An update on the biology of sphingosine 1-phosphate receptors

Abstract

Sphingosine 1-phosphate (S1P) is a membrane-derived lysophospholipid that acts primarily as an ex­tracellular signaling molecule. Signals initiated by S1P are transduced by five G protein-coupled receptors, named S1P_1–5. Cellular and temporal expression of the S1P receptors (S1PRs) determine their specific roles in various organ systems, but they are particularly critical for regulation of the cardiovascular, immune, and nervous systems, with the most well-known contributions of S1PR signaling being modulation of vascular barrier function, vascular tone, and regulation of lymphocyte trafficking. However, our knowledge of S1PR biology is rapidly increasing as they become attractive therapeutic targets in several diseases, such as chronic inflammatory pathologies, autoimmunity, and cancer. Understanding how the S1PRs regulate interactions between biological systems will allow for greater efficacy in this novel therapeutic strategy as well as characterization of complex physiological networks. Because of the rapidly expanding body of research, this review will focus on the most recent advances in S1PRs.

Sphingosine 1-phosphate [2S-amino-1-(dihydrogen pho­sphate)-4E-octadecene-1,3R-diol] (S1P) is a simple membrane-derived lysophospholipid with regulatory roles in almost all facets of mammalian biology (1). Concentrations of S1P in blood and lymph plasmas are high, in the high nanomolar to low micromolar ranges, whereas S1P concentrations in tissues are kept low, creating an S1P gradient (2). S1P signals through five highly-specific G protein-coupled receptors with nanomolar dissociation constants (3, 4). Expression patterns of the five S1P receptors (S1PRs) vary in tissues and also during development and ageing. S1P₁, S1P₂, and S1P₃ are essentially ubiquitously expressed, whereas expression of S1P₄ and S1P₅ are highly restricted to distinct cell types (4).

Production of S1P can be initiated by external or internal signals, which lead to activation of the biosynthetic pathway beginning with metabolism of membrane SM to ceramide by SMases (5, 6). Ceramide, an important signaling molecule itself, can be metabolized by ceramidase to sphingosine (Sph) (7). Sph is then phosphorylated by one of two Sph kinases (Sphks), Sphk1 or Sphk2, resulting in S1P genesis (8–10) (Fig. 1).

Fig. 1.

Synthesis and export of S1P. S1P synthesis primarily begins with metabolism of membrane SM. Once synthesized, S1P can be irreversibly degraded to phosphoethanolamine and hexadecenal by S1P lyase, or actively transported out of the cell. Once outside of the cell, S1P is found bound to ApoM or albumin. Spns2, spinster 2.

Although there are proposed intracellular roles for S1P, it is often transported out of the cell where it can act in an autocrine or paracrine manner on S1PRs (11, 12). Transport out of the cell may occur via several transporters; however, the only bona fide transporter to date is spinster 2, which is also capable of FTY720 (fingolimod/Gilenya; Novartis) export (13–22). Once outside of the cell, S1P can bind to two known carriers, albumin or ApoM (6, 23, 24) (Fig. 1). Approximately 35% of plasma S1P is bound to albumin and 65% to ApoM, which is found on a small percentage (∼5%) of HDL particles (24). This ApoM+HDL-bound S1P has been proposed as a primary contributor to the vasoprotective properties of HDLs (25–27). How albumin or ApoM deliver S1P to specific S1PRs has yet to be characterized.

AGONISTS AND ANTAGONISTS

There are several well-characterized agonists and antagonists of S1PRs; however, most compounds have been directed toward modulating the activity of S1P₁. FTY720 is the prototypical S1PR agonist and was approved by the US Food and Drug Administration as a first line oral therapy for relapsing-remitting multiple sclerosis (MS) (18, 28). Although FTY720 acts as an agonist at picomolar to nanomolar concentrations on S1P₁ and S1P_3–5, it also acts as a functional antagonist for S1P₁ by inducing receptor endocytosis and degradation of this receptor (29–31). This promiscuity may be responsible for adverse affects, such as acute bradycardia (decreased heart rate) and hypertension, seen in fingolimod-treated patients (32, 33). Initial results from rodent studies indicated that FTY720 phosphate activation of S1P₃ was responsible for both bradycardia and hypertension; however, treatment of humans with more selective agonists indicated that S1P₁ agonism was responsible for reduced heart rate, whereas S1P₃ signaling contributed to the development of hypertension (34–37). The divergent utilization of S1P₁ and S1P₃ in rodents versus primates for the regulation of these coordinated physiological functions highlights the difficulties encountered upon extrapolation from rodent model-based characterization of S1PR function to human disease therapies.

