Cells synthesize ceramides in the endoplasmic reticulum (ER) as precursors for sphingolipids to form an impermeable plasma membrane. As ceramides are engaged in apoptotic pathways, cells would need to monitor their levels closely to avoid killing themselves during sphingolipid biosynthesis. How this is accomplished remains to be established. Here we identify SMSr (SAMD8), an ER-resident ceramide phosphoethanolamine (CPE) synthase, as suppressor of ceramide-mediated cell death. Disruption of SMSr catalytic activity causes a rise in ER ceramides and their mislocalization to mitochondria, triggering a mitochondrial pathway of apoptosis. Blocking de novo ceramide synthesis, stimulating ceramide export from the ER, or targeting a bacterial ceramidase to mitochondria rescues SMSr-deficient cells from apoptosis. We also show that SMSr-catalysed CPE production, although required, is not sufficient to suppress ceramide-induced cell death and that SMSr-mediated ceramide homeostasis is critically dependent on the enzyme's N-terminal sterile α-motif or SAM domain. These results define ER ceramides as bona fide transducers of mitochondrial apoptosis and indicate a primary role of SMSr in monitoring ER ceramide levels to prevent inappropriate cell death during sphingolipid biosynthesis.
Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Appendix 1 Data availability References Decision letter Author response Article and author information Metrics Abstract Sphingomyelin is a dominant sphingolipid in mammalian cells. Its production in the trans-Golgi traps cholesterol synthesized in the ER to promote formation of a sphingomyelin/sterol gradient along the secretory pathway. This gradient marks a fundamental transition in physical membrane properties that help specify organelle identify and function. We previously identified mutations in sphingomyelin synthase SMS2 that cause osteoporosis and skeletal dysplasia. Here, we show that SMS2 variants linked to the most severe bone phenotypes retain full enzymatic activity but fail to leave the ER owing to a defective autonomous ER export signal. Cells harboring pathogenic SMS2 variants accumulate sphingomyelin in the ER and display a disrupted transbilayer sphingomyelin asymmetry. These aberrant sphingomyelin distributions also occur in patient-derived fibroblasts and are accompanied by imbalances in cholesterol organization, glycerophospholipid profiles, and lipid order in the secretory pathway. We postulate that pathogenic SMS2 variants undermine the capacity of osteogenic cells to uphold nonrandom lipid distributions that are critical for their bone forming activity. Editor's evaluation Sphingomyelin synthase 2 (SMS2) is an enzyme located in the Golgi apparatus and the plasma membrane (PM) that mediates the synthesis of sphingomyelin (SM), a critical lipid in the PM. Mutations in SMS2 underlie a rare genetic disorder of bone formation. This useful study shows that the disease mutations cause retention of SMS2 in the ER, producing SM in the wrong place and leading to a disrupted SM/cholesterol gradient in the membranes of the secretory pathway. In addition to highlighting the roles of lipid gradients in cellular signaling pathways, this study also provides cell biologists with new tools to examine lipid localization in cells. https://doi.org/10.7554/eLife.79278.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Eukaryotic membranes consist of complex lipid mixtures, with amounts and ratios of the individual lipids showing marked variations between organelles and membrane leaflets (van Meer et al., 2008; Harayama and Riezman, 2018). Although some rare lipids contribute to organelle function through stereospecific recognition by lipid binding proteins (Di Paolo and De Camilli, 2006), numerous recognition processes on or within organellar bilayers are determined by biophysical membrane properties that result from the collective behavior of the bulk lipids. Particularly striking are the lipid-induced changes in bilayer-thickness, lipid packing density and surface charge that accompany the transition from early to late organelles in the secretory pathway (Bigay and Antonny, 2012; Holthuis and Menon, 2014; Sharpe et al., 2010). These changes are highly conserved across species and provide specific cues for membrane proteins that govern vital processes such as protein secretion and signaling (Bigay and Antonny, 2012; Magdeleine et al., 2016; Zhou and Hancock, 2018). To defend the unique lipid mixtures of secretory organelles against erosion by vesicular transport, cells exploit cytosolic transfer proteins that enable specific lipids to bypass vesicular connections by mediating their monomeric exchange at contact sites between distinct organelles (Wong et al., 2019). Moreover, organelles