| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
REVIEWS |
a Departments of Medicine and Microbiology-Immunology, University of California Medical Center, San Francisco, California 94143–0711, USA
ABSTRACT
The lysophospholipid (LPL) mediators lysophosphatidic acid (LPA) and sphingosine 1-phosphate (S1P) are generated by enzymatic cleavage of stores of glycerophospholipids and sphingomyelin, respectively, in membranes of stimulated cells. LPLs are albumin bound, distributed widely in mammalian tissues, and increased in concentration by physiological activation of platelets and some other cells, tissue injury, inflammation, and neoplasia. The principal effects of LPA and S1P are growth related, including induction of cellular proliferation, alterations in differentiation and survival, and suppression of apoptosis. LPA and S1P also evoke cellular effector functions, which are dependent on cytoskeletal responses such as contraction, secretion, adhesion, and chemotaxis. The extracellular mediator activities of LPLs are transduced by subfamilies of G-protein-coupled receptors (GPCRs), of which the most completely characterized are those encoded by the endothelial differentiation genes (edgs). One homology cluster composed of Edg-1, -3, and -5 recognizes and responds to S1P, and the other cluster of Edg-2 and -4 is dedicated to LPA. Edg proteins are developmentally regulated and differ in tissue distribution, but couple similarly to multiple types of G-proteins to signal through ras and mitogen-activated protein kinase, rho, phospholipase C, and several protein tyrosine kinases. Numerous interactions between glycerophospholipids and sphingolipids are observed in their biosynthetic and signaling pathways. Many of the cellular effects of LPA and S1P are attributable to modifications in the content and/or activity of a major functional protein. Examples are increases in nuclear levels of transcription factors that regulate the serum response element, suppression of death caspase activities in apoptosis, and elevation of membrane content of heparin binding–epidermal growth factor-like growth factor, which serves as an autocrine and juxtacrine stimulus of proliferation. These ubiquitous LPL mediators of cellular growth, differentiation, and activities thus act directly through complex subfamilies of GPCRs and by regulating expression of biologically critical proteins.—Goetzl, E. J., An, S. Diversity of cellular receptors and functions for the lysophospholipid growth factors lysophosphatidic acid and sphingosine 1-phosphate. FASEB J. 12, 1589–1598 (1998)
Key Words: lysoglycerophospholipids • lysosphingophospholipids • G-protein-coupled receptors • signal transduction • apoptosis
LYSOPHOSPHOLIPIDS (LPLs)2 of several distinct classes initiate and regulate cellular proliferation and related functions by mechanisms resembling those characteristic of the major polypeptide growth factors (1–3). Although structurally distinct, LPL and polypeptide growth factors may both function as autocrine and juxtacrine mediators, bind productively to more than one receptor, and evoke many cellular responses in addition to those related to growth. Their similarities of cellular recognition and signaling include binding to highly specific cell-surface receptors and initiation of complex biochemical pathways of transduction that often involve protein tyrosine phosphorylation. In addition, both LPL and polypeptide growth factors elicit synthesis of transcriptionally active proteins, which bind to critical up-regulatory elements in promoters of multiple immediate-early growth-related genes. The major differences in cellular mechanisms between LPL and polypeptide growth factors are their respective uses of G-protein-coupled receptors (GPCRs) and tyrosine kinase receptors, which explains the sensitivity of many effects of the former to pertussis toxin. LPLs may function as intracellular second messengers of some growth-related effects, including those of specific polypeptide growth factors that evoke their synthesis. LPLs also have an apparently greater range of growth-unrelated physiological activities than polypeptide growth factors, which may be linked to cytoskeletal effector responses (4, 5).
This review will encompass our current understanding of biosynthesis of the LPLs lysophosphatidic acid (LPA) and sphingosine 1-phosphate (S1P) (
Fig. 1),
which are secreted as extracellular mediators. The properties of recently identified subfamilies of novel GPCRs and signal transduction mechanisms will be examined in relation to their specificity for LPA and S1P. Some unique interactions between the pathways for LPA and S1P generation and between LPA and S1P GPCR-dependent signaling will be described in relation to functional consequences. Our evolving awareness of the breadth of cellular effects of LPLs will be illustrated in terms of specific mechanisms by which LPA and S1P protect against apoptosis and alter sensitivity to some bacterial toxins.
|
METABOLIC GENERATION OF LPLS FOR EXTRACELLULAR SECRETION: DE NOVO BIOSYNTHESIS AND MOBILIZATION FROM STORED MEMBRANE LIPIDS
LPA and S1P both may be biosynthesized by cells either de novo through pathways of intermediate lipid metabolism or through stimulus-coupled liberation of the respective precursor glycerophospholipids and sphingolipids and subsequent enzymatic conversions ( Fig. 2).