SEW2871 is an S1P₁-specific agonist that activates ERK1/2, AKT, and Rac signals at nanomolar concentrations and induces receptor internalization and recycling; however, it has a relatively short half-life in vivo (38). AUY954 is another commonly used S1P₁-selective agonist with an EC₅₀ of approximately 1 nM, which induces phosphorylation of ERK and AKT (39). At high concentrations, AUY954 also has some activity on S1P₅ (39). Conversely, W146 antagonizes AKT and ERK phosphorylation and is the only widely utilized S1P₁-specific antagonist (40). Administration of W146 enhances vascular leakage and induces pulmonary edema (40, 41). VPC23019 is a useful in vitro tool as a dual S1P_1/3 antagonist; however, poor stability and in vivo efficacy limit its use (42–44). The only known compound with activity at S1P₂ is JTE-013, an antagonist with an IC₅₀ of approximately 20 nM, which blocks S1P₂ signaling through Rho-associated protein kinase (ROCK) and phosphatase and tensin homologue (45, 46). The S1P₂ specificity of JTE-013 has been called into question by several studies that indicate it may have activity at S1P₄ as well as non-S1PR-mediated effects (44, 47–49).

VASCULAR AND LYMPHATIC SYSTEMS

Many effects of S1P on the vasculature are due to expression of S1P₁ by the endothelium. S1P₁, originally named EDG1 (endothelial differentiation gene), was discovered during a search for immediate early genes regulating endothelial cell differentiation (50). Although S1pr1^−/− embryos developed a vascular network, they died in utero at E12.5– E14.5 due to defective coverage of large vessels by pericytes and vascular smooth muscle cells (VSMCs) (51, 52). Specifically, the aorta exhibited severe morphological abnormalities, endothelial hypersprouting, and altered VSMC recruitment and localization (Fig. 2) (53, 54). The generation of inducible cell-specific S1P₁ knockout mice has clarified the roles of endothelial cells (ECs) or VSMC S1P₁ in the regulation of postnatal vascular development, maturation, and function. In the developing retinal vasculature, S1P₁ expression is restricted to the ECs and increases with vessel maturity, as the lowest levels of expression are found at the vascular leading front (55). Postnatal deletion of EC S1pr1 did not affect mural cell recruitment or vessel coverage in the retina; however, angiogenic hypersprouting occurred, characterized by dilated vessels and increases in the number of branch points and tip cells. Induced over-expression of EC S1P₁ suppressed vascular sprouting (55). Changes in the vascular architecture of EC S1pr1^−/− mice were accompanied by increased vascular permeability, resulting from altered vascular endothelial cadherin localization at endothelial cell-cell junctions (54, 55). These data confirmed numerous earlier in vitro studies describing the necessity of EC S1P₁ for the maintenance of vascular barrier function through adherens junction formation induced by activation of Rac after G_αi coupling to S1P₁ (Fig. 2) (56, 57).

Fig. 2.

Expression of S1PRs and responses by endothelial cells. Endothelial cells express S1P₁, S1P₂, and S1P₃ protein. Endothelial cells may express different S1PRs depending on activation status.

Maintenance and formation of adherens junctions was dependent on S1P₁ signaling initiated not only by ligand, but also by fluid shear stress (Fig. 2). Examination of murine aortae found that areas of turbulent flow (the lesser curvature) had poor endothelial cell alignment and S1P₁ relocalized from the EC surface to endocytic vesicles, whereas in the descending aorta, an area of laminar flow, S1P₁ and vascular endothelial cadherin colocalized to the cell surface (55). Additionally, maintenance of vascular homeostasis by the endothelial glycocalyx, which is also susceptible to changes in flow dynamics, was dependent upon S1P₁-induced inhibition of matrix metalloproteinase (58).