like the ER harbor membrane property sensors that provide feedback to the lipid metabolic network to preserve their characteristic lipid composition when exposed to stress or metabolic insults (Levental et al., 2020; Radanović et al., 2018). Sterols and sphingomyelin (SM) are prime examples of bulk membrane lipids that are unevenly distributed between secretory organelles (van Meer et al., 2008). Sterols are rare in the ER but abundant in the trans-Golgi and plasma membrane (PM). The bulk of SM is synthesized in the lumen of the trans-Golgi from ceramides supplied by the ER and delivered by vesicular transport to the PM, where it accumulates in the exoplasmic leaflet (Hanada et al., 2003). SM is the preferred binding partner of cholesterol (Slotte, 2013). About one-third of the total cholesterol pool in the PM is sequestered by SM (Das et al., 2014; Endapally et al., 2019). Besides influencing cellular cholesterol homeostasis, SM contributes to an enhanced packing density and thickening of trans-Golgi and PM bilayers. This, in turn, may modulate protein sorting by hydrophobic mismatching of membrane spans (Munro, 1995; Quiroga et al., 2013). Moreover, an asymmetric distribution of SM across late secretory and endolysosomal bilayers is relevant for an optimal repair of damaged organelles. Lysosome wounding by chemicals or bacterial toxins triggers a rapid Ca2+-activated scrambling and cytosolic exposure of SM (Ellison et al., 2020; Niekamp et al., 2022). Subsequent conversion of SM to ceramides by neutral SMases on the cytosolic surface of injured lysosomes promotes their repair, presumably by driving an inverse budding of the damaged membrane area in a process akin to ESCRT-mediated formation of intraluminal vesicles. This SM-based membrane restoration pathway functions independently of ESCRT and may also operate at the PM (Niekamp et al., 2022). SM biosynthesis in mammals is mediated by SM synthase 1 (SMS1) and SMS2. Both enzymes act as phosphatidylcholine (PC):ceramide phosphocholine transferases, which catalyze the transfer of the phosphorylcholine head group from PC onto ceramide to generate SM and diacylglycerol (DAG) (Huitema et al., 2004). SMS1 resides in the trans-Golgi, and its deficiency in mice causes mitochondrial dysfunction and disrupts insulin secretion (Yano et al., 2013; Yano et al., 2011). SMS2 resides both in the trans-Golgi and at the PM. Its deficiency ameliorates diet-induced obesity and insulin resistance (Kim et al., 2018; Li et al., 2011; Mitsutake et al., 2011; Sugimoto et al., 2016). Removal of SMS1 or SMS2 has only a minor impact on ceramide, DAG and SM pools in tissues or cells, and the mechanisms underlying the phenotypes observed in SMS1 and SMS2 knockout mice are not well understood. Besides SMS1 and SMS2, mammalian cells contain an ER-resident and SMS-related protein (SMSr), which displays phospholipase C activity and synthesizes trace amounts of the SM analog ceramide phosphoethanolamine (Murakami and Sakane, 2021; Vacaru et al., 2009). We previously reported that SMS2 is highly expressed in bone and identified heterozygous mutations in the SMS2-encoding gene (SGMS2) as the underlying cause of a clinically described autosomal dominant genetic disorder – osteoporosis with calvarial doughnut lesions (OP-CDL: OMIM #126550) (Pekkinen et al., 2019). The clinical presentations of OP-CDL range from childhood-onset osteoporosis with low bone mineral density, skeletal fragility and sclerotic doughnut-shaped lesions in the skull to a severe spondylometaphyseal dysplasia with neonatal fractures, long-bone deformities, and short stature. The milder phenotype is associated with the nonsense variant p.Arg50*, which gives rise to a truncated but catalytically active enzyme that mislocalizes to the cis/medial-Golgi (T. Sokoya and J. Holthuis, unpublished data). However, the most severe phenotypes are associated with two closely localized missense variants, p.Ile62Ser and p.Met64Arg. Interestingly, these missense variants enhance de novo SM biosynthesis by blocking ER export of enzymatically active SMS2 (Pekkinen et al., 2019). This suggests that OP-CDL in patients with pathogenic SMS2 variants is not due to a reduced capacity to synthesize SM but rather a consequence of mistargeting bulk SM production to an early secretory organelle. How this affects the contrasting lipid landscapes and membrane properties in the secretory pathway remains to be established. In this work, we used genetically engineered cell lines and OP-CDL patient-derived fibroblasts to address the impact of pathogenic SMS2 variants p.Ile62Ser and p.Met64Arg on the lipid composition, transbilayer arrangement, and packing density of early and late secretory organelles. Toward this end, we combined shotgun