The latter pathways are considered the principal sources of free and secreted LPA and S1P. The de novo metabolic generation of LPA intracellularly proceeds from the reaction of glucose-derived dihydroxyacetone phosphate with a fatty acid-CoA through NADH- or NADPH-dependent enzymatic steps. In contrast, the bulk of LPA secreted is produced in membrane microvesicles from stimulated platelets, leukocytes, or other cells ( Fig. 2) (6). The combined actions of sphingomyelinase, which enhances glycerophospholipid availability to phospholipases, and either phospholipase D (PLD) or PLC, followed by diacylglycerol kinase, lead to high levels of phosphatidic acid (PA) in the microvesicles. Secretory-type PLA2 then converts PA to LPA for release into extracellular fluids. Cell membranes contain only barely detectable levels of free LPA, presumably as a result of conversion back to PA by LPA acyltransferase and related enzymes. Extracellular fluid phase and cytoplasmic LPAs are also rapidly degraded by one or more plasma membrane-associated phosphatidate phosphohydrolases, which are of limited specificity as S1P and ceramide 1-phosphate are cleaved as efficiently as LPA (7), and a diverse group of lysophospholipases. Lysophospholipases smaller than 50 kDa, exemplified by the Charcot-Leyden crystal protein, one or more proteins in liver cells, and mouse macrophage lysophospholipases I and II degrade LPA with greater selectivity than larger lysophospholipases (8). The latter subgroup is exemplified by type IV cytoplasmic PLA2 and Ca2+-independent type VI PLA2. PLA2 and transacylase activities of the group of larger enzymes are equally or more prominent than their lysophospholipase function.
|
The pathways of generation of S1P resemble in outline those of LPA, but more is known about the cellular compartmentalization of each component of the process (2, 9). The initial intracellular step in de novo biosynthesis of sphingoid bases is condensation of a fatty acid-CoA and serine to form 3-ketosphinganine, which is reduced and converted to a dihydroceramide, all in the endoplasmic reticulum (
Fig. 2). That very little free ceramide and no free sphingosine results from this de novo pathway provides evidence for the much greater contributions of stored sphingomyelin turnover to the secreted sphingosine (S) and S1P. Sphingomyelin degradation by any of several sphingomyelinases is evident in membranes of lysosomes and endosomes and in the plasma membrane. Conversion of sphingomyelinase-liberated ceramide to sphingosine by ceramidase occurs predominantly in the same cellular compartments as sphingomyelin degradation, whereas sphingosine kinase activity, which is the rate-determining step in formation and secretion of S1P, is greatest in the endoplasmic reticulum. One lyase, expressed most prominently in the endoplasmic reticulum, degradatively inactivates S1P by cleavage into ethanolamine phosphate and trans-2-hexadecanol. The end metabolism of S1P thus is simpler than that of LPA, where biodegradation involves a series of lysophospholipases. These findings also suggest that the intracellular concentration of S1P is determined predominantly by the ratio of activities of sphingosine kinase and lyase.
THE GROWTH-RELATED AND CYTOSKELETON-DEPENDENT EFFECTS OF LPA AND S1P
Although the concentrations of LPA and S1P in body fluids normally and in disease states have not been determined systematically, S1P but not LPA attains levels up to 0.2 µM in normal plasma, and both LPA and S1P may reach respective levels of up to 10 µM and 0.5 µM in normal serum, presumably reflecting major contributions from platelets (10, 11). LPA and S1P are the principal growth factors in serum unconditioned by cellular secretions. The cellular effects of the LPLs may be considered in two categories ( Table 1).
One cluster of growth-related activities of LPA and S1P includes stimulation of proliferation, prolongation of survival, prevention and suppression of apoptosis, and alterations in differentiation (12, 13). A second group of cellular effects of LPA and S1P are directed to functions dependent on the cytoskeleton such as shape changes, aggregation, adhesion, chemotaxis, contraction, and secretion. Activation of PLD or phosphatidylinositol 3-kinase relies on biochemical signaling events linked to the second group. This second group of responses may be evoked by lower concentrations of LPLs than the growth-related cluster or may require a concentration gradient. Some complex cellular responses to LPLs may reflect participation of both types of primary responses. For example, the effects of LPLs on cellular and tissue differentiation are attributable to both changes in expression of selected genetic programs, which share signaling pathways with other growth-related effects in the first cluster, and up-regulation of surface adhesive proteins that is related by signaling pathways more closely to the cytoskeleton-dependent group of effects in the second cluster ( Table 1).
