Sphingosine-1-phosphate: From insipid lipid to a key regulator

It is a great honor to be asked to write a “Reflections” article by one of the true icons of biochemistry, Herb Tabor. I felt humbled, especially since it follows many written by biochemists I admire and whose contributions have shaped major advances in biochemistry and molecular biology in the last century. Here I present my personal reflections on my adventure with the bioactive sphingolipid metabolite sphingosine-1-phosphate intertwined with those of my family life as a wife, mother, and grandmother. These reflections brought back many memories of events in my early career that played significant roles in determining the path I have taken for more than 40 years and that brought much fun and satisfaction into my life. It has been an exciting journey so far, with many surprises along the way, that still continues.

I was born in a small town in Israel near Tel Aviv called Holon that in Hebrew means "little sand." The sand dunes that Holon was built on are now gone, and it has become a large urban city. My parents, Lea and Baruch Arazi, raised me in a wonderful, loving home with an amazing older brother, Yossi, who adored me. I was always a very good student and loved math and science but had very little patience for any subject that required memorization. Throughout my childhood, I remember my big brother keeping me out of mischief and inspiring me to try to be as kind a human being as he was. He also saved our house from bursting into flames one day due to a fire I started with an unsupervised chemistry set experiment. So, although from an early age I hoped to become a scientist, these initial "experiments" did nothing to suggest that I would have a future as a chemist. In high school, I chose the math/science concentration, partly because of the very small class size with a 3:1 ratio of boys to girls. Although we had a lot of fun in class and after school, our curriculum was very rigorous to prepare us for admission to university. I was more interested in mathematics and physics than chemistry or biology, but we had an excellent chemistry teacher who went far beyond the traditional curriculum and encouraged us to carry out independent science projects. By this time, my science projects no longer ended up in flames and even received first-place awards. Although army service in Israel is mandatory for 18-year-olds, because of eye problems, I was not allowed to serve. Thus, when I was admitted to Hebrew University in Jerusalem and enrolled as an undergraduate student with a major in chemistry, I was much younger than the rest of the students. I was a bit intimidated by the prospect of competing with more mature students but quickly realized that being younger was not a major disadvantage. I tremendously enjoyed classes in physics and chemistry in my first year. But there was a black cloud on the horizon. During that summer, I had a retinal detachment and after surgery spent the next two vacation months in the hospital with friends and family reading Agatha Christie mysteries to me. If not for the encouragement of my eye doctor, I might have given up on my studies. When I took organic chemistry and physical organic chemistry, I was hooked. As the curriculum was not too demanding for me, I had time to carry out a research project in Professor Zvi Rappoport's lab characterizing mechanisms important for chemical reactions and identifying reactive intermediates. Graduation, summa cum laude with a degree in chemistry, was followed by the one of the most important events in my life, the birth of my daughter Shlomit, whose name means peace in Hebrew. I sometimes joked with friends that this stimulated the peace agreement between Israel and Egypt two years later. It also changed my perspective on science, and my interests shifted from theoretical physical organic chemistry to more biological and health-related research.

From the Weizmann Institute to the National Institutes of Health
The usual path to the Ph.D. in Israel is through the M.S. degree. Nevertheless, an outstanding GPA allows one to apply for the direct path to the Ph.D. I decided to apply to the Feinberg Graduate School at the Weizmann Institute of Science, the leading research institute in Israel. I interviewed with several distinguished professors there but was most impressed with Professor Meir Wilchek from the Biophysics Department. He was one of the superstars at the Weizmann who, while a visiting scientist at the National Institutes of Health (NIH), pioneered the development of affinity chromatography for protein purification and the avidin-biotin system, which still today have major impacts on biomedical research. To combine my inter-This work was supported by National Institutes of Health Grant R01GM043880 (to S. S.). The author declares that she has no conflicts of interest with the contents of this article. The content is solely the responsibility of the author and does not necessarily represent the official views of the National Institutes of Health. 1 Recipient of an endowed Mann T. and Sara D. Lowry Chair of Oncology.
To whom correspondence should be addressed. E-mail: sarah.spiegel@ vcuhealth.org.
ests and Meir's expertise in organic chemistry and affinity labeling, my thesis project was designed to chemically modify sialic acid residues on the surface of lymphocytes with radioactive N,4-dinitrophenyl-L-2,4-diaminobutyric acid hydrazide (DNP-DABH). 2 This labeled both glycoproteins and glycolipids with concomitant preservation of the mitogenic activity induced by cross-linking with anti-DNP antibody. To our surprise, I found that the major difference between stimulated committed and noncommitted DNP-modified lymphocytes is the amount of ligand attached to the cell-surface sialoglycolipids (gangliosides) rather than sialo-glycoproteins. This finding was not well-received by the members of the Biophysics Department, whose major research interest was glycoproteins and lectins that bind to their sugars. However, Meir was extremely supportive and was sure that we had discovered something very important (Fig. 1A). Indeed, my first paper describing the involvement of gangliosides in lymphocyte stimulation was published in the Proceedings of the National Academy of Sciences (1). Over the years, Meir often reminded me that this paper was also the first description of lipid capping, even before the introduction of the concept of lipid rafts. Meir's advice that I should always follow my gut feelings has carried me throughout my career.