lipidomics on purified organelles with the application of lipid biosensors and targeted solvatochromic fluorescent probes in live cells. We show that cells harbouring pathogenic SMS2 variants accumulate PM-like amounts of SM in the ER and display a disrupted transbilayer SM asymmetry. This is accompanied by significant imbalances in cholesterol organization and membrane lipid order. We also find that pathogenic SMS2 variants cause marked changes in the ER glycerophospholipid profile, including an enhanced phospholipid desaturation and rise in cone-shaped ethanolamine-containing phospholipids, potentially reflecting an adaptive cellular response to counteract SM-mediated rigidification of the ER bilayer. Our data indicate that pathogenic SMS2 variants profoundly undermine the cellular capacity to uphold nonrandom lipid distributions in the secretory pathway that may be critical for the bone forming activity of osteogenic cells. Results The IXMP motif in SMS2 is part of an autonomous ER export signal The most severe clinical presentations of OP-CDL are associated with the SMS2 missense variants p.I62S and p.M64R, which cause retention of a functional enzyme in the ER (Pekkinen et al., 2019). Ile62 and Met64 are part of a conserved sequence motif, IXMP, which is located 13–14 residues upstream of the first membrane span in both SMS1 and SMS2 (Figure 1A, B). We reasoned that this motif may be part of an ER export signal, which could explain its absence in the ER-resident SMS family member SMSr (Vacaru et al., 2009). To test this idea, we generated FLAG-tagged SMS2 constructs in which Ile62 or Met64 was replaced with Ser or Arg, respectively. Upon their transfection in HeLa cells, the subcellular distribution of the SMS2 variants was determined by fluorescence microscopy using antibodies against the FLAG tag and ER-resident protein calnexin. In agreement with our previous findings (Pekkinen et al., 2019), SMS2I62S and SMS2M64R were each retained in the ER, in contrast to wildtype SMS2, which localized to the Golgi and PM (Figure 1C; Appendix 1—figure 1). We then asked whether the IXMP motif in SMS2 can mediate ER export independently of other sorting information. To address this, we created a FLAG-tagged chimera protein, SMSr-SMS211-77, in which the region linking the N-terminal SAM domain and first membrane span of SMSr was replaced with the IXMP-containing cytosolic tail of SMS2 (Figure 1A). Contrary to SMSr, SMSr-SMS211-77 localized to the Golgi. However, SMSr-SMS211-77 variants in which Ile62 or Met64 was replaced with Ser or Arg, respectively, were retained in the ER (Figure 1D; Appendix 1—figure 1). This indicates that the IXMP motif in SMS2 is part of an autonomous ER export signal. Whether mutation of Ile62 or Met64 disrupts ER export by perturbing a linear sequence motif or the overall fold of the enzyme’s N-terminal cytosolic tail remains to be established. Figure 1 Download asset Open asset SMS2 contains an autonomous ER export signal. (A) Membrane topology of SMS family members and chimeric protein SMSr/211-77. Active site residues are marked in red. The position of a conserved IXMP sequence motif is marked by an arrow. (B) Sequence alignment of the region immediately upstream of the first membrane span (TM1) in vertebrate SMS family members. Note that human SMS2 residues Ile62 and Met64 are part of the IXMP sequence motif, which is conserved in SMS1 and SMS2, but not SMSr, across different vertebrate species. (C) HeLa cells transfected with empty vector (EV) or FLAG-tagged SMS2, SMS2I62S, or SMS2M64R were fixed, immunostained with α-FLAG (green) and α-calnexin (magenta) antibodies, counterstained with DAPI (blue) and imaged by DeltaVision microscopy. (D) HeLa cells transfected with FLAG-tagged SMSr, SMSr/211-77, SMSr/211-77I62S or SMSr/211-77M64R were fixed, immunostained with α-FLAG (green) and α-calnexin (magenta) antibodies, counterstained with DAPI (blue) and imaged by DeltaVision microscopy. Scale bar, 10 μm. Pathogenic SMS2 variants mediate bulk production of SM in the ER Metabolic labeling of patient-derived fibroblasts with 14C-choline showed that missense SMS2 variants p.I62S and p.M64R cause a marked increase in the rate of de novo SM biosynthesis (Pekkinen et al., 2019). To directly test the impact of these pathogenic mutations on the biosynthetic capacity of SMS2, we stably transduced SMS1/2 double knockout (ΔSMS1/2) HeLa cells with doxycycline-inducible expression constructs encoding FLAG-tagged SMS2, SMS2I62S, SMS2M64R or their enzyme dead isoforms SMS2D276A, SMS2I62S/D276A, and SMS2M64R/D276A, respectively. After treatment of cells with doxycycline for 16 hr, SMS2 expression was verified by immunoblot analysis and fluorescence microscopy (Figure 2A, B; Figure 2—source data 1; Appendix 