Ceramide bioactive lipids are ubiquitous and essential for all eukaryotic membranes [11]. Their metabolism in eukaryotic cells, including bone tissue, is a highly conserved process with similar spatial organelles organization, primarily involving the smooth endoplasmic reticulum (sER), the Golgi system, and lysosomes. Crosstalk between these organelles is governed by an integral network of common de novo synthetic and catabolic pathways in response to different stimuli despite the sphingolipids needed by the cell (Fig. 2). Here, we discuss principles governing bioactive ceramide metabolism’s operation relative to bone cells.

L-serine Amino Acid palmitoyl-CoA as Basis Molecules for Ceramide Synthesis

The L-serine amino acid and a long-chain fatty palmitoyl-CoA, an activated form of palmitic acid, are core molecules for sphingolipids de novo anabolic biosynthesis via a chemical reaction of condensation in the sER in all mammalian cells (Fig. 2). A study demonstrated that treatment with L-serine normalizes sphingolipids metabolism and diminishes the acceleration of bone loss caused by long-term fluoxetine use in both animal models and postmenopausal women [30]. In contrast, accelerated levels of palmitic acid enhance the receptor activator of NF-κB ligand (RANKL)-stimulated osteoclastogenesis and is sufficient to induce osteoclast differentiation even in the absence of RANKL [31]. Furthermore, palmitic acid also attenuates osteoblast differentiation and function in rat calvarial cells [32]. Emerging evidence suggests that palmitate loading results in the accumulation of pro-inflammatory ceramide, which activates PP2A-dependent dephosphorylation and activation of the AKT pathway in osteoclasts [33, 34]. Thus, de novo ceramide synthesis may protect against the accelerated accumulation of pro-inflammatory palmitic acid in sER of bone cells. However, it is also important to emphasize that mitochondria-associated membranes may also participate in de novo ceramide synthesis [11]. A recent study demonstrated that inhibiting de novo ceramide synthesis in mitochondria restores age-associated musculoskeletal tissue dysfunction [35]. Furthermore, abnormal mitochondrial fusion and fission correlate with accelerated osteoclastogenesis and diminished osteoblastogenesis in osteoporosis [36]. However, our knowledge about the impact of ceramide species generated in mitochondria versus sER on bone homeostasis and pathology remains limited.

Serine Palmitoyl-Transferase (SPT) Initiates Sphingolipids Synthesis

The amino acid, L-serine, and palmitoyl-CoA condensation form the transient intermediate 3-ketodihydrosphingosine [20]. This process is regulated by a serine palmitoyl-transferase (SPT), an enzyme complex composed of two large subunits, encoded by Sptlc1 and either Sptlc2 or -3, and a small regulatory subunit located at the rough endoplasmic reticulum membrane-bound [37]. Notably, the SPT is the only enzyme that controls de novo sphingolipids synthesis, and it uniquely serves as a critical regulator for steady state and rate-limiting fluctuations in sphingolipid metabolism [12]. Details are provided in Fig. 2.

Several enzymes regulate the SPT function in a eukaryotic cell. More specifically, a family of endoplasmic reticulum-bound Orms/ORMDL and Nogo-B proteins are negative regulators of SPT activity, sensing the ceramide levels in a eukaryotic cell [38, 39]. In contrast, overexpression of the ORMDL3 protein, one of the ORMDL isoforms, accelerated SPT activity, thereby promoting ceramide accumulation in RAW264.7 cells [38]. Since RANKL-stimulated RAW264.7 serves as osteoclast precursors, it is pleasurable that the ORMDL/ORMDL3 axis may regulate ceramide levels in osteoclast precursors.

