ORMDL in metabolic health and disease

Abstract

Obesity is a key risk factor for the development of metabolic disease. Bioactive sphingolipid metabolites are among the lipids increased in obesity. Obesogenic saturated fatty acids are substrates for serine palmitoyltransferase (SPT) the rate-limiting step in de novo sphingolipid biosynthesis. The mammalian orosomucoid-like protein isoforms ORMDL1-3 negatively regulate SPT activity. Here we summarize evidence that dysregulation of sphingolipid metabolism and SPT activity correlates with pathogenesis of obesity. This review also discusses the current understanding of the function of SPT and ORMDL in obesity and metabolic disease. Gaps and limitations in current knowledge are highlighted together with the need to further understand how ORMDL3, which has been identified as an obesity-related gene, contributes to the pathogenesis of obesity and development of metabolic disease related to its physiological functions. Finally, we point out the needs to move this young field of research forward.

1. Introduction

Metabolic diseases encompass conditions that increase the risk of heart disease, stroke and type 2 diabetes. Obesity is perhaps the most notorious risk factor for metabolic disease and is now considered a global pandemic (Blüher, 2019). Obesity is historically thought to occur following excess accumulation of neutral lipids, such as triglycerides and cholesterol. However, more recently the accumulation of sphingolipids has also been suggested to contribute to obesity (Chaurasia & Summers, 2015; Petersen & Shulman, 2017), and its comorbidities such as type 2 diabetes (Chaurasia et al., 2019; Yun et al., 2020), NAFLD and non-alcoholic steatohepatitis (Simon et al., 2019).

Several seminal studies in obese mice models perturbing key sphingolipid biosynthetic enzymes and reducing bioactive sphingolipid metabolites report beneficial effects on metabolic disease (Chaurasia et al., 2016; Chaurasia et al., 2019; Geng et al., 2015; Ijuin et al., 2022; Raichur et al., 2014; Ravichandran et al., 2019). Most of these findings emerged from inhibition or genetic manipulation of serine palmitoyltransferase (SPT), the enzyme responsible for catalyzing the rate-limiting step in de novo sphingolipid biosynthesis. Recently, it was shown that orosomucoid-like proteins (ORMDLs), of which there are 3 isoforms, form an intrinsic complex with SPT and negatively regulate its catalytic activity (Li et al., 2021; Wang et al., 2021). Intriguingly, ORMDL3 has been identified as an obesity-related gene (Pan et al., 2018). In this review we will cover studies emphasizing the involvement of SPT and ORMDL in metabolic diseases and roles of the bioactive sphingolipid metabolites ceramide and sphingosine-1-phosphate (S1P) in metabolic complications of obesity and metabolic disease. We also discuss new mechanistic insights gained from in vivo studies on modulation of the synthesis and degradation of sphingolipids. Finally, we will highlight several new studies attempting to elucidate the role of ORMDL3 in metabolic diseases.