Meir's lab was a wonderful place to learn, and he was a great role model. He usually came to the laboratory early in the morning and washed the dirty "dishes" from our previous day's organic reactions. Embarrassed by seeing the famous professor doing this menial chore, I finally got the courage to ask him whether I could do this instead of him. He politely refused and with a small grin explained that his best research ideas emerged when he was wash-ing dishes. For years, I tried to imitate him, but I am still waiting for my own eureka moments while doing housework.
As my graduate work had sparked a strong interest on gangliosides and glycosphingolipids, I spent a lot of time reading about their structures and functions. I vividly remember spending hours in the library competing with other students for access to the single printed copy of the Journal of Biological Chemistry, which was delivered by mail at least two months later than it was available in the United States. I started to write my thesis on the involvement of gangliosides in lymphocyte stimulation while on maternity leave with my son Michael, who brought additional light into my life, and he gave me inspiration and courage to seriously pursue a scientific career. As a single parent in graduate school, spending long hours in the laboratory was not always easy. I will always be very thankful and indebted to my parents, Baruch and Lea Arazi, who were just a phone call away whenever I needed their help. I was planning on doing a postdoctoral fellowship in the U.S. but the children were too young, so I stayed for another year at the Weizmann and continued working with Meir. I also initiated collaborative studies with Dr. Joseph (Yossi) Schlessinger, who pioneered the method of fluorescent photobleaching recovery measurement. We showed that the inserted fluorescent gangliosides that I prepared were free to diffuse in the plane of the membrane with diffusion coefficients much higher than those of proteins (2). Yossi, whose interests were in receptor tyrosine kinases for which he is very well-known and received numerous awards, used to kid me about wasting time on the little guys (sphingolipids) rather than on big guys (receptors) in membranes. I hope that, by now, even he appreciates their important biological functions. fellowship award enabled me to overcome the challenges of relocating to a foreign country, living on a single parent's salary, and uprooting my young children. I decided to work with Dr. Peter Fishman, whose research was focused on gangliosides in Dr. Roscoe Brady's branch. Roscoe was already world-renowned for pioneering work on biochemistry and the enzymatic basis of inherited defects in sphingolipid metabolism. Although the Fishman lab was small, around six people, there were many investigators in the Developmental and Metabolic Neurology Branch who shared my interest in understanding the biosynthesis and functions of gangliosides and other glycosphingolipids. Most of the clinicians set out to find the underlying defects in the devastating hereditary metabolic storage diseases called lipid or lysosomal storage disorders that include the sphingolipidoses. This line of research ultimately led to their development of successful enzyme-replacement therapies (3). I was more attracted to understanding ganglioside roles in cellular recognition, cell communication, and regulation of biological processes. The NIH was an exciting, stimulating research environment full of bright, interesting people who were eager to collaborate. The presence of experienced investigators was especially helpful for me to become resourceful and independent. I was fortunate to be able to collaborate with several well-known cell biology investigators, such as Drs. Joel Moss and Ken Yamada. Together, we used the fluorescent gangliosides that I developed previously at the Weizmann Institute (1, 4) as probes to study organization of fibronectin in ganglioside-deficient cells (5,6). We introduced a new idea that gangliosides can mediate the attachment of fibronectin to the cell surface and its organization into a fibrillar network (5). We also showed that the ability of a cell to organize fibronectin into an extracellular matrix depended on specific gangliosides, yet cell adhesion to fibronectin was independent (6). This work was the first to suggest that matrix organization and cell attachment and spreading are based on separate mechanisms and that these functions are associated with different cell-surface receptors.
Postdoctoral fellowships at NIH usually last for two years. Since my research was taking off, Peter suggested at the end of my second year that I should stay at NIH for an additional year as a visiting associate. This opened a new chapter in my life, as I met and married my sweetheart, Dr. Sheldon Milstien, who was a senior scientist at NIH, and decided to continue my career in the U.S. Shel has been a loyal, loving best friend who enriched my life immensely and was a partner in raising a wonderful family.