1—figure 2). Next, control and doxycycline-treated cells were metabolically labeled with a clickable sphingosine analogue for 16 hr, subjected to total lipid extraction, click reacted with the fluorogenic dye 3-azido-7-hydroxycoumarin, and analyzed by TLC. This revealed that doxycycline-induced expression of SMS2I62S and SMS2M64R, but not their enzyme-dead isoforms, fully restored SM biosynthesis in ΔSMS1/2 cells (Figure 2C). Figure 2 Download asset Open asset Pathogenic SMS2 variants support bulk production of SM in the ER. (A) HeLa SMS1/2 double KO (ΔSMS1/2) cells transduced with doxycycline-inducible constructs encoding FLAG-tagged SMS2, SMS2I62S, SMS2M64R or their enzyme-dead isoforms (D276A) were grown for 16 hr in the absence or presence of 1 μg/ml doxycycline and then subjected to immunoblot analysis using α-SMS2 and α-Na/K-ATPase antibodies. Wildtype HeLa cells served as control. (B) ΔSMS1/2 cells transduced with doxycycline-inducible FLAG-tagged SMS2M64R were treated as in (A), fixed, immunostained with α-FLAG antibody (green), counterstained with DAPI (blue) and imaged by conventional fluorescence microscopy. Scale bar, 10 μm. (C) Cells treated as in (A) were metabolically labeled with a clickable sphingosine analogue for 16 h, subjected to total lipid extraction, click reacted with the fluorogenic dye 3-azido-7-hydroxycoumarin and analyzed by TLC. (D) SM species in total lipid extracts of cells treated as in (A) were quantified by LC-MS/MS and expressed as mol% of total phospholipid analyzed. Note that the rise in dihydroSM (dhSM, d18:0/16:0) in ΔSMS1/2 cells expressing SMS2I62S or SMS2M64R (SMS2*) is likely due to competition between ER-resident ceramide desaturase (DES1) and SMS2* for dihydroceramide (dhCer, d18:0/16:0), which is synthesized de novo by ceramide synthase (CerS) from dihydrosphingosine (dhSph). Data are average ± SD of three technical triplicates. Figure 2—source data 1 Unprocessed and uncropped image files of the immunoblots. https://cdn.elifesciences.org/articles/79278/elife-79278-fig2-data1-v2.pdf Download elife-79278-fig2-data1-v2.pdf Figure 2—source data 2 Raw data of the quantitative analysis of SM species. https://cdn.elifesciences.org/articles/79278/elife-79278-fig2-data2-v2.xlsx Download elife-79278-fig2-data2-v2.xlsx Quantitative mass spectrometric analysis of total lipid extracts from wildtype and ΔSMS1/2 cells revealed that removal of SMS1 and SMS2 wiped out the entire cellular SM pool and caused a fourfold increase in glycosphingolipid (GSL) levels, consistent with a competition between Golgi-resident SM and glucosylceramide (GlcCer) synthases for ceramide substrate (Figure 2D; Figure 2—source data 2; Appendix 1—figure 3A, B; Appendix 1—figure 3—source data 1). In ΔSMS1/2 cells transduced with pathogenic SMS2I62S or SMS2M64R, addition of doxycycline fully restored the SM pool. This was accompanied by a decrease in GSL levels. Doxycycline-induced expression of SMS2 only partially restored the SM pool, presumably because SMS2, unlike its pathogenic isoforms, has no direct access to ER-derived ceramides and must compete with GlcCer synthase for ceramides delivered to the Golgi. Moreover, ΔSMS1/2 cells expressing SMS2I62S or SMS2M64R contained three- to fourfold higher levels of dihydroceramide (Cer d18:0/16:0) and dihydroceramide-based SM (SM d18:0/16:0) than wildtype or SMS2-expressing ΔSMS1/2 cells (Figure 2D; Figure 2—source data 2; Appendix 1—figure 3B; Appendix 1—figure 3—source data 1), which suggests that ER-resident pathogenic SMS2 variants compete with ceramide desaturase DES1 for dihydroceramide substrate synthesized in the ER. All together, these data indicate that pathogenic SMS2 variants support bulk production of SM in the ER. Lipidome analysis of ER and PM isolated from cells expressing pathogenic SMS2 variants We next asked whether pathogenic SMS2 variants that mediate bulk production of SM in the ER interfere with the ability of cells to generate a SM/cholesterol concentration gradient along the secretory pathway. To address this, we analyzed the lipid composition of ER and PM purified from wildtype or ΔSMS1/2 cells that express either SMS2 or the pathogenic variant SMS2M64R. For purification of the ER, cells were lysed and a post-nuclear supernatant was incubated with an antibody against calnexin (Figure 3A). This was followed by incubation with secondary antibody-conjugated paramagnetic microbeads. For PM isolation, the surface of cells was treated with a non-membrane permeant biotinylation reagent before cell lysis (Figure 4A). A post-nuclear supernatant was then directly incubated with streptavidin-conjugated paramagnetic microbeads. The microbeads were applied to columns packed with ferromagnetic spheres (μMACS columns) and the bound