2. Sphingolipid Metabolism

Cellular signalling and the maintenance of eukaryotic cell membranes depend on the availability of sphingolipids within the body (Hannun & Obeid, 2018). Aberrant sphingolipid biosynthesis is a hallmark of many disease pathologies (Hannun & Obeid, 2018). De novo biosynthesis of sphingolipids begins in the endoplasmic reticulum (ER) with the condensation of an amino acid (predominantly serine) with a fatty-acyl CoA (predominantly palmitoyl-CoA) to form 3-ketosphinganine, catalyzed by SPT (Figure 1). Once formed, 3-ketosphinganaine is rapidly reduced by a reductase to form sphinganine, which is the backbone of all sphingolipids. Sphinganine is then N-acylated by six ceramide synthases (CerS1-6) with affinities for different chain length fatty acyl CoAs, forming dihydroceramides with saturated or monounsaturated fatty acid chains ranging from 14 to 26-carbons in length. Introduction of a 4-5 double bond by dihydroceramide desaturase (Degs1) yields ceramides, which are a central hub of complex sphingolipid synthesis (Merrill, 2002). Highly hydrophobic ceramides are transported from the ER to the Golgi apparatus either by the ceramide transfer protein (CERT) for the formation of sphingomyelin (SM), or to the cis-Golgi by vesicular transport where it is glycosylated by glucosylceramide synthase to glucosylceramide, which is then trafficked further via FAPP2 to form lactosylceramide that is sequentially glycosylated to complex glycosphingolipids. Ceramides can also be produced by the salvage pathway whereby internalized membrane complex sphingolipids reach the lysosomal compartment and are degraded by acidic hydrolases to ceramides and subsequently by ceramidase to sphingosine which can be re-acylated by CerS to reform ceramides in the ER (Hannun & Obeid, 2018). Alternatively, sphingosine can be phosphorylated by sphingosine kinases (SphK1 and SphK2) to yield sphingosine-1-phosphate (S1P). The irreversible catabolism of S1P via S1P lyase on the ER to produce ethanolamine-1-phosphate and hexadecenal is the sole exit pathway for sphingolipid degradation. S1P can be exported from cells to activate five G protein-coupled, sphingosine-1-phosphate receptors (S1PR1-5) in autocrine or paracrine manners. S1P also has intracellular functions (Spiegel & Milstien, 2011) (Figure 1). For example, in the nucleolus it regulates histone acetylation and gene expression (Hait et al., 2009). Even though many of these biosynthetic steps are physiologically important, it is perhaps the activity of SPT and S1P lyase regulating de novo biosynthesis and degradation of sphingolipids, respectively, that are critical for cell function.

Figure 1. Illustration of sphingolipid metabolism.

SPT initiates de novo sphingolipid biosynthesis as described in the text. The ceramide produced is transported to the Golgi for further complex sphingolipid biosynthesis. In the salvage/recycle pathway, internalized sphingolipids reach the lysosome for degradation to sphingosine, which can be converted back to ceramide. Alternatively, sphingosine can be phosphorylated by SphK to S1P. When it reaches the ER, S1P is irreversibly degraded by S1P lyase. S1P can also be exported out of cells to activate S1P receptors. SPT: serine palmitoyltransferase; KDSR: 3keto-dihyrosphingosine reductase; CerS, ceramide synthase; Des, dihydroceramide desaturase.

3. SPT and functions of ORMDL in regulation of sphingolipids

SPT, which catalyzes the rate-limiting step of sphingolipid biosynthesis, is an enzyme complex composed of two large subunits, encoded by Sptlc1 and either Sptlc2 or Sptlc3, and small regulatory subunits. The heterodimers comprised of SPTLC2 harboring the PLP-binding lysine residue and SPTLC1 provides the active site and the catalytic core of the enzyme. In addition, the accessory subunits SPTssa and SPTssb are important for substrate specificity for acyl-CoAs and increase enzymatic activity by reducing the distance between the catalytic PLP-binding lysine residue of SPTLC2 and the location of the substrate acyl-CoA at the membrane (Li et al., 2021, Wang et al., 2021). SPTssa enables the use 16-carbon acyl-CoA, whereas SPTssb incorporates specificity for an 18-carbon acyl-CoA (Han et al., 2009). SPTLC1 is essential for activity, whereas SPTLC2 and SPTLC3 are partly redundant and differ in their enzymatic properties. Whereas SPTLC2 specifically forms C18, C19, and C20 long chain bases while SPTLC3 yields broader range of products (Lone et al., 2020).