Mice with endothelium-specific deletion of S1P₁ developed severe pathology in a model of renal ischemia/reperfusion injury, both in the kidneys and the liver, characterized by elevated plasma creatinine, alanine transferase, and tissue necrosis (59). Conversely, of the five S1PRs, S1P₂ mRNA in the kidney was most increased upon renal ischemia/reperfusion, and mice deficient in S1P₂ developed significantly less pathology compared with WT controls (60). When S1pr2^−/− mice were treated with the S1P₁ antagonist, W146, before ischemia/reperfusion, they were no longer protected from renal injury, suggesting that S1P₁ and S1P₂ in the renal vasculare endothelium play protective and injurious roles, respectively, in kidney injury and disease (60).

The pro-inflammatory tendency of S1P₂ is supported by in vitro studies suggesting a paracrine feedback loop involving EC TNFα induction of S1P₂ expression leading to activation of nuclear factor (NF)-κB and increases in intracellular adhesion molecule (ICAM)-1 and vascular cell adhesion molecule (VCAM)-1 (61). In vivo studies utilizing S1pr2^−/− mice and a model of acute inflammation, endotoxemia, further support the conclusion that S1P₂ is an important regulator of vascular activation and therefore, permeability (62). Induction of endotoxemia in mice lacking S1pr2 in the stroma and not in the bone marrow (BM) compartment resulted in decreased vascular permeability, VCAM-1 and ICAM-1 expression, and more rapid resolution (62). Similarly, in vitro, S1P₂ actively suppressed angiogenic sprouting through leukemia-associated RhoGEF (LARG) activation of RhoC (63). These recent studies reaffirm the conclusion that an antagonistic relationship exists between S1P₁ and S1P₂ in the vascular endothelium during tissue injury and disease.

Lymphatic endothelium also expresses S1PRs, although more interest has focused on the role it may play in S1P metabolism (21, 64). Examination of murine iliac collecting lymph vessels demonstrated that while S1P does not induce nitric oxide or prostaglandin release, signaling via S1P₂ regulates tonic contractility of lymph vessels, as shown using S1P₂ inhibition by JTE013 (65).

IMMUNE SYSTEM

S1PRs regulate many aspects of immune cell biology. The best known is the regulation by S1P₁ of lymphocyte migration out of the secondary lymphoid organs into the blood and lymph (Fig. 3) (66). Regulation of migration occurs by S1P₁ counteracting the retention signals provided by the chemokine receptor CCR7 (67). However, this is not the only role for S1P₁ in lymphocytes, and roles for the other four S1PRs in the immune system have recently been revealed.

Fig. 3.

Expression of S1PRs and responses by cells of the acquired immune system. T cells express S1P₁ and S1P₄, B cells express S1P₁, S1P₂, S1P₃, and S1P₄, and NK cells express S1P₁ and S1P₅. Cells do not necessarily express all of the illustrated S1PRs at one time, but may have differential expression during different stages of maturation or activation.

The contribution of S1PRs to regulation of the immune response has been studied extensively in the context of experimental autoimmune encephalomyelitis (EAE), the most commonly used animal model of MS (68). Although EAE and MS are considered to be primarily diseases of the immune system, the role of S1PRs on neural cells is also gaining an appreciation and will be discussed later. FTY720 is a Sph analog that is phosphorylated, acts on S1P_1,3–5, and was the first US Food and Drug Administration approved oral therapy for MS (69). The presumed mechanism of action has been the trapping of autoreactive T and B cells in the lymphoid organs, away from the central nervous system (70, 71). However, T cell S1P₁ may also regulate the activation and differentiation status of these immune cells. Deletion of T cell S1P₁ significantly suppresses the ability of these cells to be polarized to T-helper (Th)17 in vitro (72). Conversely, when EAE was induced in mice expressing an internalization-defective S1P₁ (S5A), this significantly increased polarization of T cells to the Th17 phenotype resulting in increased disease pathology and immune cell infiltration into the CNS (72).

