Date Published: March 7, 2019
Publisher: Public Library of Science
Author(s): Franck Talmont, Lionel Moulédous, Marion Baranger, Anne Gomez-Brouchet, Jean-Marie Zajac, Clarence Deffaud, Olivier Cuvillier, Anastassia Hatzoglou, Paulo Lee Ho.
Although sphingosine-1-phosphate receptor 1 (S1P1) has been shown to trigger several S1P targeted functions such as immune cell trafficking, cell proliferation, migration, or angiogenesis, tools that allow the accurate detection of endogenous S1P1 localization and trafficking remain to be obtained and validated. In this study, we developed and characterized a novel monoclonal S1P1 antibody. Mice were immunized with S1P1 produced in the yeast Pichia pastoris and nine hybridoma clones producing monoclonal antibodies were created. Using different technical approaches including Western blot, immunoprecipitation and immunocytochemistry, we show that a selected clone, hereinafter referred to as 2B9, recognizes human and mouse S1P1 in various cell lineages. The interaction between 2B9 and S1P1 is specific over receptor subtypes, as the antibody does not binds to S1P2 or S1P5 receptors. Using cell-imaging methods, we demonstrate that 2B9 binds to an epitope located at the intracellular domain of S1P1; reveals cytosolic and membrane localization of the endogenous S1P1; and receptor internalization upon S1P or FTY720-P stimulation. Finally, loss of 2B9 signal upon knockdown of endogenous S1P1 by specific small interference RNAs further confirms its specificity. 2B9 was also able to detect S1P1 in human kidney and spinal cord tissue by immunohistochemistry. Altogether, our results suggest that 2B9 could be a useful tool to detect, quantify or localize low amounts of endogenous S1P1 in various physiological and pathological processes.
Sphingosine 1-phosphate receptor 1 (S1P1) is part of the sphingosine 1-phosphate (S1P) receptor family, which comprises five G-protein coupled receptors (GPCR, S1P1, S1P2, S1P3, S1P4, and S1P5, S1P1-5). This receptor family, firstly named, endothelial differentiation gene (EDG) family of lipid receptors, also comprises lysophosphatidic acid (LPA) receptors. S1P1-5 bind the switterionic lysophospholipid S1P, with low nanomolar affinities, share sequence, and genomic structure similarities [1–3]. S1P1 was originally detected in human umbilical vein endothelial cells (HUVEC) treated by phorbol 12-myristate 13-acetate . S1P1 signaling pathway includes coupling to the Gi/o proteins family and hence inhibition of adenylyl cyclase, activation of phosphatidylinositide 3-kinase and phospholipase C . Analysis of transcripts indicates that S1P1 is strongly expressed in adipose tissues, spleen, lung, brain, liver, and heart and poorly represented in skeletal muscle, thymus, uterus, and kidney of adult mice . When S1PR1 gene was ablated in the germ line of mice it resulted in a lethal effect in utero . In fact S1P1 has a vital role in vascular development and lethality in mice was due to a defect in blood vessels development . S1P1 has also an essential function in cell migration, in particular in the drain of T cells from the thymus to the blood and surrounding lymphoid structures . More particularly, the activation of S1P1 signaling pathway with an agonist prevents the recruitment and migration of lymphocytes to sites of inflammation by the loss of ability to perceive S1P gradient concentration. The drug FTY720 (Fingolimod, Gilenya) which activates S1P1 leading to impaired lymphocyte migration is currently used for the treatment of relapsing remitting multiple sclerosis . This drug is phosphorylated, in vivo, and the resulting FTY720-P binds to S1P1 to activate receptors as a true agonist. Nevertheless, this process leads to the internalization of S1P1 that are not recycled at the membrane thus blocking the egress of lymphocytes. S1P1 is also implicated in cancer-related processes such as neovascularization in a tumor microenvironment context, cell migration, survival, transformation and progression . Thus, the development of accurate tools for the detection, quantitation and localization of S1P1 is mandatory to understand the implication of this receptor