Sphingosine-1-phosphate (S1P)/S1P1 receptor axis in macrophages: role in atherogenesis and inflammation

BACKGROUND AND AIM High density lipoproteins (HDLs) are an important factor protecting against coronary heart disease. The anti-atherogenic effect of HDLs are mainly attributed to its capacity to mediate cholesterol transport from the atherosclerotic plaque through the interaction with cholesterol transporters expressed in macrophages. In addition, HDLs inhibit inflammatory and oxidative processes within atherosclerotic plaque. Several studies documented that HDLs act as a carrier of bioactive lipid molecules such as sphingosine1-phosphate (S1P) and raised the possibility that S1P contributes to atheroprotective effects exerted by these lipoproteins. Current evidence supporting the notion that S1P, acting as a constituent of HDL, exerts anti-atherogenic effects derives primarily from experiments in vitro, whereas findings in animal models of atherosclerosis are scares and partly discrepant. In particular, the use of pharmacological agonists for amplification of S1P signaling is hampered by poorly defined side effects. In addition, potential anti-atherogenic mechanisms related to interaction of S1P with specific cells implicated into the atherosclerosis, such as macrophages, have not been examined yet. More specific approaches focusing in the endogenous S1P signaling are necessary to demonstrate in vivo beneficial effects of this sphingolipid on atherosclerosis and inflammation and to delineate the role of macrophages. Nevertheless, the S1P receptor subtype mediating the atheroprotective effects of S1P and the underlying molecular mechanisms remain enigmatic. Based on these evidences, my PhD project aim was to provide information on S1P/S1P1-receptor axis in atherosclerosis disease, investigating the effect of S1P on atherosclerotic animal models with S1P signaling specifically amplified in macrophages and/or monocytes. To this purpose, I have conducted several in vivo and in vitro studies, in order to get insights into molecular mechanisms underlying anti-atherogenic effect of S1P1 receptor and to understand the effect of its signalling on the known atherosclerosis relevant functions of macrophages, such as polarization, apoptosis and innate activation. METHODS AND RESULTS Mice overexpressing S1P1-receptor (S1P1) in macrophages and/or in monocytes were generated by Cre-LoxP technique. Mice expressing the murine S1P1 gene under a promoter containing a floxed blocking element were crossed with mice expressing Cre recombinase under the control of lysozyme (Lys2Cre) or F4/80 (Adgre1Cre) promoters. As assessed by RT-PCR, Western blot and flow cytometry, the heterozygote offspring S1pr1-LysMCre and S1pr1-F4/80Cre overexpressed S1P1 in monocytes and macrophages or only in macrophages, respectively. The Gi/o-mediated response (cAMP suppression) was enhanced in these macrophages. Taken together, these results unequivocally document that mice heterozygotes overexpress functional S1P1 receptor in the desired cell lineage and that this property is transplantable. Bone marrows from these or control mice were transplanted into Low-density lipoprotein receptor-deficient mice (Ldl-r-/- mice) and resulting chimeras were fed a Western diet (containing 0.5% of cholesterol) for 14 weeks. The following experiments have been conducted at the Universitätsklinikum of Münster (Germany) in collaboration with University Hospital of Leipzig, (Germany). Morphological studies showed a significant reduction of atherosclerotic lesion formation, both in the aortic root and brachiocephalic artery, accounting for 50% and 90% respectively, in S1pr1-LysMCre mice compared to control and non-significant reduction in S1pr1-F4/80Cre mice. Afterwards, we have tried to explain this regression by molecular and functional analysis of macrophages. To start, we analysed whole genome from these mice, using an Illumina mouse microarray. Primary analysis of these data showed an up-regulation of some genes and transcription factors important to the process of M2 polarization, such as CD163, Clec7a, Klf4, in macrophages isolated from S1pr1-LysMCre and S1pr1-F4/80Cre mice. Afterwards, the different expression of these genes has been validated also by RT-PCR. The functional macrophage phenotype has been evaluated by examining surface markers implicated in polarization process and activation and differential gene and transcription factor expression, by mean of flow cytometry and qPCR technique, respectively. Our results showed