material was eluted after the columns were thoroughly washed. The purity of isolated ER and PM was assessed by immunoblot and lipidome analysis. Figure 3 Download asset Open asset Cells expressing pathogenic variant SMS2M64R accumulate SM in the ER. (A) Workflow for affinity purification of the ER from HeLa cells expressing wildtype or pathogenic SMS2 variants. (B) HeLa wildtype (WT) or ΔSMS1/2 cells transduced with empty vector (EV) or doxycycline-inducible SMS2 or SMS2M64R were treated with doxycycline (1 μg/ml, 16 hr), lysed and used to purify the ER as in (A). Whole cell lysates (WC) and purified ER were subjected to immunoblot analysis using antibodies against SMS2 and various organellar markers. Samples were loaded on an equivalents basis. (C) Lipid composition of whole cell lysates (WC) and ER purified from HeLa wildtype cells (WT) was determined by mass spectrometry-based shotgun lipidomics. Levels of the different lipid classes are expressed as mol% of total identified lipids. CE, cholesteryl esters. (D) Lipid composition of whole cell lysates (WC) and ER purified from cells as in (B) was determined as in (C). (E) Comparative analysis of PC unsaturation and chain length in ER purified from ΔSMS1/2 cells expressing SMS2 or SMS2M64R. The graphs show total numbers of double bonds (left) or carbon atoms (right) in the two acyl chains. All data are average ± SD, n=4. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 by paired t test. Figure 3—source data 1 Unprocessed and uncropped image files of the immunoblots. https://cdn.elifesciences.org/articles/79278/elife-79278-fig3-data1-v2.pdf Download elife-79278-fig3-data1-v2.pdf Figure 3—source data 2 Raw data of the quantitative analysis of lipid species in whole cells or isolated ER. https://cdn.elifesciences.org/articles/79278/elife-79278-fig3-data2-v2.xlsx Download elife-79278-fig3-data2-v2.xlsx Figure 4 Download asset Open asset Lipid composition of the PM of cells expressing wildtype or pathogenic SMS2 variants. (A) Workflow for affinity purification of the PM from HeLa cells expressing wildtype or pathogenic SMS2 variants. (B) HeLa wildtype (WT) or ΔSMS1/2 cells transduced with empty vector (EV) or doxycycline-inducible SMS2 or SMS2M64R were treated with doxycycline (1 μg/ml, 16 h), lysed and used to purify the PM as in (A). Whole cell lysates (WC) and purified PM were subjected to immunoblot analysis using antibodies against SMS2 and various organellar markers. Samples were loaded on an equivalents basis. (C) Lipid composition of whole cell lysates (WC) and PM purified from HeLa wildtype cells (WT) was determined by mass spectrometry-based shotgun lipidomics. Levels of the different lipid classes are expressed as mol% of total identified lipids. (D) Lipid composition of whole cell lysates (WC) and PM purified from cells as in (B) was determined as in (C). All data are average ± SD, n=5. (E) Comparative analysis of SM unsaturation and chain length in PM purified from ΔSMS1/2 cells expressing SMS2 or SMS2M64R. Data on SM unsaturation and chain length in ER purified from ΔSMS1/2 cells expressing SMS2M64R are included as reference. The graphs show total numbers of double bonds (left) or carbon atoms (right) in the sphingoid base and acyl chain. All data are average ± SD, n=5. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 by paired t test. Figure 4—source data 1 Unprocessed and uncropped image files of the immunoblots. https://cdn.elifesciences.org/articles/79278/elife-79278-fig4-data1-v2.pdf Download elife-79278-fig4-data1-v2.pdf Figure 4—source data 2 Raw data of the quantitative analysis of lipid species in whole cells or isolated PM. https://cdn.elifesciences.org/articles/79278/elife-79278-fig4-data2-v2.xlsx Download elife-79278-fig4-data2-v2.xlsx As shown in Figure 3B; Figure 3—source data 1, ER eluates contained calnexin but were devoid of protein markers of the PM (Na/K-ATPase), lysosomes (LAMP1), and mitochondria (pMito60). As expected, ER eluates from cells expressing pathogenic variant SMS2M64R contained readily detectable levels of the protein. In contrast, no traces of SMS2 were found in ER eluates from cells expressing the wildtype protein. As there is no specific lipid marker for the ER, using a lipidomics approach to confirm that pull-down with anti-calnexin antibody indeed isolates the ER is not trivial. However, the ER is known to synthesize ceramides whereas SM is primarily produced in the trans-Golgi and accumulates in the PM. In line with the immunoblot data, ER eluates from wildtype cells displayed a fivefold higher ceramide/SM ratio than total cell lysates (Figure 3C; Figure 3—source data 2). Moreover, ER eluates were largely devoid of lipids that are normally concentrated in mitochondria (cardiolipin, CL), PM (SM, cholesterol), and lysosomes (bis(monoacyl-glycerol)phosphate, here quantified together with the isobaric phosphatidylglycerol and reported as BMP/PG). Immunoblot analysis of the PM eluates revealed that they contain Na/K-ATPase but lack protein markers of the ER (calnexin), lysosomes (LAMP1), and mitochondria (pMito60; Figure 4B; Figure 4—source data 1). As expected, PM eluates from cells expressing wildtype SMS2 contained readily detectable amounts of the protein. On the other hand, PM eluates from ΔSMS1/2 cells expressing the pathogenic variant SMS2M64R were devoid of this protein. Moreover, lipidome analysis of PM eluates revealed significantly elevated levels of lipids that are typically concentrated in the PM (i.e. SM, cholesterol, PS) and a fivefold reduction in the ceramide/SM ratio relative to total cell lysates (Figure 4C; Figure 4—source data 2). Lipids primarily associated with lysosomes and mitochondria (PG, CL) were largely absent. Cells expressing pathogenic SMS2 variants accumulate SM in the ER Using the pull-down approaches described above, we next determined the lipid composition of the ER and PM isolated from wildtype and ΔSMS1/2 cells expressing SMS2 or SMS2M64R. The ER from ΔSMS1/2 cells expressing SMS2 had a lipid composition similar to that of the ER from wildtype cells. In contrast, the ER from SMS2M64R-expressing cells contained sevenfold higher SM levels, that is ~10 mol% SM instead of ~1.5 mol% of all identified lipids (Figure 3D; Figure 3—source data 2). This increase in ER-bound SM was accompanied by a twofold rise in DAG levels and a significant drop in the amount of PC and ceramide, consistent with the presence of a catalytically active SM synthase in the ER. Interestingly, expression of SMS2M64R also led to a marked (1.8-fold) increase in ER-associated PE levels. In contrast, ER levels of cholesterol and other bulk lipids were largely unaffected. However, we noticed that expression of SMS2M64R enhanced unsaturation of bulk phospholipid in the ER, as indicated by a significant rise in di-unsaturated PC at the expense of saturated and mono-unsaturated PC species (Figure 3E; Figure 3—source data 2). PC chain length, on the other hand, was largely unaffected. Strikingly, SMS2M64R expression also caused a sharp increase in ER-bound ceramide-1-phosphate (Cer1P; Figure 3D; Figure 3—source data 2). Moreover, cellular Cer1P levels were essentially abolished in SM synthase-deficient cells, indicating that production of Cer1P is tightly coupled to SM biosynthesis. The PM from ΔSMS1/2 cells expressing SMS2 had a SM content similar to the PM from wildtype cells (~10 mol%). In comparison, the PM from ΔSMS1/2 cells expressing SMS2M64R had a slightly reduced SM content (~8 mol%) even though the total SM content of these cells was considerably higher (Figure 4D; Figure 4—source data 2). PM-associated levels of cholesterol and other bulk lipids did not show any major differences among the various cell lines, except for a lack of SM and elevated PC and GSL content in SMS-deficient cells. The PM from all four cell lines contained significantly elevated levels of saturated PC species in comparison to the ER. In addition, the PM from ΔSMS1/2 cells expressing SMS2M64R contained fourfold higher levels of dihydroSM (represented by SM species with 0 double bonds; Figure 4E; Figure 4—source data 2), consistent with the ER residency of this enzyme. Collectively, these data indicate that pathogenic SMS2 variants disrupt the SM gradient along the secretory pathway and cause substantial changes in the lipid profile of the ER. To confirm that cells expressing pathogenic SMS2 variants accumulate SM in the ER, we next used an engineered GFP-tagged version of equinatoxin II (Eqt) as non-toxic SM reporter in live cells. To enable detection of SM inside the secretory pathway, the reporter was equipped with the N-terminal signal sequence of human growth hormone and tagged at its C-terminus with oxGFP, yielding EqtSMSS (Deng et al., 2016). A luminal Eqt mutant defective in SM binding, EqtSolSS, served as control. When expressed in human osteosarcoma U2OS cells, both EqtSMSS and EqtSolSS showed a reticular distribution that overlapped extensively with the ER marker protein VAPA (Figure 5A). However, upon co-expression with SMS2M64R, EqtSMSS but not EqtSolSS displayed a distinct punctate distribution that coincided with the ER network. EqtSMSS-containing puncta were not observed in cells expressing the enzyme-dead variant SMS2M64R/D276A, indicating that their formation strictly relies on SM production in the ER (Figure 5A). To verify that the EqtSMSS-positive puncta mark ER-resident pools of SM, U2OS cells co-expressing EqtSMSS and SMS2M64R were subjected to hypotonic swelling as described before (King et al., 