|
LPA and, in some instances, S1P, have at least four different types of effects on cellular proliferation and other growth-related functions. The first, which predominates in most types of cells, is proliferation mediated by serum response element (SRE) -driven recruitment of immediate-early response genes coupled to growth (14). The proliferation-stimulating and survival-enhancing activities of LPA and S1P may be expressed in paracrine, juxtacrine, or autocrine settings. The second is the ability of LPA or S1P to induce cellular production and secretion of one or more polypeptide growth factors. One example is the elicitation of temporally distinct production and secretion of transforming growth factors 
and ß by LPA-stimulated keratinocytes, which neutralizing antibodies demonstrated account respectively for the stimulation of proliferation and differentiation by LPA (15). The third effect is sensitization of some types of cells to the proliferation-stimulating activity of a polypeptide growth factor, as revealed by the synergistic effects of the polypeptide and a marginally active concentration of LPA. This phenomenon is most likely to be observed in types of cells such as mesangial cells, for which LPA alone has only weak activity at high concentrations (16). The fourth effect of LPA and S1P on cellular growth-related functions is inhibition of proliferation of some myelocytes (17). This inhibition of proliferation is resistant to pertussis toxin, accompanied by an increase in [cAMP]i, in contrast to the usual Gi-mediated decrease in [cAMP]i, mimicked by cAMP analogs and adenylyl cyclase activators, which act additively with LPA, and enhanced by phosphodiesterase inhibitors.
The second group of cellular effects of LPLs are dependent principally on responses of the cytoskeleton and cell-surface adhesive proteins ( Table 1). One subgroup of such functions encompasses enhanced interaction of cells with other cells and with connective tissue matrices such as aggregation of platelets, epithelial and endothelial cells, formation of adherens junctions, and concurrently increased expression of P- and E-cadherins, and cell-surface assembly of focal adhesion complexes and fibronectin. Another subgroup consists of changes in cellular shape or motility exemplified by smooth muscle contraction, vasoconstriction, neurite retraction, chemotaxis, and invasion of cellular monolayers, which may depend in part on responses of the first subgroup (18, 19). Cellular transmembrane endocytic and exocytic events constitute a distinct third subgroup of LPL effects, such as ion secretion and neurotransmitter release.
Specific biochemical prerequisites have been delineated for some of the cellular effects of LPLs. For example, stimulation of rapid isometric contraction of fibroblasts by LPA correlates temporally with phosphorylation of the regulatory light chain of myosin II, as in smooth muscle, and both are dependent on Mg2+-ATP (20). Another example of LPL effects of the second group on fibroblasts is LPA stimulation of tyrosine phosphorylation of p125 focal adhesion kinase (p125FAK), which is completely abolished by either pretreatment of the fibroblasts with cytochalasin D or their suspension in protein-free medium. In contrast, activation of p42 mitogen-activated protein (MAP) kinase by LPA is a response of the first group, which was shown to be completely independent of cytoskeletal integrity and interactions with extracellular matrices (21). The distinctiveness of biochemical prerequisites for responses of the two major groups was further emphasized by the lower concentrations of LPA needed to activate MAP kinase than to evoke phosphorylation of p125FAK and the sensitivity of only the former to pertussis toxin. In the contraction-migration subgroup of the second group of LPL effects, LPA induced constriction of cerebral arterioles and inhibited their vasodilatory responses to hypercapnia or isoproterenol; the latter action was associated with prevention of rises in [cAMP]i (22). Pertussis toxin abolished both the inhibition of rises in [cAMP]i and suppression of vasodilation, suggesting a shared dependence on Gi.
The results of numerous early investigations have provided insights into the novel biological roles of LPLs and their interactions with polypeptide growth factors, which share with LPLs many pathways for signaling nuclear events. LPLs may act as paracrine, juxtacrine, or autocrine mediators and as intracellular messengers. An example of the paracrine mechanism was the production of LPA by activated platelets and its local effects on functions of vascular constituents, including endothelial and smooth muscle cells (1). Most recently, it has been demonstrated that 2-adrenergic stimulation of adipocytes evokes generation and secretion of LPA in concentrations sufficient for paracrine stimulation of preadipocyte spreading and proliferation (23). At the other extreme, S1P acts as an intracellular messenger of the signal from platelet-derived growth factor (PDGF), but not epidermal growth factor (EGF) (24). Inhibitors of sphingosine kinase, the activity of which is the principal determinant of S1P concentration, prevented PDGF- but not EGF-induced generation of S1P, activation of MAP kinase, recruitment of cyclin-dependent kinases, phosphorylation of Crk, and stimulation of activator protein 1 (AP-1) DNA binding activity (24–26). These biochemical defects were reversed by addition of S1P, supporting the proposed second messenger role of S1P. There are relevant examples of signaling interactions between LPLs and PDGF in relation to cellular responses other than growth. Stimulation of assembly of focal adhesions and formation of actin stress fibers in fibroblasts by LPA requires tyrosine phosphorylation of p125FAK, paxillin, and p130, and is inhibited by concentrations of PDGF that disrupt actin stress fibers (27). That reciprocal regulation of effects of PDGF by S1P may also be functionally relevant was shown by the ability of extracellular S1P to disassemble actin filaments and prevent focal adhesions, thereby inhibiting arterial smooth muscle cell spreading, lamellar extension, and chemotaxis elicited by PDGF (28). The specificity of S1P action was demonstrated by the lack of effect on arterial smooth muscle cell MAP kinase or PI 3-kinase activities or proliferation evoked by PDGF.