The enzymatic activity of SPT is also homeostatically controlled by a family of 3 ORMDL proteins ORMDL1, ORMDL2 and ORMDL3, which are in a complex with SPT and negatively regulate its catalytic activity. ORMDLs are conserved across kingdoms. However, phosphorylation-mediated control of ORM interactions with SPT is unique to prokaryotes, whereas eukaryotic ORMDL1-3 do not contain these phosphorylation sites (Ikushiro et al., 2001; Roelants et al., 2011; Shimobayashi et al., 2013). Recent seminal findings have propelled our understanding of the relationship between the catalytic component SPT and its inhibitory regulatory components ORMDLs in eukaryotes (Li et al., 2021; Wang et al., 2021). Cryo-EM structures depict ORMDL3 at the center of the SPT-ORMDL complex forming direct interactions with the major subunits of SPT via hydrogen bonds, salt bridges and Van Der Walls contacts (Li et al., 2021), thereby stabilizing the complex and blocking binding of the substrate to SPT (Li et al., 2021; Wang et al., 2021) (Figure 2). It was suggested that ORMDLs directly sense the levels of ceramide within the cell (Davis et al., 2019). High cellular levels of ceramides cause conformational changes in N-terminus of ORMDLs to block the acyl-CoA binding tunnel of SPT (Figure 2), thus inhibiting de novo sphingolipid biosynthesis and further sphingolipid accumulation (Li et al., 2021; Wang et al., 2021). The importance of the interactions between ORMDL and SPTLC subunits was highlighted by single SPTLC1 mutations that impaired the interaction with OMRDL leading to increased sphingolipid synthesis and causing amyotrophic lateral sclerosis (Lone et al., 2022).

Figure 2. illustration of the SPT-ORMDL complex.

(A) Structures of ORMDL-SPT complex (PDB 7K0M) from Wang et al 2021. The symmetric dimers are comprised of SPTLC1, SPTLC2, SPTssa and ORMDL3. Insert is a 45° rotation to the left showing the catalytic core. In blue is the N-terminus of ORMDL3 that can cap the catalytic core in response to high concentrations of ceramide. (B) Model of the active SPT complex, two dimers each containing of the two major SPT subunits, SPTLC1 and SPTLC2; the small subunit ssSPTa. High cellular ceramide was suggested to cause conformational changes to enhance binding of ORMDL to the SPT complex. The N terminus of ORMDL projects into the acyl-CoA binding tunnel (shown as a circle), explaining how ORMDL can inhibit SPT activity.

The complete SPT-ORMDL complex is comprised of a symmetrical dimer of protein assemblies of SPTLC1, SPTLC2 (or SPTLC3), SPTssa (or SPTssb) and ORMDL. The entire complex is tethered to the ER via α-helices from SPTLC1, SPTssa and ORMDL, further confirming ORMDL is an authentic component of the complex. Moreover, in the absence of one ORMDL isoform, such as ORMDL3, the remaining two have been shown to compensate for this loss to retain inhibition of SPT (Green et al., 2021b).

What sphingolipid is sensed by the SPT complex remains of great debate. On the one hand, there is very strong evidence that ORMDLs can directly sense cellular ceramide (Davis et al., 2019; Gupta et al., 2015; Siow & Wattenberg, 2012). This was shown by the addition of C6- or C8-ceramide to intact cells or to isolated membranes containing the SPT-ORMDL complex which leads to a reduction in SPT activity. Interestingly, SPT activity was retained in cells pre-treated with the CerS inhibitor FB1 and treated with sphingosine, confirming that ceramide or a downstream metabolite is responsible for homeostatic regulation of SPT. This mechanism was conserved across eukaryotes and prokaryotes, with yeast similarly displaying inhibition of SPT activity in response to phytoceramide (Davis et al., 2019). On the other hand, a recent report suggested that ORMDLs regulate SPT activity by sensing S1P levels (Sasset et al., 2023). In this report, treatment of human umbilical vein endothelial cells with S1P, but also with C16-ceramide significantly decreased serine incorporation. However, the inhibitory effect of C16-ceramide on SPT activity was abolished by the sphingosine kinase inhibitor SKI-II, implying that C16-ceramide must be converted to sphingosine and subsequently phosphorylated to S1P to produce the inhibitory effect. It was also demonstrated that local production of S1P and its extracellular transport via SPNS2 was sufficient to stimulate S1PR1/3 and the AKT pathway to stabilize ORMDL (Sasset et al., 2023). Further investigation is needed to determine whether the S1P sensing mechanism is ubiquitous in eukaryotic cells, or specific to endothelial cells, which are mainly controlled via S1PR1 signalling and are exposed to high levels of S1P (Xiong & Hla, 2014).