S1P₁ is also expressed on CD4 T cells isolated from human rheumatoid arthritis patients (73). S1P enhances TNFα-induced expression of the receptor activator of nuclear factor kB (RANK) ligand by these cells, an effect replicated in a synovial cell-like cell line, MH7 (73). In collagen-induced models of rheumatoid arthritis, a S1P₁-specific antagonist prevented or ameliorated disease by upregulating lymphocyte CD69 expression, which downregulates S1P₁ surface expression, blocking thymic egress (73–75).

S1P₁ also affects other populations of T cells, such as T regulatory cells (T_reg), which, as the name implies, play an important role in controlling immune responses and T memory cells (76, 77). S1P₁ suppresses T_reg development via the AKT/mammalian target of rapamyacin pathway and affects their migration from the thymus and out of the periphery by counteracting CCR7 retention signals, similar to the mechanism regulating the egress of effector T cells from lymph nodes (67, 77, 78). S1P₁ signals may also modulate nuclear localization of the transcription factor forkhead box P3, which is necessary for T_reg generation (78). In human patients, FTY720 significantly increased the number of T_reg while decreasing central memory T cells (79). In a specific subset of T memory cells, nonlymphoid resident memory cells (T_RM), cytokines that induce the T_RM phenotype also downregulate the transcription factor Krüppel-like factor2 and its target gene, S1pr1 (80). Subsequently, T_RM are unable to sense S1P in circulation and are maintained in the periphery.

Although S1P₁ has been the focus of much research, not much is known of the roles of the other S1PRs. In CD8 effector T cells, S1P₄ may influence their trafficking to lymph nodes (LNs), although it appears not to be a primary regulator (81). S1pr4^−/− mice have decreased Th17 T cell polarization; however, reduced Th17 differentiation is likely T cell extrinsic and primarily due to functions of S1P₄ in dendritic cells (DCs) (81).

S1PR expression choreographs many aspects of B cell subset localization within lymphoid organs, thereby affecting their functionality; however, there are some direct effects of S1P signaling on B cell survival (Fig. 3) (82, 83). While S1P₁ has some regulatory functions in B cells, it appears that S1P₂ has a greater impact on these cells. Aged S1pr2^−/− mice develop diffuse large B cell lymphoma (DLBCL), characterized by increased germinal center (GC) B cells and spontaneous GC formation, which correlates with an approximate 26% mutation incidence for S1PR2 in human DLBCL (84). Under homeostatic conditions, S1P₂ signals via G_12/13 to activate Rho/ROCK, antagonizing activation of AKT and pro-survival signals (82). B cell S1P₂ also regulates follicular positioning of B cells by directing their clustering to GC in response to follicular DC-derived S1P (82, 85). The ability of follicular B cells to exit the follicle is, however, dependent upon S1P₁ expression (86). Additionally, trafficking of marginal zone (MZ) B cells between the MZ and the follicle is regulated by S1P₁, which maintains these cells in the MZ in order for them to capture blood-borne antigens (86–88).

Studies of nonobese diabetic mice have shown that upregulation of S1P₃ by MZ B cells and their T2 MZ precursors may also play a role in enhancing MZ retention in these mice (89, 90). S1P₃ has already been shown to regulate B cell migration in vitro, but not in vivo, in WT mice (83, 87). However, it may be important for positioning of immature B cells and their progenitors within the BM, whereas S1P₁ participates in directing their migration from the BM parenchyma into sinusoids and subsequently into circulation (83).