DIVERSE SPECIFIC CELLULAR RECEPTORS FOR LPA AND S1P
The existence of GPCRs for LPLs was suggested by the ligand structural dependence of their effects, ligand-induced desensitization of the responses of some cells, and pertussis toxin inhibition of their stimulatory activity for cellular proliferation. The requirement for removal of the serum source of LPLs from cell cultures to elicit optimal responses to exogenous purified LPA and S1P, the ability of virtually every type of cell to recognize and respond to exogenous purified LPA and/or S1P, and the very high level of their unspecific binding to cultured cells have complicated the search for LPL-specific GPCRs. Two recent developments have permitted the cloning of one subfamily of highly homologous GPCRs for LPA and S1P. One was the appreciation that expression of at least one neural subset of LPA-dedicated GPCRs is developmentally regulated in mouse brain (29). Another was the successful application of SRE-based reporter constructs for functional identification (30).
Retrovirally immortalized clones of mitotically active neurons from the ventricular zone (vz) of mouse embryonic cerebral cortex, which lack superficial processes during mitosis and are the source of most cortical neurons in adult brain, selectively expressed mRNA encoding a novel GPCR (29). The cDNA encoding this GPCR initially was designated vzg-1, and renamed edg-2 based on the highest homology of its amino acid sequence with a human orphan GPCR termed Edg-1 protein, earlier found to be encoded by an immediate-early gene expressed after endothelial cell activation (31). Mouse Edg-2 protein was identified as an LPA receptor because its overexpression led to increased specific binding of LPA and LPA induced the rounded shape of mitotic vz neurons. The highest level of embryonic cerebral expression of edg-2 was in the vz, as assessed by in situ hybridization and the distribution of Edg-2 GPCRs correlated with that of mitotic activity.
LPA stimulation of cellular proliferation is dependent on induction and binding of both serum response factor (SRF) and ternary complex factor (TCF) proteins to SRE present in the promoter sequences of many growth-related genes. These observations suggest that cells transfected with luciferase-encoding plasmids bearing multiple tandem copies of a cDNA for SRE would permit expression cloning of LPA GPCRs. Double transfection of HEK293 cells with an SRE-luciferase plasmid and either pools of cDNA from a human lung library or cDNAs from a panel encoding GPCRs homologous to human Edg-1, sheep Edg-2, or one of many GPCRs of known specificity for other mediators led to identification of the human Edg-2 protein as a highly specific GPCR for LPA (30). Mammalian reporter cells expressing human Edg-2 responded to LPA and structurally related lysoglycerophospholipids, but not to other LPLs or a range of sphingolipids. Overexpression of human Edg-2 protein in CHO cells increased specific binding of 3H-LPA. Similar overexpression of the human Edg-2 protein in rat hepatoma cells of the HTC line, which show no baseline biochemical responses to LPA, allowed LPA to elicit concentration-dependent increases in [Ca2+]i in two independent assays (32).
A sequence-based search for other GPCRs specific for LPA and S1P resulted in the discovery of these activities for Edg-3, -4, and -5 proteins ( Table 2).
Human Edg-4 protein was encoded by a cDNA that had been deposited in GenBank dbEST, and is 46% identical and 72% similar in amino acid sequence to human Edg-2 protein (33). As for Edg-2 protein, Edg-4 is a GPCR for LPA. Jurkat T cell transfectants overexpressing Edg-4 protein show increased specific binding of 3H-LPA and LPA concentration-dependent activation of an SRE-luciferase reporter gene with an EC50 of 10 nM. The reporter signals transduced by Edg-4 GPCRs were about threefold higher than those from Edg-2 GPCRs. PA evoked signals approximately one-half the magnitude of those seen with LPA, perhaps resulting from partial conversion of PA to LPA, whereas no increases were elicited in either Edg-2 or Edg-4 transfectants by S1P, lysophosphatidyl-choline, -ethanolamine, or -serine. Maximal reporter signals from Edg-2 and Edg-4 GPCRs were suppressed partially by either pertussis toxin or the rho-directed clostridial C3 exotoxin and almost completely by a combination of the two toxins. In contrast to their similar ligand specificity and signaling properties, Edg-2 and Edg-4 are distributed differently in human tissues, as assessed by Northern blots ( Table 2). Edg-2 mRNA is expressed at highest levels in brain and heart tissues, at moderate levels in gastrointestinal, genitourinary and many other tissues, but not in liver. In contrast, Edg-4 mRNA is expressed at the highest level in leukocytes and at a moderate level in testicular tissues, but was not detected in brain or heart tissues.
