki demonstrated and suggested the existence of an iSMase in the myelin fraction prepared from rat brains, which was optimally activated at pH 7.0 (Yamaguchi and Suzuki, 1978). In 1994, Okazaki et al. observed iSMase activity in HL-60 cells treated with 1α,25-dihydroxyvitamin D3 (Okazaki et al., 1994). The enzymatic activity was present in the cytosolic fraction and was inhibited by the addition of Cu²⁺, Fe²⁺, and Zn²⁺ ions. However, these groups did not successfully identify iSMase, primarily because of its instability. Recently, we characterized iSMase in the brain. We partially purified iSMase from the cytosolic fraction of the bovine brain (Lee et al., 2011). However, as in the previous two groups, the enzymatic activity was unstable during the purification process. Intriguingly, it was moderately inhibited by the addition of GW4869, similar to its effects on mt-iSMase.

In the present study, the mt-iSMase activity was relatively stable during purification. After mitochondrial isolation, the enzymatic activity did not decrease for at least a week. Thus, we prepared the fraction (T-S100) extracted from mitochondria using Triton X-100 for five days (Fig. 4A). This activity did not require antioxidants; however, it was disrupted by the addition of DTT (Fig. 7D). mt-iSMase was only detected in the mitochondrial fraction and not in the fractions from whole-brain homogenates to S21 (Fig. 1C). Certain extramitochondrial substances may mask enzyme activity. This is another important characteristic of mt-iSMase that requires further investigation.

In this study, we developed a novel purification method for mt-iSMase using biotin-SM. mt-iSMase was bound to biotin-SM but did not hydrolyze it at 4°C for at least 2 h (Supplementary Fig. S4A). Pull-down mt-iSMase was extracted from biotin-SM using 0.005% Triton X-100 at 4°C, indicating that mt-iSMase released SM by interacting with Triton X-100. This highly efficient pull-down purification removed abundant proteins, increasing the specific activity by approximately 6-fold (Figs. 5B and 5D, Table 2). Moreover, this method was applied to the purification of nSMase2 (Supplementary Fig. S3), implying that it is universally applicable to the purification of SMases.

GW4849 was originally identified as a non-competitive nSMase2 inhibitor (Luberto et al., 2002; Marchesini et al., 2003). However, it may also inhibit nSMase1 (Menck et al., 2017), indicating that GW4869 is non-specific. In the present study, mt-iSMase activity was inhibited by GW4869 (Fig. 1E). Given that GW4849 has protective effects against cytochrome c release-mediated cell death (Luberto et al., 2002), the cellular and in vivo effects of GW4869 may involve the inhibition of mt-iSMase, especially in the regulation of mitochondrial ceramide levels, which are related to apoptosis. However, further studies are required to confirm these findings.

The newly purified mt-iSMase showed appropriate characteristics for localization to the mitochondrial IMS. First, IMS exhibits an approximate pH of 6.88 (Porcelli et al., 2005), which appears to be suitable for mt-iSMase, which has an optimum pH of 6.5 (Fig. 2A). Second, the ionic strength of IMS is lower than 130 mM (Cortese et al., 1991), which is close to the optimal salt concentration (approximately 100 mM) of mt-iSMase (Supplementary Fig. S2). Third, most IMS proteins are located in oxidized forms through interactions with Mia40, which is an IMS receptor/oxidoreductase (Backes and Herrmann, 2017). mt-iSMase also appeared to be active in the oxidized state, given that it was inactivated by DTT treatment (Fig. 7D). Therefore, mt-iSMase possesses the appropriate characteristics to function in the IMS and possibly to produce ceramides from mitochondrial SM pool during apoptosis.

Mitochondrial ceramides play a crucial role in MOMP and release of cytochrome c from IMS through pore formation in the mitochondrial outer membrane (Birbes et al., 2001). Mitochondrial ceramides may also be involved in BAX translocation and BCL-2 inactivation in the intrinsic apoptosis pathway (Hannun and Obeid, 2018). Moreover, ceramides are related with cell differentiation, cell proliferation, senescence, necrosis, necroptosis, autophagy, mitophagy, cytoskeleton rearrangement, insulin resistance, and cellular metabolism through their interactions with various proteins such as protein phosphatase 2A, protein phosphatase 1 catalytic subunit alpha, kinase suppressor of RAS, protein kinase C, and AKT (Galadari et al., 2015; Hannun and Obeid, 2018; Kim et al., 2022; Lee et al., 2022). Thus, future studies must investigate whether mt-iSMase is involved in ceramide-mediated signals.

The imbalance of Mg²⁺ homeostasis is pathologically associated with neurodegenerative diseases such as Alzheimer’s disease, Parkinson’s disease, and demyelination (Maier et al., 2022). Abnormal changes in mitochondrial Mg²⁺ concentration affect the viability of neuronal cells. Decrease in mitochondrial Mg²⁺ levels due to apoptotic stimuli induces MOMP, the release of cytochrome c, and cell death (Yamanaka et al., 2019). Given that mt-iSMase activity was suppressed in the presence of Mg²⁺ (Fig. 7B), the decrease of mitochondrial Mg²⁺ activates mt-iSMase, leading to ceramide generation and cell death. Whereas, the Mg²⁺ dependency of MA-nSMase might be unrelated to apoptosis.

In summary, we purified mt-iSMase from the mitochondria of rat brain using biotin-SM pull-down and sequential column chromatographic methods with approximately 13.6% yield and 6,130-fold purity. mt-iSMase activity was observed in the mitochondrial IMS. Several biochemical characteristics of mt-iSMase differ from those of previously identified SMases, including Mg²⁺ independence, subcellular localization, pH dependency, and stability. Based on these data, we infer that purified mt-iSMase is a novel SMase. The primary structure and underlying molecular mechanisms of mt-iSMase activity are expected to be elucidated in the future. We believe that this finding would provide insights into the development of therapeutic strategies for neuronal diseases.

This research was supported by the National Research Foundation (NRF) grant funded by the Korea government (MSIT and MOE, grant Nos. NRF-2020R1A2C1102124 and 2019R1A6A1A10072987, respectively).

J.M.C., Y.P., K.H.A., S.K.K., J.H.W., J.M.J., S.Y.J., J.H.L., I.C.S., and Z.F. performed the experiments. J.M.C., Y.P., E.M.J., and D.K.K. prepared the figures and wrote the manuscript. All the authors have read and approved the final version of the manuscript.

The authors have no potential conflicts of interest to disclose.

Purification of mitochondrial Mg²⁺-independent sphingomyelinase (mt-iSMase) from the mitochondria of rat brain using Mono Q FPLC (fast protein liquid chromatography) and Superose 6 gel filtration

Purification of mitochondrial Mg²⁺-independent sphingomyelinase (mt-iSMase) from mitochondria of rat brain using biotinylated sphingomyelin (biotin-SM) pull-down method