Natural killer (NK) cells are considered innate lymphoid cells that develop from lymphoid progenitors in the BM, but do not undergo genomic changes that occur in the B or T cell receptor genes (91, 92). They are important for anti-tumor immunity and are prolific producers of IFNγ (92). Mouse NK cells have low levels of transcript for S1pr1, S1pr2, and S1pr4 and high S1pr5 mRNA levels (Fig. 3) (93, 94). S1P₅ normally antagonizes NK CXCR4 BM retention signals, and S1pr5^−/− mice have decreased numbers of NK cells in the periphery and increased numbers in lymph nodes and BM due to defective migration (93, 95). This phenotype is also observed in the mouse model of Niemann-Pick disease type C, a lysosomal storage disorder presenting as an accumulation of cholesterol and sphingolipids in the lysosome and decreased concentrations of circulating S1P in human patients (96, 97). Studies utilizing FTY720 indicated that S1P₁ also contributes to NK cell migration from LN to lymph, but the contribution is relatively minor compared with that of S1P₅, which is not subject to CD69 regulation (75, 94). Decidual NK (dNK) cells are a specialized NK cell subset that regulates trophoblast invasion during early pregnancy by secreting pro-angiogenic and growth factors, including vascular endothelial growth factor (VEGF) (98). S1P₁ and S1P₅ are increased in human dNK cells compared with circulating NK cells, and S1P₅ expression decreases after the first trimester (99). FTY720 treatment decreased dNK S1P₅ expression, VEGF production, and trophoblast invasion in vitro (99).

Macrophages are important sentinel cells that develop from monocytes to fight infection and repair damaged tissue (100). S1PRs expressed by monocytes and macrophages regulate their migration and activation, and the receptors responsible are cell subtype- and situation-specific (Fig. 4). In general, S1P₁ and S1P₃ appear to induce migration toward S1P, whereas S1P₂ expression repulses macrophages from S1P (101, 102). S1pr2^−/− mice on a pro-atherogenic genetic background (Apoe^−/−) developed significantly less atherosclerosis, accompanied by decreased macrophage and monocyte retention in atherosclerotic plaques, indicating effects on migration, tissue retention, and activation (103). In comparison, S1pr3^−/− mice on the same Apoe^−/− background do not have altered development of atherosclerosis, but do have decreased monocytes and macrophages with atherosclerotic lesions (101). In WT mice, treatment with FTY720 results in decreased circulating monocytes; however, use of the S1P_1/4/5 agonist, BAF312, yielded similar results, both at homeostasis and during EAE, indicating that S1P₃ is not the sole regulator of monocyte circulation (104). This could be a cell subtype-specific effect, or dependent on environment, as local administration of FTY720 appeared to enhance recruitment of anti-inflammatory pro-angiogenic monocytes (105). This supports an earlier report that macrophage S1P₃ induces a pro-regenerative phenotype in a model of renal ischemia/reperfusion (106).

Fig. 4.

Expression of S1PRs and responses by cells of the innate immune system. Monocytes and/or macrophages express S1P_1–4, neutrophils express S1P₁, S1P₃, and S1P₄, eosinophils and MCs express all S1PRs, and DCs express S1P₁, S1P₃, and S1P₄. Cells do not necessarily express all of the illustrated S1PRs at one time, but may have differential expression during different stages of maturation or activation.

A report utilizing the zymosan peritonitis model proposed that the resulting apoptotic neutrophils induced S1P₁ expression on recruited macrophages and that S1P₁ is necessary for emigration from the inflamed peritoneum, but has no role in efferocytosis or activation (107). S1P₂ on alveolar macrophages may regulate their phagocytic capacity, as S1pr2^−/− alveolar macrophages displayed decreased phagocytosis of the fungus Cryptococcus neoformans due to decreased expression of Fc receptors necessary for phagocytosis of antibody-opsonized fungus (108).

Neutrophils are the first immune cell line of defense and can shape the immune response (109). Neutrophils express mRNA for all S1PRs; however, the level of expression and the ability of S1P to affect changes in their responses depend upon their activation status (Fig. 4) (110). More recently, it was reported that S1P lyase (Sgpl)^−/− mice are unable to degrade S1P and have neutrophilia (111). Although S1P₄ deficiency in Sgpl knockouts resulted in circulating neutrophil numbers that were close to WT, S1P₄ was not specifically deleted in neutrophils, raising the possibility that multiple cell types were responsible for the effect. Specific deletion of neutrophil S1P₁ did not normalize neutrophil numbers in Sgpl^−/− mice. However, in rat models of hyperalgesia dependent upon neutrophil infiltration, S1P₁ was necessary for neutrophil recruitment (112). Specific S1P₁ antagonism blocked neutrophil infiltration, whereas agonism increased sensitivity.