Keto-Sphinganine Reductase and Ceramide Synthases Control the Generation of Dihydroceramide

The 3-ketodihydrosphingosine is immediately converted into dihydrosphingosine (dhSPH) by 3-ketosphinganine reductase followed by N-acylation by one of six ceramide synthases (CerS1-6/ longevity assurance, Lass1-6) to form dihydroceramide or ceramide. Formation of ceramide or dihydroceramide is dependent on the dihydrosphingosine or sphingosine as a substrate with a saturated or monounsaturated fatty acid of 14–26 carbons [40]. Importantly, these CerS show distinct preferences for the different fatty acyl-CoA substrates generating distinct ceramides with unique N-linked fatty acids (Fig. 3). More specifically, CerS1 selectively regulates the synthesis of C18-dihydroceramide, whereas CerS5 and CerS6 result in the preferential formation of C16-dihydroceramide. It was recently suggested that inhibition of CerS6, as well as CerS1 and CerS5, may serve as an attractive therapeutic regimen for treating metabolic diseases, which are also linked to accelerated inflammatory osteolysis [41]. Finally, it was also suggested that CerS enzymes may localize to different regions of the sER or distinct membranes, such as the nuclear membrane, mitochondrial-associated membranes, and others, which needs to be confirmed in further studies [11]. Recently published studies demonstrated that CERS1 abundance declined with aging and accelerated obesity-induced insulin resistance in musculoskeletal tissue [42, 43]. Furthermore, it was also demonstrated that pharmacological or genetic inhibition of CERS1 in aged mice blunts myogenesis and deteriorates aged skeletal muscle mass and function, which is associated with the occurrence of morphological features typical of inflammation and fibrosis [42].

Dihydroceramide Desaturase Forms the Ceramides, a Central Metabolic Sphingolipid Hub

The removal of two hydrogens from a fatty acid chain of the dihydroceramide by the dihydroceramide desaturase enzymes (DES1 or DES2) results in the formation of ceramide species with 4,5-trans double bond in their sphingoid backbone [12, 37]. More specifically, a double bond is inserted by the Δ4-desaturase activity of the DES1 enzyme, resident in virtually all cells’ endoplasmic reticulum. DES2 is highly expressed only in the skin, kidneys, and intestines; DES2 demonstrates the C-4 hydroxylase activity that enables it to produce phytoceramides with barrier protection function [44, 45]. A study revealed that DES1 global knock-out mice show a normal skeletal structure, indicating that DES2 may still serve; however, these mice demonstrate multiple physiologic anomalies such as weight loss and growth impairment [46]. In contrast, ablation of DES1 may serve as a target for treating metabolic diseases [47]. Altogether, our knowledge about the role of DES1 and DES2 in bone physiology and pathology is limited and needs to be addressed further. In addition to a canonical DES1/DES2-created ceramides with a Δ4-double bond, a recent study detected that fatty acid desaturase type 3 (FADS3) introduces an additional double-bond at the Δ14-position creating d18:2 sphingadienine [48]. Surprisingly, the levels of sphingadienine are linked with gender, being on average ∼30% higher in females with type-2 diabetes [48]. However, the role of atypical d18:2 sphingadienine in sex-associated bone pathology is unclear and needs to be addressed in further studies. Finally, it is also important to emphasize that the prevalence of double bonds in different ceramide bioactive sphingolipid species alters the biophysical properties of the molecules, modifying their elastic properties and packing behavior compared to dihydroceramide [49]. However, both ceramides and dihydroceramides can serve as substrates for the enzymes that produce complex sphingolipids, such as dihydro-sphingomyelins and dihydro-glucosylceramides [50].