|
A similar approach identified Edg-3 and Edg-5 (originally designated H218) as GPCRs for S1P (34). In Jurkat T cells overexpressing Edg-3 protein, S1P evoked SRE-luciferase reporter signals more than twofold greater than those for structurally related sphingoid ligands, whereas no signal above empty vector background was induced by C6-ceramide, psychosine, or LPA. In Edg-5 transfected Jurkat T cells, reporter signals of similar maximal magnitude were elicited by S1P and related sphingoids, but none were induced by C6-ceramide, psychosine, or LPA. The reporter responses to S1P in both Edg-3 and Edg-5 transfectants were detectable at 1 nM, had EC50 values of 10 nM, and were optimal at 0.1 µM. In contrast, Jurkat T cell transfectants overexpressing Edg-1 GPCRs manifested weaker reporter signals of only a mean of 1.7-fold higher than vector alone control levels with 1 µM S1P. The Jurkat T cell results were confirmed in Xenopus oocytes, where both Edg-3 and Edg-5 transduced significant increases in calcium efflux evoked by S1P and dihydro-S1P, but not the other lipids. Although early analyses of Edg-1 GPCRs in oocytes showed no detectable calcium efflux signals (34), subsequent studies in other systems established that Edg-1 GPCR is a S1P receptor (35, 36). Northern blot analyses showed strong expression of mRNA encoding both Edg-3 and Edg-5 in heart tissues, with Edg-3 in blood leukocytes, and at lower levels when otherwise widely expressed, whereas Edg-5 was also prominent in most tissues and Edg-1 was strongly expressed in most mammalian tissues.
The Edg subfamily of GPCRs for LPA or S1P and other lysosphingolipids that have been characterized structurally and functionally may be considered in two major homology clusters based on amino acid sequence identities ( Table 2). Human Edg-2 and Edg-4 exhibit 46% identity and 72% similarity of amino acid sequences, and Edg-1, Edg-3, and Edg-5 exhibit 45–60% identity of amino acid sequences. In contrast, the level of amino acid sequence identity between the two clusters is lower, in the range of 31–34%. Some common structural features, unusual in the superfamily of GPCRs, are shared by all members of the Edg protein subfamily of GPCRs. For example, all Edg proteins lack a cysteine in the first extracellular loop found in most other GPCRs. Each Edg protein also has distinguishing structural elements that have not yet been related to any aspects of ligand binding or signaling. One example for Edg-4 is the substitution of alanine for proline in the usual seventh transmembrane NPXXY sequence, which is conserved in the other Edg proteins and most GPCRs. The two clusters of Edg proteins show the next highest partial homology with the cannabinoid receptor subfamily, suggesting the possibility of a common ancestral gene (37). The genomic organization of edg genes has not been fully elucidated.
The functional properties of Edg protein GPCRs in each of the two homology clusters so far also support the fundamental biological relevance of this tentative classification ( Table 2) (37). Edg-2 and Edg-4 protein GPCRs of the first cluster bind LPA specifically and transmit numerous cellular signals in response to LPA, but not lysosphingolipids. Edg-1, Edg-3, and Edg-5 protein GPCRs of the second cluster bind and signal responses to S1P, but not LPA. Interest in the structural homology between Edg protein GPCRs and receptors for cannabinoids is intensified by findings that sn-2 arachidonylglycerol is an endogenous cannabinoid ligand (reviewed in ref 37).
Xenopus oocytes and murine cells express a second type of GPCR for LPA, termed PSP24, which is not structurally homologous to Edg protein GPCRs but shows modest homology with the GPCR for phospholipid platelet-activating factor (38, 39). PSP24 specifically transduces LPA-evoked oscillatory Cl- currents by activation of the inositol triphosphate-Ca2+ system ( Table 2). In mouse tissue, PSP24 is most strongly expressed in the central nervous system. The growth-related and other functions of PSP24 GPCRs, additional members of the subfamily, and any interactions with Edg protein GPCRs remain to be elucidated in mammals. Future studies may identify LPL recognition by some other GPCRs with structural homology to Edg proteins lower than that observed within the subfamily.
EDG PROTEIN MECHANISMS OF TRANSDUCTION OF LPL SIGNALS
The array of proteins and biochemical pathways linking occupancy of Edg GPCRs by PALs to growth-related and cytoskeletal-dependent responses of cells is as complex as that serving any subfamily of receptors ( Fig. 3).