Eosinophils and mast cells (MCs) are both involved in anti-parasite immune responses and allergic immunity (113). Eosinophils from mice over-expressing interleukin-5, an eosinophil growth factor, express high levels of S1P₃ and demonstrate increased chemotactic responses to S1P in vitro (Fig. 4) (114). In a model of allergic rhinitis, FTY720 treatment significantly decreased the numbers of infiltrating MCs and eosinophils, resulting in resolution (115). In vitro, FTY720 induced MC apoptosis in a dose-dependent manner (115). Similar to lymphocytes, S1P₁ regulates MC migration toward the antigen, whereas S1P₂ regulates their activation status upon FcεRI ligation, inducing degranulation and CCL2 secretion (116).

DCs are professional antigen-presenting cells and as such, are required for proper induction and direction of the acquired immune response (117). Both human and mouse DCs express mRNA for S1P_1–5 and exhibit varied responses to S1P stimulation in vitro and in vivo (Fig. 4) (118, 119, 120). Langerhans cells, skin resident DCs, require S1P₁ for migration to LN, whereas kidney resident DCs require S1P₃ for maturation in ischemia/reperfusion (121, 122). This is also the case in models of sepsis, where DC S1P₃ is required for interleukin-1β production (123). In EAE, although S1P₁ agonism decreased disease pathology, it did not affect entry into the CNS of a subset of DCs (plasmacytoid DCs). However, plasmacytoid DCs in the CNS were necessary for the efficacy of S1P₁ agonist treatment (124).

S1P₄ was cloned from mature human DCs, yet not much is known about the role this receptor plays in these cells (125). In models of autoimmune disease, Th2-type immune responses such as allergic airway inflammation and cutaneous hypersensitivity, S1pr4^−/− mice had increased pathology and up to 50% increase in DCs in draining LN after topical antigen application (81). This implies that S1P₄ may antagonize S1P₁ in DCs, regulating their ability to migrate from the periphery after antigen uptake.

NERVOUS SYSTEM

Neural progenitors express mRNA for S1P_1–5 and respond to S1P stimulation with induction of Ca²⁺ mobilization (Fig. 5) (126). S1P regulates embryonic nervous system development, as the neuroepithelial layers of the developing telencephalon in S1pr1^−/− embryos have significantly increased apoptosis and decreased mitosis (127). S1P₂ may also play a role in regulating neural progenitors, as postischemic administration of the S1P₂ antagonist JTE-013 or short hairpin RNA against S1P₂ significantly increased progenitor migration to the ischemic region (128). This indicates that S1P₂ may repel neural progenitors from areas of high S1P concentration in the same manner as it regulates macrophage migration (102). Indirectly, S1P signaling on astrocytes affects neural progenitors by increasing lamin production, thereby encouraging maturation and neurite outgrowth by progenitors (129). Interestingly, neural stem cells were protected from radiation-induced apoptosis by nanomolar FTY720 treatment in vitro, although it is unknown which receptor is involved in this protection (130).

Fig. 5.

Expression of S1PRs and responses by neural cells. Neural progenitors express S1P₁ and S1P₂, neurons express S1P₁ and S1P₃, oligodendrocytes express S1P₁ and S1P₅, and astrocytes express S1P₁ and S1P₂. S1P₁ couples exclusively to G_αi. S1P₂ and S1P₃ can couple to G_αi, G_α12/13, or G_αq, and S1P₅ can couple to G_αi or G_α12/13. Cells do not necessarily express all of the illustrated S1PRs at one time, but may have differential expression during different stages of maturation or activation.

Although analyses of entire mouse dorsal root ganglion found that S1P₃ was the most highly expressed S1PR, single cell mRNA analysis of individual neurons found that S1P₁ was most highly expressed, regardless of neuronal subtype, indicating that high expression of S1P₃ occurs in ganglion cell types other than neurons (Fig. 5) (131, 132). One group found that pain responses induced by intradermal S1P injection or models of postoperative pain were significantly decreased in S1pr3^−/− mice, whereas minimal differences were seen in S1pr1^−/− mice (131); however, another group found that mice lacking S1P₁ specifically in nociceptor neurons were protected from S1P-induced pain (133). Finally, in the murine model of the neurodevelopmental disease, Rett syndrome, FTY720 or S1P₁-specific agonist SEW2871 in vivo treatment increased neuron production of brain-derived neurotrophic factor and decreased neurological symptoms (134).