an increase of MHCII, CD206, CD115 and Dectin-1 signal in macrophages isolated from S1pr1-LysMCre and S1pr1-F4/80Cre mice and a consequent reduction of CD86 and CD93, two of M1 surface marker. The favourable effects exerted by S1P1 on cholesterol uptake and transport were likely related to the increased expression and activity of a transcription factor liver X receptor (LXR) and its targets, which is a master regulator of cell cholesterol handling. Moreover, LXR may regulate the expression of other important modulators of macrophage biology in lesions undergoing regression, such as Arg-1, IRF8, Fizz-1, Lgmn and Ym-1. In particular, Arg-1 levels are promoted through binding of the hematopoietic transcription factors IRF8 and PU.1. This evidence is confirmed also in our macrophages and monocytes that showed an increase of expression of PU.1 and IRF8 (S1pr1-LysMCre mice and S1pr1-F4/80Cre mice), evaluated by qPCR and Flowcitometry (FACs). Furthermore, to validate LXR activation, we analysed the expression of two cell surface markers, CD244 and CD226 by FACs analysis; we demonstrated that the expression of these molecules at the cell surface of macrophages, overexpressing S1P1, was significantly increased after LXR agonist treatment (22-OH Hydroxycholesterol or Desmosterol). Moreover, the effect of S1P on M2 polarization in vitro was examined also by evaluation of anti-inflammatory cyto/chemokine secretion in peritoneal macrophages and serum, such as IL-1RA, IL10, IL4, IL-5 and CCL-20. Collectively, these results showed that S1P1 overexpression, induces M2a alternative macrophage phenotype and promotes an anti-inflammatory response, by the production and secretion of above mentioned cyto/chemokines. We also evaluated the effect of S1P signalling on macrophages apoptosis and efferocytosis. To investigate apoptosis in vitro, our macrophages were exposed to factors triggering endoplasmatic reticulum (ER)-stress, such as free cholesterol overload or combined exposure to Fucoidan and Thapsigargin. Apoptosis was evaluated later as the phosphatidylserine externalization using FITC-labelled Annexin-V as indicator and determination of activity of caspase-3 and -12 with a fluorogenic substrate. Efferocytosis in vitro was studied in macrophages exposed to apoptotic RAW264.7 cells fluorescently labelled with Calcein. Moreover, we evaluated the expression of membrane receptors involved in efferocytosis, such as MerTK and Axl-1 by mean of qPCR and flow cytometry. In vivo aortic section has been retrospectively assessed for apoptotic bodies (TUNEL-positive material in lesional macrophages) and for the expression of efferocytosis and signals. Our results indicate that S1P1 overexpression promotes efferocytosis and inhibits ER-stress-induced apoptosis in macrophages. Both effects may additionally contribute to the anti-atherogenic action exhibited by S1P. In line with influencing the functional phenotype of macrophages, the S1P1 overexpression also alters monocyte functions. Mice overexpressing S1P1 in monocytes (S1pr1-LysMCre) showed expanded Ly6chi monocyte population characterized by the increased expression of proteins involved in chemotaxis and interaction with endothelial cells, such as CCR5, VLA-4 and PSGL-1, a relentless binding of VCAM1 and P-selectin, an enhanced migration towards RANTES and adhesion to TNFα-activated endothelium. These findings led to assume that a high recruitment of S1P1 overexpressing Ly6Chi monocytes to atherosclerotic lesions, is dependent on their capacity to migrate along the chemokine gradient, which thereby accounts, along with reduced apoptosis, for the increased number of macrophages in plaques from S1pr1-LysMCre mice. CONCLUSION Collectively, these results documented that the amplification of S1P1-dependent signaling in monocytes and macrophages countervails the development of vascular lesion in a murine model of atherosclerosis. The underlying molecular mechanism involves the appearance of a novel macrophage phenotype, in which the parallel activation of transcription factors PU.1/IRF8 and LXR coordinates several anti-atherogenic pathways including enhanced secretion of anti-inflammatory cytokines, cholesterol disposal and efferocytosis as well as reduced ER stress-induced apoptosis. Further investigations will be required to understand, whether and to which extent the atheroprotective mechanisms of S1P identified here may contribute to the beneficial effect of this lysosphingolipid on cardiovascular risk inferred from observational studies in humans.