Previously, it was shown that an increase in intracellular free cholesterol induces the translocation of ORMDLs out of the ER and targets them to autophagosomes and lysosomal degradation (Wang et al., 2015). Others suggested that constitutive proline hydroxylase-mediated ubiquitination of Pro137 in the C-terminus of OMRDL led to its retro-translocation from the ER to the cytosol for proteasomal degradation, allowing for constant steady-state sphingolipid biosynthesis (Sasset et al., 2023). In this regard, prokaryotic orthologue Orm2 has also been shown to be degraded by comparable ER export and endosome- and Golgi-associated degradation via Dfm1 (Bhaduri et al., 2022; Schmidt et al., 2019). Even though the degradation of ORMDLs control their levels and function, whether the sphingolipid sensed by the SPT-ORMDL complex is involved in ORMDL degradation is still not known.

4. Ormdl3 in asthma pathogenesis and ER stress

Perhaps the most studied disease involving ORMDL is asthma. In a genome wide association study on childhood asthma, Moffatt et al. first identified several single nucleotide polymorphisms in the 17q21 locus that led to elevated levels of the ORMDL3 transcript (Moffatt et al., 2007), which has now been shown to positively correlate with Th2 cytokine levels and asthma severity (Schedel et al., 2015). The pathogenic role for ORMDL in asthma has largely been attributed to its effect on ER-stress in promotion of the unfolded protein response (UPR) and autophagy (Chen et al., 2022; Dang et al., 2017; Li et al., 2021; Ma et al., 2015; Oyeniran et al., 2015). The UPR, an adaptive response triggered by accumulation of unfolded proteins in the ER, is comprised of activation of 3 key signalling cascades: transcription factor 6 (ATF6); eukaryotic translation initiation factor 2α (elF2α) and its phosphorylating enzyme protein kinase R like endoplasmic reticulum kinase (PERK); and inositol requiring enzyme 1 (IRE1) responsible for splicing X-box binding protein (XBP1). The role for ORMDLs in promoting the UPR stemmed from the inhibitory effect of ORMDL3 on the sarcoendoplasmic reticulum calcium ATPase 2b, a downstream target of the ATF6 signalling cascade that reduced ER-mediated calcium release and cross-activation of PERK (Cantero-Recasens et al., 2010). Furthermore, overexpression of ORMDL3 in cells increased ATF6 levels (Li et al., 2021; Miller et al., 2012), an event which has been recapitulated in vivo using ORMDL3 transgenic mice and resulted in increased airway responsiveness characteristic of asthma (Miller et al., 2014). On the other hand, ablation of OMRDL3 in mice decreased expression of XBP1 of the ATF6 branch that correlated with protection from allergic airway disease (Loser et al., 2017). Moreover, ORMDL3 expression is elevated following cigarette smoke-induced human airway smooth muscle cell injury, similarly driving an ER-stress response dependent on ATF6 and PERK activation (Chen et al., 2022). Additional roles for ORMDL3 induction of ATF6 have been indicated in the survival of splenic B cells, whereby increased Beclin-1 was responsible for increasing levels of autophagy and supressing apoptosis (Dang et al., 2017).

The notion that ORMDL can have both beneficial effects to limit de novo sphingolipid biosynthesis, whilst also promoting excess autophagy, may be attributed to the level of ORMDLs present. Indeed, removal of ORMDL allows for the increased formation of ceramide (Debeuf et al., 2019; Song et al., 2022) and conversely, overexpression reduces ceramide biosynthesis in vitro (Kiefer et al., 2015) and in vivo (Debeuf et al., 2019). However, modest increases in ORMDL3 expression have been documented to decrease ceramide production, while more pathophysiological increases in ORMDL3 expression, as seen in allergic asthma, promoted ceramide biosynthesis (Oyeniran et al., 2015), possibly due to activation of the UPR. This may represent an evolutionary conserved protective mechanism for prokaryotic ORMs and eukaryotic ORMDLs. It is plausible that once a threshold of sphingolipid abundance and/or ORMDL activity is reached, there is a switch in ORMDL function to activate the UPR leading to apoptosis.