Several unifying elements are well supported by data from many types of investigations. First, Edg GPCRs couple productively to three different types of G chains and ß/
dimers. Interactions with Gi, which accounts for the sensitivity to pertussis toxin, mediate decreases in [cAMP]i, stimulation of protein kinases that recruit the ras-raf cascade leading to activation of MAP kinases, and tyrosine kinase-dependent induction of a specific tyrosine phosphatase (40–42). Gq/11 initiates phospholipase C (PLC) activation, which liberates inositol triphosphate (IP3) to mobilize intracellular Ca2+ and is capable of activating MAP kinases directly or in some types of cells through intermediate induction by diacylglycerol (DAG) of protein kinase C (PKC) activity ( Fig. 3) (43). G12/13 initiates the rhoGTP pathways that contribute to SRE-mediated transcription and mediate the cytoskeleton-dependent functions, as well as activation of PLD and PI 3-kinase (44). The Gß/ dimers may participate directly by recruiting PLC and rhoGTP. Second, initiation of SRE-dependent transcription requires involvement of both ras and rho pathways for respective production and nuclear translocation of TCF and SRF ( Fig. 3). Third, G12/13 induction of rhoGTP-coupled activities is the sole mechanism for eliciting cytoskeleton-dependent effector responses (44). Many other aspects of these signaling cascades have not been elucidated definitively. The relative roles of Gß/ dimers and Gq/11 in stimulation of PLC and subsequent mobilization of intracellular Ca2+, a requirement for PKC in PLC recruitment of MAP kinase, the relative contributions of PLC and the rasGTP-coupled pathways to MAP kinase activation, how intracellular S1P and S1P bound to Edg-1, -3, and -5 stimulate increases in [Ca2+]i separately and in cooperation, and the dependence of PLD activation on G12/13-rhoGTP and Gq/11 mechanisms all are still open questions and appear to vary with the specific type of cell and GPCR (Fig.3).
|
There also are many types of interactions between LPA and S1P in both the biosynthetic pathways and signaling events ( Fig. 4).
S and S1P enhance generation of PA by three separate mechanisms (2, 9, 45). PLD-dependent conversion of membrane microvesicle phospholipids to PA is enhanced by sphingomyelinase conditioning of the membranes and through stimulation of PLD activity by S and S1P, as well as LPA (45). S and S1P also regulate the interconversions of PA and DAG by stimulating the reaction to and inhibiting that from PA, which favors the formation of LPA from PA and blocks breakdown of LPA to mAG ( Fig. 4A). Many cooperative and antagonistic interactions are evident between LPA and S1P ( Fig. 4B). These are exemplified by their conjoint capacity to promote increases in AP-1 transcriptional regulatory activity attributable to dimers of phosphorylated jun and fos (26). The phosphorylation of jun by JNK is evoked by S1P binding to an Edg GPCR, which recruits the rac1 and cdc42 members of the rho family capable of activating JNK. The generation of fos through SRE-induced transcription is mediated by LPA binding to an Edg GPCR that elicits production and nuclear translocation of TCF and SRF. LPA mobilization of PLC-dependent signals is also modulated by S and S1P ( Fig. 4B) (42). S and S1P suppress activation of PKC by PLC by inhibiting formation of DAG and the activity of PKC. Cooperative increases in [Ca2+]i are mediated by PLC-generated IP3 and directly by intracellular S1P.
|
REGULATION OF CELLULAR EFFECTOR PROTEINS BY LPA and S1P: AN INTEGRATING HYPOTHESIS
Many of the cellular effects of LPA and S1P appear to be transduced by Edg protein receptors, and in some settings by another subfamily of PSP24 GPCRs, but the roles of each Edg protein and its signaling pathways in any one type of response have not been defined in relation to target cell specificity. The principal effects of LPA and S1P on cells are stimulation of proliferation and promotion of the related functions of survival, suppression of apoptosis, modification of differentiation, and elicitation of a range of effector responses. Where mechanisms have been elucidated, each of these effects is attributable to changes in amount or activity of one or two cellular functional proteins ( Table 3).
Edg protein signals from LPA or S1P initiate proliferation by inducing synthesis of the transcriptional proteins SRF and TCF, which together bind to the SRE DNA sequence in promoters of many growth-related genes. LPA evokes secretion of transforming growth factor by some types of cells, resulting in autocrine and paracrine stimulation of proliferation. LPA and S1P both increase expression of plasma membrane localized heparin binding EGF-like GF (HB-EGF) in T cells, macrophages, and some other types of cells (46). The binding by HB-EGF of diphtheria toxin (DT) suppresses protein synthesis and is presumptively cytotoxic. In the absence of DT, however, HB-EGF binds to EGF receptors in order to evoke cellular proliferation principally by a juxtacrine mechanism and binds to matrix proteoglycans in order to enhance cellular adhesion and related responses.