Non-Canonical De Novo Ceramide Synthesis and Bone Pathology

When L-serine levels are low, L-alanine or L-glycine is used as a substrate by SPT to yield non-canonical 1-deoxy(methyl)-ceramides (doxCer). These sphingolipids species lack the hydroxyl group (1-deoxy-ceramides), or both hydroxyl and methylene group (1-deoxymethyl-ceramides) on the primary carbon of their sphingoid base. During catabolism, doxCer is degraded by ceramidase to form 1-deoxysphingosine (1-deoxySO). However, 1-deoxySO is not phosphorylated to form S1P. This prevents its cleavage to hexadecenal by S1P-lyase, meaning that the non-canonical catabolic pathway participates in doxCer degradation [51]. Importantly, it was assumed that the same set of enzymes metabolizes 1-deoxySA as canonical sphingoid bases and that 1-deoxySO, like sphingosine, bears a (4E) double bond that DES1 introduces [52]. Published evidence also indicates that these atypical species of sphingolipids were found to exist in the cells at a meager amount compared to other sphingolipid species [53,54,55]. It was further demonstrated that doxCer drives cancer suppression as well as promotes neuropathy and cellular dysfunction through an atypical cell death program, e.g., activation of caspase 3 and 12 or by causing endoplasmic reticulum stress [37, 56, 57]. In contrast, increased plasma levels of doxSLs are also directly implicated in the development and progression of hereditary sensory and autonomic neuropathy type 1 and diabetes type 2 [58]. Therefore, we may speculate that doxSLs also accelerate inflammatory osteolysis, which needs to be investigated in further studies.

Ceramide Transporter Proteins and Bone Health

De novo synthesized ceramides are primarily transported from sER to the Golgi by the ceramide transfer protein, CERT, at sER-trans-Golgi contact sites to form sphingomyelin Fig. 2). CERT specifically recognizes natural d-erythro ceramides and efficiently transfers the natural ceramide isomers dihydroceramide and phytoceramide and also various ceramide molecular species having C14-C20 amide-acyl chains [59]. No effect of CERT on the transport of sphingosine, sphingomyelin, cholesterol, or phosphatidylcholine was reported. To our knowledge, it was demonstrated that knockdown of CERT impairs insulin signaling in the presence of palmitate-induced lipotoxicity, whereas overexpression of CERT reduces ceramide level by 4.1-fold and restores insulin sensitivity in palmitate-treated C2C12 myoblasts [60]. In contrast, the role of CERT in bone homeostasis is unknown and needs to be established in further studies. In addition to the CERT-dependent ceramide transport, they could also be transported to cis-Golgi by vesicle transport and further glycosylated to glycol-ceramide by Glucosylceramide synthase (GCS), which is then further trafficked in the Golgi by four-phosphate adaptor protein 2 (FAPP2) to form a complex of glycol- and lacto-sphingolipids. It is speculated that FAPP2 may also be involved in mediating the proinflammatory effects of ceramide on various mammalian cells, including bone cells [61].

Critical Role of Mammalian-Derived Ceramides on Osteoblastogenesis and Osteoclastogenesis

Published evidence indicates that ceramide serves as a second messenger in promoting inflammation and cell death downstream of inflammatory cytokines, such as tumor necrosis factor-α (TNF-α) [37, 62]. The role of ceramides and other sphingolipids in bone physiology and pathology has been extensively studied in the past decades and is well-addressed in recently published studies [12, 24]. Indeed, ceramide suppresses bone formation [62]. In contrast, published evidence also demonstrated a positive correlation between elevated levels of C16:0, C18:0, C18:1, and C24:1 ceramide and bone resorption markers, as well as hip fracture [24, 63]. In addition, an experimental mouse model of postmenopausal osteoporosis demonstrates accelerated levels of ceramide species in collected bone femurs [64].