2020). After incubation for 5 min in hypotonic medium, the ER’s fine tubular network transformed into numerous micrometer-sized vesicles. In SMS2M64R-expressing cells, the membranes of these ER-derived vesicles were extensively labelled with EqtSMSS (Figure 5B). In contrast, in hypotonic cells co-expressing EqtSMSS with SMS2M64R/D276A or EqtSolSS with SMS2M64R, the reporter was found exclusively in the lumen of ER-derived vesicles, indicating that Eqt staining of the ER membrane critically relies on catalytically active SMS2M64R and a SM-binding competent reporter. In agreement with the ER lipidome analyses, these results demonstrate that cells expressing SMS2M64R accumulate bulk amounts of SM in the ER. Moreover, our finding that hypotonic swelling of SMS2M64R-expressing cells transforms the ER-associated punctate distribution of EqtSMSS to a more uniform labeling of the ER bilayer suggests that alterations in membrane curvature and/or lipid packing may affect the lateral organization of Eqt-SM assemblies. Figure 5 Download asset Open asset Luminal SM reporter EqtSMSS enables visualization of an ER-resident SM pool in SMS2M64R-expressing cells. (A) Human osteosarcoma U2OS cells co-transfected with mCherry-tagged VAPA (ER, magenta) and empty vector (control), SMS2M64R or SMS2M64R/D276A and luminal GFP-tagged SM reporter EqtSMSS or its SM binding-defective derivative, EqtSolSS (Eqt, green), were incubated in isotonic medium (100% Optimem) for 5 min and imaged by spinning disc confocal microscopy. (B) Cells treated as in (A) were incubated in hypotonic medium (1% Optimem) for 5 min and then imaged by spinning disc confocal microscopy. Scale bar, 10 μm. Pathogenic SMS2 variants break transbilayer SM asymmetry SM adopts an asymmetric distribution across the bilayers of late secretory and endolysosomal organelles, with the bulk of SM residing in the luminal/exoplasmic leaflet. However, using GFP-tagged EqtSM as cytosolic SM reporter (EqtSMcyto), we found that perturbation of lysosome or PM integrity by pore-forming chemicals or toxins disrupts SM asymmetry by triggering a rapid transbilayer movement of SM catalyzed by Ca2+-activated scramblases (Niekamp et al., 2022). To perform its central task in membr
Abstract SMSr/SAMD8 is an ER-resident ceramide phosphoethanolamine synthase with a critical role in controlling ER ceramides and suppressing ceramide-induced apoptosis in cultured cells. SMSr-mediated ceramide homeostasis relies on the enzyme’s catalytic activity as well as on its N -terminal sterile α-motif or SAM domain. Here we report that SMSr-SAM is structurally and functionally related to the SAM domain of diacylglycerol kinase DGKδ, a central regulator of lipid signaling at the plasma membrane. Native gel electrophoresis indicates that both SAM domains form homotypic oligomers. Chemical crosslinking studies show that SMSr self-associates into ER-resident trimers and hexamers that resemble the helical oligomers formed by DGKδ-SAM. Residues critical for DGKδ-SAM oligomerization are conserved in SMSr-SAM and their substitution causes a dissociation of SMSr oligomers as well as a partial redistribution of the enzyme to the Golgi. Conversely, treatment of cells with curcumin, a drug disrupting ceramide and Ca 2+ homeostasis in the ER, stabilizes SMSr oligomers and promotes retention of the enzyme in the ER. Our data provide first demonstration of a multi-pass membrane protein that undergoes homotypic oligomerization via its SAM domain and indicate that SAM-mediated self-assembly of SMSr is required for efficient retention of the enzyme in the ER.
ABSTRACT Sphingomyelin is a dominant sphingolipid in mammalian cells. Its production in the trans -Golgi traps cholesterol synthesized in the ER to promote formation of a sphingomyelin/sterol gradient along the secretory pathway. This gradient marks a fundamental transition in physical membrane properties that help specify organelle identify and function. We previously identified mutations in sphingomyelin synthase SMS2 that cause osteoporosis and skeletal dysplasia. Here we show that SMS2 variants linked to the most severe bone phenotypes retain full enzymatic activity but fail to leave the ER owing to a defective autonomous ER export signal. Cells harboring pathogenic SMS2 variants accumulate sphingomyelin in the ER and display a disrupted transbilayer sphingomyelin asymmetry. These aberrant sphingomyelin distributions also occur in patient-derived fibroblasts and are accompanied by imbalances in cholesterol organization, glycerophospholipid profiles and lipid order in the secretory pathway. We postulate that pathogenic SMS2 variants undermine the capacity of osteogenic cells to uphold nonrandom lipid distributions that are critical for their bone forming activity.