in the regulation of numerous physiological and pathological processes. Besides commercial antibodies used by research groups, which are mainly rabbit polyclonal, generated with peptidic antigens and badly characterized, the analysis of scientific literature on S1P1 allows selecting anti-S1P1 antibodies demonstrating rather good efficacy. The murine anti-S1P1 monoclonal IgG, called E49  was produced using an Escherichia coli-derived human S1P1 full-length antigen. Another interesting antibody was the rabbit anti-S1P1 polyclonal antibody H60 raised against amino acids 322–381 of S1P1 of human origin [9, 12, 13]. Unfortunately, all these antibodies were discontinued. In this context, we have generated a murine monoclonal anti-S1P1 antibody using a purified protein produced in the methylotrophic yeast Pichia pastoris model . Mice were immunized with purified S1P1 and nine hybridoma clones secreting specific S1P1 monoclonal antibodies (MAbs) were produced. Among these, 2B9 was selected and further characterized. This antibody specifically recognizes human recombinant cmyc-S1P1 and S1P1-Green Fluorescent Protein, as well as human and mouse native S1P1s. We provide evidence that 2B9 recognizes endogenous S1P1 in murine embryonic fibroblasts (MEF), BT-549 breast cancer cell line and HUVEC cells. The binding of 2B9 to S1P1 is specific since the knocking down of the receptor in cells leads to the loss of signal. Furthermore, 2B9 was able to detect S1P1 by immunohistochemistry in human tissue. Finally, 2B9 binds to the intracellular part of the receptor, reveals cytoplasmic and membrane bound S1P1 as well as receptor internalization upon S1P and FTY720-P stimulation.
S1P1-5 are differentially expressed under various physiological and pathological conditions [30, 31]. Despite the plethora of studies describing the cellular functions of S1P receptors, tools that allow detection, quantification and localization of these receptors are unfortunately scarce. In fact, most of the currently available antibodies directed at GPCRs are not correctly validated neither for specificity nor for the detection of endogenously expressed receptors [16, 32, 33]. A challenge for GPCR receptor studies and more particularly S1P receptors studies is the specific and accurate detection of endogenous proteins. To this purpose, we developed and characterized a new monoclonal anti-S1P1 antibody, named 2B9. Based on our previous work [14, 15, 34], we expressed and purified the human S1P1 in the Pichia pastoris model, which allows a large availability of pure proteins. Immunization of mice with a SDS-solubilized S1P1 induced an immune reaction against foreign antigens and nine hybridoma cell lines producing antibodies were established. The advantage of the hybridoma model is the unlimited production of the same monoclonal antibody while polyclonal antibody production is animal-dependent. This is the case for the commonly used rabbit anti-S1P1 antibody sc-25489 (H60) from Santa Cruz Biotechnology Company. This antibody was used in a great number of S1P1 studies [9, 13, 25, 26, 28, 35–42] but is not anymore commercially available. In this work, we show that 2B9 specifically binds to S1P1 expressed either as endogenous or recombinant in various cell lines. 2B9 detected, by Western blot, recombinant mouse and human S1P1 receptors expressed in CHO and HEK cells as well as the endogenous human and mouse receptors present in cell lines of different origin, like primary cells (HUVEC and MEF) and cancer derived cells (BT-549). Variation in the molecular weight of S1P1 among cell lines can be explained at least in part by posttranslational modification such as glycosylation [43–45]. S1P1 exhibit one potential glycosylation site located at position 30 (asparagine) in the amino-terminus extra-cellular part of the receptor. Asparagine 30 N-glycosylation was formally detected in recombinant HEK cells  but not in recombinant Pichia pastoris. In general, proteins expressed in P. pastoris have shorter glycosylation chains than those expressed in Saccharomyces cerevisiae thus making P. pastoris a more attractive host for the expression of recombinant proteins.