At a bone cellular level, it was also demonstrated that a low concentration of cell-permeable C2-ceramide upregulated osteoblast survival [65]. Furthermore, this study also demonstrated that direct application of C2-ceramide at a concentration of > 10− 6 M significantly reduced osteoblast viability in a dose- and time-dependent manner. Our group reported that a cell-permeable C6-ceramide accelerates osteoclastogenesis using RANKL-primed RAW264.7 cells in vitro [25]. In contrast, no effects of cell non-permeable C2- and C6-dihydroceramide on osteoblastogenesis and osteoclastogenesis, respectively, were observed [25, 65], indicating that eukaryotic dihydroceramide may have a limited effect on bone cells. However, our group also reported that a structurally unique bacterial-derived dihydroceramide, iso-C17:0-Dihydroceramide-1-phosphoglycerol, penetrates the cellular membrane of osteoclast precursors and dramatically accelerates osteoclastogenesis in various in vitro and in vivo models [25, 27]. Furthermore, we also observed that bacterial-derived dihydroceramide upregulates the levels of intracellular mammalian ceramides (unpublished observations). Therefore, it is clear that our knowledge about the role of eukaryotic and bacterial-derived dihydroceramide on bone cells is limited and needs to be erased in further studies. When taken together, these results strongly suggest that ceramide species upregulate osteoclastogenesis and suppress osteoblastogenesis.

Degradation of Ceramide by Ceramidases and Bone Health

The status between cell inflammatory and anti-inflammatory status is controlled by ceramide and sphingosine-1-phosphate. Notably, sphingosine-1-phosphate is exclusively generated from the catabolism of ceramides and is mainly controlled by ceramidases (Fig. 2). More specifically, ceramidases (acylsphingosine deacylase, glycosphingolipid ceramide deacylase) cleave a fatty acid from ceramide, producing sphingosine, which in turn is phosphorylated by a sphingosine kinase to form sphingosine-1-phosphate [20, 66, 67]. Therefore, we can postulate that ceramidases control the balance between ceramide and sphingosine-1-phosphate in bone and other tissues. Ceramidases are classified into the acid, neutral, and alkaline ceramidase subtypes according to the pH optima for their catalytic activity (see reviewers by [20, 66,67,68]).

Acid Ceramidase and Bone Health

Acid ceramidase is the most widely studied ceramidase [20]. Under physiological conditions, it is actively secreted extracellularly from macrophage lysosomes [69]. Genetic mutation in aCDase Asah1 promotes the accumulation of pro-inflammatory ceramides and is implicated in the rare Farber Disease [70]. Patients with Farber disease demonstrate accelerated levels of multiple bone diseases, including e.g., erosion of bone near joints, osteoporosis, and different types of peripheral osteolysis [71]. We demonstrated that diminished acid ceramidase/ASAH1 mRNA expression in patients with periodontal bone lesions [72]. In contrast, we also observed that acid ceramidase exhibits anti-inflammatory effects in OECM1 human cells exposed to P. gingivalis challenge [72]. Surprisingly, our group also reported that a structurally unique pro-osteoclastogenic dihydroceramide isolated from oral bacteria P. gingivalis diminishes the expression of acid ceramidase and accelerates accumulation of pro-inflammatory ceramides in OECM-1 in vitro [21]. Therefore, we conclude that acid ceramidase is critical in crosstalk between bacterial-derived dihydroceramide and host ceramides. Furthermore, our group and others suggested a novel therapeutic regimen using recombinant acid ceramidase protein-based therapy that reduces the levels of pro-inflammatory ceramides and elevates anti-inflammatory sphingosine-1-phosphate.

Role of Neutral and Alkaline Ceramidases in Bone Health

Neutral ceramidase (ASAH2) was detected in mitochondria and an O-glycosylation membrane protein. It is only expressed at the intestinal brush border of the small intestine [68]. Our group observed no neutral ceramidase expression in periodontal lesions either [72]. Therefore, neutral ceramidase’s role in controlling intracellular ceramide in bone cells may be limited. However, neutral ceramidase-based therapy may still demonstrate beneficial effects in controlling extracellular ceramide-mediated osteolysis.