Lysosomes are vital organelles vulnerable to injuries from diverse materials. Failure to repair or sequester damaged lysosomes poses a threat to cell viability. Here we report that cells exploit a sphingomyelin-based lysosomal repair pathway that operates independently of ESCRT to reverse potentially lethal membrane damage. Various conditions perturbing organelle integrity trigger a rapid calcium-activated scrambling and cytosolic exposure of sphingomyelin. Subsequent metabolic conversion of sphingomyelin by neutral sphingomyelinases on the cytosolic surface of injured lysosomes promotes their repair, also when ESCRT function is compromised. Conversely, blocking turnover of cytosolic sphingomyelin renders cells more sensitive to lysosome-damaging drugs. Our data indicate that calcium-activated scramblases, sphingomyelin, and neutral sphingomyelinases are core components of a previously unrecognized membrane restoration pathway by which cells preserve the functional integrity of lysosomes.
Ceramides are essential precursors of sphingolipids with a dual role as mediators of apoptotic cell death. Previous work revealed that the ER-resident ceramide phosphoethanolamine (CPE) synthase SMSr/SAMD8 is a suppressor of ceramide-mediated apoptosis in cultured cells. Anti-apoptotic activity of SMSr requires a catalytically active enzyme but also relies on the enzyme's N-terminal sterile α-motif or SAM domain. Here, we demonstrate that SMSr itself is a target of the apoptotic machinery. Treatment of cells with staurosporine or the death receptor ligand FasL triggers caspase-mediated cleavage of SMSr at a conserved aspartate located downstream of the enzyme's SAM domain and upstream of its first membrane span. Taking advantage of reconstitution experiments with SMSr produced in a cell-free expression system, specific caspase-inhibitors and gene silencing approaches, we show that SMSr is a novel and specific substrate of caspase-6, a non-conventional effector caspase implicated in Huntington's and Alzheimer's diseases. Our findings underscore a role of SMSr as negative regulator of ceramide-induced cell death and, in view of a prominent expression of the enzyme in brain, raise questions regarding its potential involvement in neurodegenerative disorders.
Abstract Lysosomes are vital organelles vulnerable to injuries from diverse materials. Failure to repair or sequester damaged lysosomes poses a threat to cell viability. Here we report that cells exploit a sphingomyelin-based lysosomal repair pathway that operates independently of ESCRT to reverse potentially lethal membrane damage. Various conditions perturbing organelle integrity trigger a rapid calcium-activated scrambling and cytosolic exposure of sphingomyelin. Subsequent metabolic conversion of sphingomyelin by neutral sphingomyelinases on the cytosolic surface of injured lysosomes promotes their repair, also when ESCRT function is compromised. Conversely, blocking turnover of cytosolic sphingomyelin renders cells more sensitive to lysosome-damaging drugs. Our data indicate that calcium-activated scramblases, sphingomyelin, and neutral sphingomyelinases are core components of a previously unrecognized membrane restoration pathway by which cells preserve the functional integrity of lysosomes.
SM is a fundamental component of mammalian cell membranes that contributes to mechanical stability, signaling, and sorting. Its production involves the transfer of phosphocholine from phosphatidylcholine onto ceramide, a reaction catalyzed by SM synthase (SMS)1 in the Golgi and SMS2 at the plasma membrane. Mammalian cells also synthesize trace amounts of the SM analog, ceramide phosphoethanolamine (CPE), but the physiological relevance of CPE production is unclear. Previous work revealed that SMS2 is a bifunctional enzyme producing both SM and CPE, whereas a closely related enzyme, SMS-related protein (SMSr)/SAMD8, acts as a monofunctional CPE synthase in the endoplasmic reticulum. Using domain swapping and site-directed mutagenesis on enzymes expressed in defined lipid environments, we here identified structural determinants that mediate the head group selectivity of SMS family members. Notably, a single residue adjacent to the catalytic histidine in the third exoplasmic loop profoundly influenced enzyme specificity, with Glu permitting SMS-catalyzed CPE production and Asp confining the enzyme to produce SM. An exchange of exoplasmic residues with SMSr proved sufficient to convert SMS1 into a bulk CPE synthase. This allowed us to establish mammalian cells that produce CPE rather than SM as the principal phosphosphingolipid and provide a model of the molecular interactions that impart catalytic specificity among SMS enzymes.