|
Effects on differentiation appear to involve alterations in both transcription of fundamental genetic programs and expression and activities of integrins and other adhesive proteins that influence cell–cell interactions during ontogeny. Edg protein-mediated protection of some types of cells from apoptosis by LPA and S1P requires changes in the cellular levels and subcellular distribution of proteins of the BcL superfamily and/or activities of one or more caspases (47, 48). Similarly, LPL induction and regulation of a range of cellular effector functions, including adhesion, migration, signaling, and generation of other mediators, requires production, assembly, and phosphorylation of several focal adhesion proteins, including paxillin and p125FAK. PLD and PI 3-kinase also are evoked by cytoskeleton-dependent transduction of signals from Edg proteins to rho-GTP—in some instances ras-GTP ( Fig. 3)—and contribute to phospholipid metabolism. PLD generates additional PA precursor for conversion to LPA; PI 3-kinase yields phosphatidylinositol-3,4,5-triphosphate, which participates in the activation of protein kinase B, MAP kinases, and Rac-related GTPases (49–51). Thus, alterations in regulatory proteins characterize many of the growth-related and cytoskeleton-dependent activities of LPA and S1P.
PATHOPHYSIOLOGICAL IMPLICATIONS
A comprehensive discussion of the pathophysiological effects of LPL mediators in disease states is beyond the scope of this review. It seems useful, however, to consider briefly how the principal activities of LPA and S1P observed in vitro might be manifested in vivo and contribute to or protect against the basic pathogenetic processes of some diseases. The high concentration of S1P in plasma and of both LPA and S1P in serum suggests that endothelial injury, local production of LPLs by activated platelets, and extravasation of intravascular fluid will introduce LPLs into tissues at concentrations sufficient to evoke inflammation and promote angiogenesis and wound healing. These effects of LPLs would be facilitated by their attraction and activation of macrophages, which contribute an array of growth factors and other relevant cytokines. LPLs in this setting would restore tissue integrity by inducing proliferation of fibroblasts and endothelial cells directly and cooperatively with polypeptide growth factors. In contrast, the capacity of LPA and S1P to promote endothelial and smooth muscle cell proliferation could accelerate and intensify restenosis after coronary artery angioplasty. Local delivery of LPL agonists thus may be beneficial in tissue repair and angiogenesis, whereas LPL receptor antagonists could diminish excessive cellular proliferation to limit vascular sclerosis and functionally deleterious fibrosis in other conditions.
LPA and S1P suppression of apoptosis by several mechanisms predicts many potentially practical applications for LPL agonists, which may include increased life span of organs donated for transplantation, protection of cardiac myocytes injured by ischemia (52), and preservation or even regeneration of neurons in a spectrum of degenerative diseases. It is possible that stimulation of growth and suppression of apoptosis may promote development and spread of tumors. Both LPA and S1P affect malignant cells in vitro. However, there is insufficient information currently for meaningful speculation about possible roles of LPLs in cancer. All clinical avenues require a greater understanding of mechanisms of regulation of expression of LPL receptors, elucidation of their separate and interacting signaling pathways, and development of potent and bioavailable agonists and antagonists for LPL receptors.
ACKNOWLEDGMENTS
The authors are grateful to Bethann Easterly for preparation of the illustrations and editorial work. This research was supported by grant HL-31809 from the National Institutes of Health (E.J.G.) and a Grant-in-Aid from the American Heart Association (S.A.).
FOOTNOTES
1 Correspondence: Immunology and Allergy, UB8B, Box 0711, University of California Medical Center, 533 Parnassus, San Francisco, CA 94143–0711, USA. E-mail: egoetzl{at}itsa.ucsf.edu 
2 Abbreviations: AP-1, activator protein 1; edg, endothelial differentiation gene; LPA, lysophosphatidic acid; S1P, sphingosine 1-phosphate; LPL, lysophospholipid; PL, phospholipase; GPCR, G-protein-coupled receptor; PA, phosphatidic acid; S, sphingosine; FAK, focal adhesion kinase; MAP, mitogen-activated protein; PDGF, platelet-derived growth factor; EGF, epidermal growth factor; HB-EGF, heparin binding– EGF-like GF; DT, diphtheria toxin; IP3, inositol triphosphate; DAG, diacylglycerol; PKC, protein kinase C; SRE, serum response element; TCF, ternary complex factor; SRF, serum response factor; vz, ventricular zone.