Alkaline ceramidases encoded by ACER1, ACER2, and ACER3 genes. All the biochemically characterized members of the ACER family are localized to the sER, the Golgi complex, or both [67]. Alkaline ceramidases ACER1-3 play a crucial role in regulating the levels of (dihydro/phyto)ceramides, sphingoid base (dihydrosphingosine, sphingosine, phytosphingosine), and their phosphates in mammalian cells [67]. The impact of ACER1 on bone physiology and pathology may be limited due to its high expression in keratinocytes [73]. The role of each member of the ACER2 and ACER3 in ceramide metabolism and accumulation of sphingosin-1-phosphate in circulation is discussed in earlier reviews [20, 67, 68]. Notably, a study demonstrated that human ACER2 and ACER3 may have a redundant role in regulating the levels of unsaturated long-chain ceramides in cells and tissues [67, 74].

How Do Formation and Transport of Sphingosine-1-Phosphate Affect Bone Health?

The primary role of sphingosine kinase is to generate sphingosine-1-phosphate from sphingosine by phosphorylation. Two isoforms of sphingosine kinase exist (SphK1 and SphK2), and they catalyze the same reaction but with different subcellular localizations [75]. SphK1 is mainly a cytosolic enzyme, while SphK2 is localized in different cell compartments, such as the nucleus, endoplasmic reticulum, and mitochondria [76]. Although SphK1 stimulates growth and survival, it was also found that SphK2 promotes apoptosis in different cell types and inhibits cell proliferation [17]. A study reported that Sphk1 expression and activity is increased during RANKL-induced osteoclastogenesis [77]. Furthermore, no effect of SphK1 loss of function on osteoclasts was observed in mice at 12 weeks of age. In contrast, Sphk2 loss-of-function leads to a dramatic decrease in bone density (osteopenic phenotype) in both male and female mice [78]. Indeed, it was recently suggested that the Sphk1,2/sphingosine-1-phosphate axis plays a vital role in bone by affecting the balance of bone metabolism through the regulation of different signaling pathways in osteoblasts and osteoclasts [17].

Although ceramidases and sphingosine kinases function in removing excess effect of ceramide, sphingosine-1phosphate is a potent signaling molecule. Intracellularly generated sphingosine-1phosphate is transported out of cells by several ATP-binding cassette transporters and non-ATP-dependent organic ion transporter named spinster homologue 2 (SPNS2) [37]. It was shown that the hormone calcitonin reduces the expression of SPNS2 on osteoclasts, decreasing sphingosine-1phosphate secretion [18]. Moreover, SphK and SPNS2 are upregulated in osteoblasts and chondrocytes during differentiation, thereby inhibiting sphingosine-1-phosphate signaling and diminished matrix mineralization [79]. Intriguingly, patients with spondyloarthritis, characterized by increased bone mineralization, demonstrate elevated serum levels of sphingosine-1-phosphate, suggesting that sphingosine-1-phosphate accelerates bone formation [79].

Sphingosine-1-phosphate acts in an autocrine or paracrine fashion as a ligand for membrane G-protein coupled receptors, namely, S1PR1-5 [37]. Notably, S1PR5 is expressed only in the central nervous system [37]. Multiple reviewers describe molecular mechanisms of sphingosine-1-phosphate/S1PR1-4 axis actions in bone formation and remodeling well and are out of the scope of this study [12, 24, 80]. Importantly, a recently published study reported increasing levels of sphingosine-phosphate accelerates bone vascular density via sphingosine-1-phosphate/S1PR3/VEGF signaling pathway, providing the basis for the development of new osteoanabolic therapies [81]. Of note, both S1PR1/S1PR3 and S1PR2 agonists significantly downregulated the expression of osteogenic genes and suppressed AKT activation, resulting in an attenuated osteogenic capacity of dental pulp cells [82]. Furthermore, our group recently reported that Porphyromonas gingivalis-derived phosphoethanolamine dihydroceramide inhibits expression patterns of S1PR3 mRNA in human OECM-1 cells in vitro [21]. Therefore, further studies are needed to determine the role of bacterial-derived dihydroceramide in sphingosine-1-phosphate signaling in bone formation and remodeling. Altogether, sphingosine-1-phosphate is a coupling factor between osteoclasts and osteoblasts as we play a key role in vasculogenesis and angiogenesis.