5. Ceramide and S1P in metabolic disease

Ceramide and S1P have emerged as two key obesogenic sphingolipids, with their dysregulation leading to aberrant tissue and circulating concentrations that have been shown to correlate with rodent and human metabolic disease (Chaurasia et al., 2016; Chaurasia et al., 2019; Geng et al., 2015; Haus et al., 2009; Ito et al., 2013; Kowalski et al., 2013). Although concomitantly correlating with metabolic disease, these two sphingolipids have distinct and sometimes opposing pathophysiological roles (Fang et al., 2019; Juchnicka et al., 2021). Strong correlations have been established in diet-induced obese rodent models and obese human patients between circulating ceramide and obesity (Haus et al., 2009; Turpin et al., 2014). Furthermore, ceramides have been shown to cause a myriad of effects (Figure 3), notably insulin resistance (Chaurasia et al., 2016; Chaurasia et al., 2019), but also other comorbidities, such as hepatic steatosis (Chaurasia et al., 2019), type 2 diabetes mellitus (Chaurasia et al., 2019; Holland et al., 2007), non-alcoholic steatohepatitis (Hajduch et al., 2021), and cardiovascular disease (Jensen et al., 2022). Ceramides generated during the progression of metabolic disease can activate protein phosphatase 2A (PP2A) and protein kinase Cζ (PKCζ), and both have been shown to inhibit activation of Akt and thus reduce insulin signaling and glucose uptake (Chaurasia & Summers, 2015; Powell et al., 2003). Ceramide can also directly impair insulin secretion by blocking insulin gene expression in pancreatic islets (Kelpe et al., 2003) and cause cytochrome c release from the mitochondria of pancreatic β-cells to promote their apoptosis (Lang et al., 2011).

Figure 3. Regulation of key metabolic pathways important in obesity By ceramide and S1P.

Ceramide and S1P are increased in obesity. In turn, these important sphingolipid metabolites regulate key metabolic processes that influence obesity and subsequently metabolic diseases.

Ceramide-mediated Inhibition of oxidative phosphorylation and mitochondrial fragmentation has also been reported in adipocytes (Raichur et al., 2014; Turpin et al., 2014), which has been substantiated by the treatment of adipocytes with short-chain ceramide analogs or inhibition of ceramide-biosynthetic enzymes (Chaurasia et al., 2016; Turpin et al., 2014). In addition, ceramide can regulate the thermogenic program in white (beiging) and brown (browning) adipocytes through the attenuation of thermogenic gene expression required for non-shivering thermogenesis (Chaurasia et al., 2016).

While ceramide has collectively gained negative connotations with regards to metabolic disease, recent evidence indicates distinct ceramide species, rather than the total mass, are responsible for disease development (Hammerschmidt & Bruning, 2022). To highlight this, Turpin et al., showed that the predominant ceramide species in rodent adipose tissue was C16-ceramide, which correlated with high expression of its synthesizing enzyme CerS6 in obese patients (Turpin et al., 2014). The negative association of C16-ceramide with disease has now been substantiated by others in mice fed a high fat diet showing that CerS6 knock-down and impaired synthesis of C16-ceramide led to reduced body fat, improved glucose tolerance, and insulin sensitivity (Raichur et al., 2019). Conversely, C24:1-ceramide is reduced in liver, heart and plasma in models of diet-induced obesity, whereby dietary supplementation of C24:1-ceramide negated body weight gain, improved glucose tolerance, improved insulin sensitivity and fatty acid oxidation in the liver (Keppley et al., 2020). Moreover, deletion of alkaline ceramidase 3 in a diet-induced mouse model of obesity increased liver C18:1-ceramide and decreased oxidative stress and severity of non-alcoholic steatohepatitis (Wang et al., 2020). These findings underline the possibility that pharmacologically targeting specific ceramide species may be used as a therapeutic in the context of metabolic disease. For ceramide to contribute to the pathogenesis of metabolic disease, a 4-5 double bond must first be inserted into its precursor dihydroceramide, and perturbation of this metabolic step alone ameliorated insulin resistance and hepatic steatosis in obese ob/ob mice (Chaurasia et al., 2019). Indeed, inhibition of Degs1, which increased the dihydroceramide: ceramide ratio, blocked adipocyte differentiation and lipid deposition (Barbarroja et al., 2015). Moreover, treatment of diet-induced obese mice with fenretinide, which induced the polyubiquitination and degradation of Degs1 (Alsanafi et al., 2021), blocked ceramide accumulation in liver and skeletal muscle and suppressed insulin resistance and hepatic steatosis (Bikman et al., 2012), together with increased circulating adiponectin (McIlroy et al., 2013).