Oligodendrocytes are the myelinating cells of the CNS and the primary cell type affected in MS and in the mouse EAE model (135). Process retraction, Rho/ROCK-mediated inhibition of immature oligodentrocyte precursor migration, and G_i/AKT-mediated survival in mature oligodendrocytes occurs via S1P₅ (Fig. 5) (136, 137). Ex vivo studies using cerebellar slice cultures indicated that S1PR agonism, particularly S1P₁, could prevent or reverse demyelination, explaining the ability of FTY720 to induce remyelination and process extension in the same system (138, 139). Data from a different in vitro system, myelinated neurospheres, indicated that FTY720 decreased microglial activation and oligodenrocyte apoptosis, and induced remyelination primarily by S1P₅ agonism (140). An in vivo study provides conflicting evidence to these in vitro studies, reporting no effects on myelin repair with FTY720 treatment; however, the models of demyelination utilized in both the in vitro and in vivo studies were induced chemically and were meant to exclude possible effects of immune or vascular cells (141). As such, they cannot model complex neuroinflammatory disease and care must therefore be taken when attempting to extrapolate results to in vivo disease, such as EAE or MS.

The resident immune cells of the CNS, microglia, express all S1PRs (142). In vitro studies indicated that FTY720 downregulated production of pro-inflammatory molecules by microglia while increasing neurotrophic factor production, resulting in an overall neuroprotective phenotype (142). FTY720 also inhibited secretory vesicle mobility and exocytic release by astroglia, thus inhibiting the release of pro-inflammatory mediators by this cell type, as well (143). Astrocytic gliosis also occurs in EAE and MS (Fig. 5) (71). In vitro treatment of a human astrocyte cell line with FTY720 suppressed S1P-induced production of pro-inflammatory cytokines (144). In vivo, specific deletion of astrocyte S1P₁ resulted in decreased EAE pathology and a loss of FTY720 efficacy, indicating that the primary target of FTY720 during EAE was S1P₁ specifically on astrocytes (145). Additionally, in a model of spinal cord injury, FTY720 affected the later stages of vascular permeability and astrogliosis, partially through agonism of S1P₁ (146). Another target of FTY720, S1P₃, was also found on reactive astrocytes in human MS lesions and upregulated by lipopolysaccharide stimulation of astrocytes in vitro, although it is unknown if expression of S1P₃ is protective or pathogenic in the context of MS/EAE (147). Mice deficient in the one S1PR not targeted by FTY720, S1P₂, are prone to seizures resulting in 40% mortality and have enhanced hippocampal gliosis accompanied by behavioral defects (148). Importantly, MS patients treated with fingolimod show reduced brain volume loss and lesional activity, suggesting the importance of S1PR pathways in neuroprotection (149–151).

The blood brain barrier (BBB) forms through unique interactions between brain endothelial cells, astrocyte foot-processes, and pericytes, and regulates interactions between the immune and nervous systems (152). Alterations in the BBB are implicated or present in numerous neurological diseases, including MS, stroke, and dementias (153). S1P₅ was highly expressed by human brain capillary endothelial cells, and antagonism of S1P₅ in an in vitro model of BBB decreased vascular permeability and monocytic transmigration (154). Studies of FTY720 treatment in the context of transient cerebral ischemia and reperfusion have demonstrated neuroprotection in mouse and rat models; however, these effects may be due to effects on interactions between the neurovasculature and immune cells (155, 156). FTY720 treatment reduced brain edema as well as expression of the vascular adhesion molecule, ICAM-1, resulting in decreased neutrophil infiltration (155). Additionally, when transient cerebral ischemia was induced in lymphocyte-deficient Rag1^−/− mice, the protective effect of FTY720 was lost, further implying that FTY720-mediated protection is due to effects on the neurovasculature and its interactions with immune cells (156). Conversely, a study utilizing a model of permanent cerebral ischemia demonstrated no effect on pathology with FTY720 treatment, whereas another group demonstrated efficacy after delaying FTY720 treatment for 3 days after photothrombosis induction, with increased functional capacity and decreased astrogliosis (157, 158). Thus, protection by FTY720 may be dependent on the method of ischemia induction and temporal regulation of cell activation and recruitment.