A major interest of the Fishman laboratory at that time was cholera toxin (CT), the causative agent of the diarrheal disease cholera. Although its structure was not yet known, it was accepted that CT was composed of an A subunit and a homopentameric B subunit with different functions. Peter's goal was to understand how the A subunit ADP-ribosylates and activates the G s protein, leading to increased adenylate cyclase activity and cAMP and its role in toxin actions (7). I, on the other hand, was fascinated by the observation that the pentameric B submit binds to ganglioside GM1. Together, we made an important discovery (published in Science (8)) that the B subunit, which is multivalent and binds exclusively to five molecules of ganglioside GM1, was mitogenic for lymphocytes. Mitogenesis depended on the direct interaction of the B subunit with GM1 on the surface of the cells. This was the first demonstration that endogenous plasma membrane ganglioside GM1 in lipid microdomains can transmit a signal across the plasma membrane to induce cell proliferation (8). The B subunit of CT labeled with a fluorescent tag is still used to identify lipid microdomains/lipid rafts. We later observed that the B subunit inhibits the growth of Ras-transformed fibroblasts, whereas untransformed cells exhibit opposing responses to the B subunit, depending on their state of growth (9). We concluded that endogenous gangliosides may be bimodal regulators of signals of cell growth and raised the possibility that other physiological processes could also be triggered by interactions with gangliosides on the cell surface.

Moving to independence and discovery of the enigmatic signaling lipid sphingosine-1-phosphate
Despite all of these studies, I still did not understand then how the signal could be transduced from the outer leaflet of the plasma membrane, where gangliosides reside, across the cytoplasm to the nucleus to regulate DNA synthesis and proliferation (10 -12). Luckily, the Department of Biochemistry and Molecular Biology at Georgetown University Medical School had an opening, and I decided, as a new assistant professor, to tackle this intriguing puzzle (13,14). Since I had received my first grant, the start-up source had disappeared into thin air. Instead, I got some old equipment from retired faculty, and my daughter Shlomit helped me set up my first small lab of 400 square feet. With help from students and Shel, we painted the whole lab a nice clean white.
I was fascinated by the idea raised by Drs. Robert M. Bell and Yusuf Hannun that the sphingolipid metabolite sphingosine might be a direct inhibitor of protein kinase C (PKC) (15), a key enzyme in signaling that was known then to play a critical role in cell growth regulation. They proposed that in addition to the well-known lipid signaling molecule diacylglycerol, which is derived from metabolism of glycerolphospholipids and stimulates PKC, sphingolipid metabolism produces the bioactive metabolite sphingosine that inhibits it. However, with one of my first rotation students, we surprisingly found that sphingosine stimulates rather than inhibits cell proliferation. Our results unexpectedly demonstrated that sphingosine acts as a positive regulator of cell growth in a fundamentally different, PKC-independent pathway (16,17). Obviously, the "big guys" in the field did not readily accept this idea, and it took some time before Al Merrill and Yusuf Hannun became my best colleague friends. Ignoring criticisms, and with the conviction that we were on the right track, we next set out to determine how sphingosine affects cell growth. In fact, we observed that it is not sphingosine itself, but, rather, it becomes rapidly converted to a unique phospholipid. Before the era of mass spectrometry (MS), thin-layer chromatography (TLC) was the main method used to separate and identify lipids. Two-dimensional TLC analysis revealed that sphingosine induces the formation of an unidentified 32 P-labeled phospholipid spot that did not co-migrate with any of the known phospholipids. After much effort, we showed that this mystery compound, which I originally nicknamed "schmutz" (Yiddish for dirt), is sphingosine-1-phosphate (S1P) (18). It was then that my career began to take off, suggesting that sometimes gold can be found even in a dirt pile ( Fig. 2A). We found that S1P is produced from sphingosine by an uncharacterized sphingosine kinase (SphK) and has much higher potency than sphingosine itself as a mitogen and in mobilization of calcium from internal stores (18). Thus, a tantalizing link between sphingolipid metabolism and cellular proliferation emerged from our discovery that the sphingolipid metabolite S1P is not only a component of the final degradative step in the metabolism of all sphingolipids, as was shown by Professor Stoffel in the early 1970s (19,20), but is also a novel bioactive sphingolipid that regulates cellular proliferation (18,21). This early study on S1P was well-received by the glycosphingolipid community, and Professor Hakomori, the guru of glycosphingolipids, invited me in 1991 to present a plenary lecture at the 11th International Symposium on Glycoconjugates (Fig.  3). Intriguingly, sphingosine was named by Thudichum in 1884 after the Greek mythological creature the Sphinx because of its enigmatic nature (22). More than a century later, we are just beginning to unravel the S1P riddle.