Elevated S1P is also a pathological feature of human and rodent obesity and similarly to ceramide tightly correlates with fat mass and insulin resistance (Blachnio-Zabielska et al., 2012; Geng et al., 2015; Kowalski et al., 2013). However, unlike many ceramide species that are widely considered to promote metabolic disease, the role of S1P is more complicated, potentially arising from its intracellular and extracellular actions and the presence of specific S1P receptors on different metabolic tissues (Green et al., 2021a). Despite positively correlating with obesity, S1P has been shown to produce beneficial effects contrasting those of ceramide on insulin signalling in some tissues (Figure 3). For example, S1P generated in liver hepatocytes can activate Akt to enhance insulin signalling and triglyceride storage. Likewise, overexpression of SphK1 induces glucose uptake via Akt phosphorylation, and conversely, inhibition or reduced expression of SphK1 prevented this (Ma et al., 2007). In contrast, it has been reported that in vitro and in vivo treatment with palmitate induced S1P production that inhibited insulin signalling in the liver through S1PR2 activation (Fayyaz et al., 2014). Moreover, the S1P/S1PR3 axis suppresses Akt and insulin signalling in muscles through IL6 production (Ross et al., 2013). Similarly, opposing functions for SphK2 in metabolic diseases have been reported. On one hand, hepatic-overexpression of SphK2 in a diet-induced mouse model of obesity led to a reduction in liver and circulating ceramide that correlated with a reduction in hepatic lipid accumulation and improved insulin sensitivity (Lee et al., 2015). On the other hand, deletion of SphK2 preserved insulin production and prevented progression of diabetes by protecting pancreatic β-cells from lipoapoptosis (Song et al., 2019).

In addition to insulin signalling, many studies have reported pathophysiological links between S1P and hepatic steatosis, whereby deletion of SphK1 ameliorated diet-induced hepatic steatosis in mice (Chen et al., 2016). Moreover palmitate-induced production of S1P has been reported to activate hepatic stellate cells and activation of S1PR3 that induced α-smooth muscle actin expression in the pathogenesis of NAFLD (Al Fadel et al., 2016; Liu et al., 2011). However, it was reported recently that estrogen stimulated release of S1P only from female hepatocytes, preventing TGFβ-induced fibrosis and expression of collagen in hepatic stellate cells via S1PR3 (Montefusco et al., 2022). It was suggested that this may contribute to sexual dimorphism in NAFLD through an anti-fibrogenic function of the S1P/S1PR3 axis (Montefusco et al., 2022). This highlights the incomplete understanding of the functions of SphK/S1P axis in metabolic disease.

Although global deletion of SphK1 (Chen et al., 2016; Wang et al., 2014), SphK2 (Ravichandran et al., 2019) or enzymes governing ceramide synthesis, such as Degs1 (Chaurasia et al., 2019) or CerS6 (Raichur et al., 2019; Turpin et al., 2014), had beneficial effects on metabolic disease, targeting these enzymes will not alter the total sphingolipid pool, which increases with overnutrition (Kang et al., 2013; Russo et al., 2013). Therefore, in recent years research has been on targeting the SPT-ORMDL complex governing entry into the sphingolipid pool in metabolic diseases.