INVOLVEMENT OF S1PRs IN CANCER AND ONCOGENESIS

S1PRs have also been implicated in cancer pathogenesis, playing roles in tumor maintenance similar to their roles in maintenance of homeostasis, such as modulation of survival and proliferation (159–161). WT hamster lung fibroblasts were protected from nutrient deprivation-induced apoptosis by expression of S1P₁, which induced the anti-apoptotic protein Mcl1 via the phosphoinositide 3-kinase and PKC pathways (162). Lung adenocarcinoma cell lines respond to S1P with increased proliferation and invasion through S1P₃-mediated expression of epidermal growth factor receptor (EGFR) (163).

Estrogen receptor positive (ER⁺) breast cancer cells also responded to S1P via S1P₃ to coordinately regulate EGFR localization and signaling (164). High expression of S1P₁ or S1P₃ by ER⁺ breast cancer cells correlated with poor prognosis and high S1P₁ expression induced decreased expression of pro-apoptotic markers (165, 166). In ER⁻ breast cancer cells, S1P₄ expression activated the ERK1/2 pathway and correlated with poor prognosis (167). In vitro, several breast cancer cell lines respond to S1P or S1P₁ agonist SEW2871 with increased proliferation (168).

Another malignancy that S1P signaling may play a prominent role in is colonic inflammation and the resultant cancer (169). In a model of ulcerative colitis, considered a possible precursor for colon cancer, increased colonic bleeding and mortality resulted from S1P₁ deletion (170). In a model of colitis-associated cancer, S1P₁ signaling was necessary for persistent activation of nuclear factor-kB and signal transducer and activator of transcription3 transcription factors needed for maintaining the chronic inflammatory state and could be blocked by FTY720 treatment (171). In human colon cancer cells, expression of the chemotherapeutic resistance and cancer stem cell marker CD44 was regulated by S1P₂-induced ERK phosphorylation (172). Interestingly, FTY720 treatment impaired the mucosal immune response to the extracellular bacterium, Citrobacter rodentium, including decreased DC numbers, as well as macrophages and T cells in the colon, while increasing bacterial burden (173). These data suggest that FTY720 or other S1PR modulators could be beneficial or detrimental, depending upon how they influence the immune response.

In prostate adenocarcinoma, Sphk1-derived S1P activated AKT pro-survival pathways through activation of S1P₂ (174). AKT and Bcl-associated death promoter pro-survival pathways were also reduced by FTY720 administration to neuroblastoma cells in an in vitro and an in vivo xenograft model, resulting in decreased cancer cell viability (175).

S1PR expression in several hematological malignancies has also been described, including S1P₁ expression by classical Hodgkin’s lymphoma (CHL) cells, B cell chronic lymphocytic leukemia (B-CLL) cells, and activated B cell-like DLBCL cells (176–178). Chronic myeloid leukemia (CML) cells expressed S1P₂, which resulted in increased stability of the B cell receptor-Abl1 fusion protein and subsequently, increased proliferation (179). Expression of S1PRs by blood cancer cells may directly regulate their survival or by controlling the localization of cells within permissive environments such as the lymph nodes.

CONCLUDING REMARKS

S1PRs are gaining appreciation as powerful modulators of homeostasis and pathogenesis. In all biological systems, S1PRs play some role in regulating cell survival, migration, phenotype, activation status, and proliferation. In the current review, we have attempted to summarize the most recent advances in the field of S1PR biology and to provide novel insights into the biological responses regulated. As more cell-specific animal models of gene deletion or over-expression are created, and agonists and antagonists with greater S1PR subtype specificity are developed, further studies with such tools will clarify the contributions of specific S1PRs in each physiological or pathological context. This is especially true of the less explored members of the S1PR family, S1P₄ and S1P₅. Additionally, we anticipate that the development of more compounds for clinical use will expand our understanding of the complex signaling networks regulated by S1PRs and their role in human homeostasis and disease.