3 The references represent citations of reports of some highlights in the field and are not intended to be comprehensive.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  H. Jiang, S. Rhee, C.-H. Ho, and F. Grinnell Distinguishing fibroblast promigratory and procontractile growth factor environments in 3-D collagen matrices FASEB J, July 1, 2008; 22(7): 2151 - 2160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Cho, H. H. Kim, M. J. Lee, S. Lee, C.-S. Park, S.-J. Nam, J.-J. Han, J.-W. Kim, and J. H. Chung Phosphatidylserine prevents UV-induced decrease of type I procollagen and increase of MMP-1 in dermal fibroblasts and human skin in vivo J. Lipid Res., June 1, 2008; 49(6): 1235 - 1245. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Omori, Y. Shikata, K. Sarai, N. Watanabe, J. Wada, N. Goda, N. Kataoka, K. Shikata, and H. Makino Edaravone mimics sphingosine-1-phosphate-induced endothelial barrier enhancement in human microvascular endothelial cells Am J Physiol Cell Physiol, November 1, 2007; 293(5): C1523 - C1531. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Lattin, D. A. Zidar, K. Schroder, S. Kellie, D. A. Hume, and M. J. Sweet G-protein-coupled receptor expression, function, and signaling in macrophages J. Leukoc. Biol., July 1, 2007; 82(1): 16 - 32. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Chidlow Jr., D. Shukla, M. B. Grisham, and C. G. Kevil Pathogenic angiogenesis in IBD and experimental colitis: new ideas and therapeutic avenues Am J Physiol Gastrointest Liver Physiol, July 1, 2007; 293(1): G5 - G18. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Zhang, A. Bialkowska, R. Rusovici, S. Chanchevalap, H. Shim, J. P. Katz, V. W. Yang, and C. C. Yun Lysophosphatidic Acid Facilitates Proliferation of Colon Cancer Cells via Induction of Kruppel-like Factor 5 J. Biol. Chem., May 25, 2007; 282(21): 15541 - 15549. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Brault, F. Gobeil Jr., A. Fortier, J.-C. Honore, J.-S. Joyal, P. S. Sapieha, A. Kooli, E. Martin, P. Hardy, A. Ribeiro-da-Silva, et al. Lysophosphatidic acid induces endothelial cell death by modulating the redox environment Am J Physiol Regulatory Integrative Comp Physiol, March 1, 2007; 292(3): R1174 - R1183. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-P. Xu, J. Yin, and F.-S. X. Yu Lysophosphatidic Acid Promoting Corneal Epithelial Wound Healing by Transactivation of Epidermal Growth Factor Receptor Invest. Ophthalmol. Vis. Sci., February 1, 2007; 48(2): 636 - 643. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. N. Lorenz, L. J. Arend, R. Robitz, R. J. Paul, and A. J. MacLennan Vascular dysfunction in S1P2 sphingosine 1-phosphate receptor knockout mice Am J Physiol Regulatory Integrative Comp Physiol, January 1, 2007; 292(1): R440 - R446. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Zhang, Z. Liu, and K. E. Meier Lysophosphatidic acid as a mediator for proinflammatory agonists in a human corneal epithelial cell line Am J Physiol Cell Physiol, November 1, 2006; 291(5): C1089 - C1098. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Skaznik-Wikiel, T. Kaneko-Tarui, A. Kashiwagi, and J. K. Pru Sphingosine-1-Phosphate Receptor Expression and Signaling Correlate with Uterine Prostaglandin-Endoperoxide Synthase 2 Expression and Angiogenesis During Early Pregnancy Biol Reprod, March 1, 2006; 74(3): 569 - 576. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. A. Neidlinger, S. K. Larkin, A. Bhagat, G. P. Victorino, and F. A. Kuypers Hydrolysis of Phosphatidylserine-exposing Red Blood Cells by Secretory Phospholipase A2 Generates Lysophosphatidic Acid and Results in Vascular Dysfunction J. Biol. Chem., January 13, 2006; 281(2): 775 - 781. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. L. Minnear, L. Zhu, and P. He Sphingosine 1-phosphate prevents platelet-activating factor-induced increase in hydraulic conductivity in rat mesenteric venules: pertussis toxin sensitive Am J Physiol Heart Circ Physiol, August 1, 2005; 289(2): H840 - H844. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Yun, H. Sun, D. Wang, R. Rusovici, A. Castleberry, R. A. Hall, and H. Shim LPA2 receptor mediates mitogenic signals in human colon cancer cells Am J Physiol Cell Physiol, July 1, 2005; 289(1): C2 - C11. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Yin and M. A. Watsky LPA and S1P Increase Corneal Epithelial and Endothelial Cell Transcellular Resistance Invest. Ophthalmol. Vis. Sci., June 1, 2005; 46(6): 1927 - 1933. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Yang, W. W. Zhong, N. Srivastava, A. Slavin, J. Yang, T. Hoey, and S. An G protein-coupled lysophosphatidic acid receptors stimulate proliferation of colon cancer cells through the {beta}-catenin pathway PNAS, April 26, 2005; 102(17): 6027 - 6032. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Sun, J. Nishihira, T. Yoshiki, M. Kondo, Y. Sato, F. Sasaki, and S. Todo Macrophage Migration Inhibitory Factor Promotes Tumor Invasion and Metastasis via the Rho-Dependent Pathway Clin. Cancer Res., February 1, 2005; 11(3): 1050 - 1058. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