6. ORMDL in metabolic disease

Much focus has recently been on limiting de novo synthesis with regards to metabolic disease (Figure 4), however much of this research has been conducted on either modulating expression of the SPTLC major subunits or pharmacological inhibition of SPT catalysis (Bonezzi et al., 2019; Chaurasia et al., 2016; Ussher et al., 2010). Treatment with myriocin or adipose tissue-specific deletion of SPTLC2 led to reduced adipose dihydroceramide/ceramide levels and significantly altered adipose tissue and morphology (Chaurasia et al., 2016). Using a diet-induced obesity mouse model it was shown that limiting de novo sphingolipid synthesis led to increased beiging of white adipose tissue, increased mitochondrial activity, and improved insulin sensitivity that was predominantly attributed to the increased presence of anti-inflammatory M2 macrophages within the adipose tissue (Chaurasia et al., 2016). Furthermore, mice lacking SPTLC2 in uncoupling protein-1 (UCP1) expressing cells had reduced ceramide accumulation that correlated with reduced weight gain and diminished energy expenditure in an obesogenic-diet model. These findings suggested that ceramides are necessary and sufficient for diet-induced impairment in thermogenic adipocyte function (Chaurasia et al., 2021).

Figure 4. Effects of modulation of de novo sphingolipid biosynthesis on metabolism.

Summary of studies describing modulation of de novo sphingolipid biosynthesis and their significant biological effects on metabolic tissues.

In addition to adipose, SPTLC has also gained recent attention in the liver with regards to insulin resistance, NAFLD and the progression to hepatocellular carcinoma. Flepatic Sptlc2 was upregulated in response to ER-stressors or an obesogenic-diet, which led to similar increases in ceramide production that inhibited phosphorylation of insulin receptor β and thus elevated fasting glucose levels (Kim et al., 2022). In addition, SPTLC3 expression has been shown to positively correlate with liver fibrosis and the progression of NAFLD to hepatocellular carcinoma (Ijuin et al., 2022).

Historically, interfering with SPTLC subunit expression or catalytic activity of SPT has highlighted the importance of de novo sphingolipid synthesis in metabolic disease. ORMDL is now considered a regulatory intrinsic part of the SPT-ORMDL complex, although only a few studies have investigated its functions in metabolic disease (Li et al., 2021; Wang et al., 2021). A previous publication showed that ORMDL3 expression negatively correlated with BMI (Pan et al., 2018), which has now been substantiated in additional models of obesity (Lee et al., 2020; Pan et al., 2018; Song et al., 2022). In this regard, transgenic and knock-out mouse have shown that ORMDL3 negatively regulate ceramide levels in vivo (Clarke et al., 2019; Debeuf et al., 2019). Whole-body ablation of ORMDL3 in mice has recently been shown to increased susceptibility of mice to weight gain and insulin resistance when fed an obesogenic diet (Song et al., 2022). Furthermore, ORMDL3 knock-out allowed for the accumulation of ceramide that impaired the induction of UCP1 expression and other thermogenic genes in brown and white adipose tissue, following either administration of a β3 adrenergic agonist, or exposing mice to prolonged periods of cold. The reduction in thermogenesis and increase in body weight in Ormdl3 knock-out mice deficiency was reversed by treatment with myriocin, which inhibits the de novo ceramide production (Song et al., 2022). The impairment in adipose tissue thermogenesis and insulin sensitivity due to ORMDL3 deficiency was attributed to the tissue build-up of ceramides (Song et al., 2022). This is consistent with earlier reports from Chaurasia et al., on effects of inhibiting SPT activity (Chaurasia et al., 2016; Chaurasia et al., 2019; Ijuin et al., 2022). In contrast to this apparent beneficial effect of ORMDL3 on obesity, another study highlighted a potential negative role in metabolic function (Lee et al., 2020). It was shown that unlike pancreatic islets from obese human donors that have reduced ORMDL3 expression, ORMDL3 expression was elevated in leptin-deficient ob/ob mice that spontaneously develop obesity. Furthermore, treatment of isolated primary pancreatic islets from ob/ob mice or in vivo administration of leptin reduced ORMDL3 expression, which correlated with reduced body weight gain. Mechanistically, in β cells, knock-down of ORMDL1, 2 or most significantly 3 caused increased cleavage of the pro-apoptotic markers, caspase-3 and poly (ADP-ribose) polymerase (Lee et al., 2020). These two contradictory studies collectively demonstrate that deeper understanding is required into the apparent opposing role of ORMDL in metabolic function.

Of note, there are no studies focussed on the upregulation of ORMDL to limit de novo sphingolipid synthesis in targeting metabolic disease. Given the evidence described above, it is tempting to suggest that limiting SPT activity via increased ORMDL activity and/or expression may have therapeutic potential for limiting obesogenic ceramide/S1P accumulation. However, complications remain with such an approach. Firstly, there are no small molecule activators of ORMDL aimed at increasing the capacity of N-terminal-blocking on the SPT catalytic core. Therefore, genetic overexpression of ORMDL remains the only viable investigative option. Secondly, evidence from SPTLC1 or 2 knock-out mouse studies indicate that de novo sphingolipid biosynthesis is essential for the survival of adipocytes (Alexaki et al., 2017). In addition, although knock-out of SPTLC2 reduced adipose tissue mass, perhaps due to adipocyte cell death, fatty acids that would usually be stored in adipocytes were instead shunted to the liver leading to hepatosteatosis and insulin resistance (Lee et al., 2017). These two studies collectively indicate sphingolipid biosynthesis is a necessary metabolic process and may explain why in a previous study, brown adipocyte specific ablation of SPTLC2 similarly led to a reduction in brown adipose tissue mass (Chaurasia et al., 2021). It should be pointed out that the roles of ORMDL3 in UPR and ER stress, and in regulation of ceramide homeostasis are not mutually exclusive (James et al., 2019; Oyeniran et al., 2015). However, it remains to be investigated as to whether these mechanisms occur in adipocytes in the context of obesity, since elevated ER stress is a key feature causing impaired adipokine release (Ryu et al., 2020) and chronic inflammation in adipose tissues (Kawasaki et al., 2012).

7. Conclusions and future perspectives

The literature herein undoubtedly highlights important roles for bioactive sphingolipids in metabolic functions of adipose and liver tissues, and their implications for obesity and metabolic disease. While much focus has been placed on limiting de novo sphingolipid biosynthesis to ameliorate metabolic disease, a complete block of biosynthesis is not a viable therapeutic option. In this regard, sphingolipid biosynthesis is required for the survival of adipocytes (Alexaki et al., 2017) and intestinal cells (Li et al., 2018), as well as endothelial cell function for vascular development (Kuo et al., 2022). These studies indicate rather than a complete block on SPT activity, tighter control of de novo sphingolipid biosynthesis via marginally increasing ORMDL activation and/or expression could offer a more viable therapeutic option. This would limit the build-up of ceramide and/or S1P, without inducing excess ER-stress from elevated ORMDL expression (Oyeniran et al., 2015) that further exacerbate sphingolipid accumulation and disease progression.

Despite great advances in understanding the pathophysiological role of ORMDLs in asthma, their role in metabolically active tissues, particularly adipose and liver with respect to obesity and metabolic disease remain understudied. However, following the identification of ORMDL3 as an obesity-related gene, studies into the metabolic roles of ORMDLs have begun to appear in the literature. We hope that future research into the mechanistic actions of ORMDL on SPT, the identification of small molecule activators and further elucidating its roles in metabolically active tissues may help to move this young field of research forward.

Abbreviations

ATF6

activating transcription factor 6

CerS

ceramide synthase

CERT

ceramide transfer protein

Degs1

dihydroceramide desaturase

elF2α

eukaryotic translation initiation factor 2α

ER

endoplasmic reticulum

IRE1

inositol requiring enzyme 1

ORMDL

orosomucoid-like protein

PERK

protein kinase R like endoplasmic reticulum kinase

S1P

sphingosine-1-phosphate

S1PR

sphingosine-1-phosphate receptor

SM

sphingomyelin

SphK

sphingosine kinase

SPT

serine palmitoyltransferase

SPTLC

serine palmitoyltransferase long chain base subunit

UPR

unfolded protein response

XBP1

X-box binding protein 1