I have been fortunate to attract many outstanding postdoctoral fellows to my lab over the years, beginning with the talented Dr. Ana Olivera from Spain, who later became my lifelong colleague and dear friend. Ana found that potent mitogens, like platelet-derived growth factor (PDGF), stimulate SphK, leading to S1P formation. When SphK activity is inhibited, not only do S1P levels become reduced, but, importantly, DNA synthesis induced by PDGF is greatly suppressed (23). Our report in Nature on this provided the first clue to a missing link between the plasma membrane (where growth factor receptors are found) and cellular proliferation, namely S1P. We suggested that S1P has properties that make it a suitable candidate to function as a messenger in this link: It elicits diverse cellular responses; its turnover is rapid; its level in cells is low and increases transiently in response to growth factors; and it releases calcium from internal stores independently of inositol trisphosphate. This was the first suggestion that activation of SphK and S1P production play important roles in signal transduction pathways triggered by growth factors (23). My journey to understanding the paradigm that sphingolipid metabolites serve as signaling molecules, particularly S1P, which is now the most thoroughly characterized mediator in this field, was recognized in 2009 by the ASBMB Avanti Award in Lipids (Fig. 4).

The sphingolipid rheostat, sphingosine kinases, and the beginning of a long-term research partnership
In 1995, I was nominated to organize the Glycolipid and Sphingolipid Biology Gordon Conference. I was flattered since I was among the youngest in the group, and Professor Sen-itiroh Hakomori, who nominated me, was one of my idols. I enthusiastically embraced this task. In organizing this conference, I hoped to increase participation of female scientists and the focus on bioactive sphingolipid metabolites. It was a great honor for me to organize this and many other GRC and FASEB meetings, and it has been very satisfying to see how the field has grown and changed over the past 20 years, especially the large increase in female participation (see the 1998 and 2014 GRC meetings in Ventura in Fig. 5). In the early 1990s, several laboratories, particularly those of Drs. Yusuf Hannun and Richard Kolesnick, showed that another sphingolipid metabolite, ceramide, the backbone of all sphingolipids and a precursor of S1P, is an important regulatory component of apoptosis and cell growth arrest induced by stress stimuli and cytokines. Two of my postdoctoral fellows, Olivier Cuvillier from France and Burkhard Kleuser from Germany, were fascinated by those studies. Together with Dr. Silvio Gutkind at NIH, we showed that inhibition of ceramide-mediated apoptosis by activation of PKC results from stimulation of SphK and the concomitant increase in S1P (24). This led us to suggest the concept of a "sphingolipid rheostat" in which the dynamic balance between these counteracting sphingolipid metabolites and consequent regulation of different family members of mitogen-activated protein kinases are major determinants of cell fate. In other words, elevation of ceramide signals cell death, whereas S1P elevation signals cell growth and survival (24). Since then, many reports have substantiated the importance of the sphingolipid rheostat in cell-fate determination in health and diseases. Moreover, modulation of the rheostat has emerged as a focal point for treatment strategies to battle cancer.
The aforementioned results also implicate SphK as a key regulator of the "rheostat" (24 -27). Even though gene cloning today is straightforward, it was close to impossible without sequence knowledge in the old days. Therefore, we knew how important it was to purify SphK so that we could sequence, clone, and characterize this unknown enzyme. This was far from easy, and one night, when I was complaining again over dinner about our lack of progress, my husband Shel, an experienced protein chemist, finally gave in and agreed to help. He mentioned then that, since SphK appeared to be a soluble enzyme and he had purified many of them, this should take just a few weeks. Although he was overly optimistic, as it took us two years rather than weeks, this was the beginning of a wonderful collaboration that still continues. Shel, with the help of Ana Olivera and Dr. Takafumi Kohama, a visiting scientist from Japan, succeeded in purifying SphK more than 600,000-fold from rat kidneys, a workup that left the lab stinking for many weeks. After obtaining some peptide sequences from the second important "schmutz" of my career, we were able to clone SphK1 (28,29) (Fig. 2B) and, later, another isoform that we named SphK2 (30). This provided the field with a valuable tool to begin examining the biological functions of these SphKs and helped convert Shel to a sphingolipidologist, and he became my "partner in crime." As with all potent bioactive mediators, the levels of S1P are tightly regulated, not only by its synthesis, being catalyzed by SphKs, but also by its degradation. Through a collaboration with Dr. Suzanne Mandala from Merck, we cloned and characterized two specific S1P phosphatase isoforms that convert S1P back to sphingosine (31)(32)(33). This piqued Suzanne's interest in S1P and later led to a seminal drug discovery that will be discussed later.

Resolving the S1P enigma: Discovery of S1P receptors
For almost a decade, I was intrigued and puzzled by the pleiotropic actions of S1P. Not only does it stimulate growth when added to cells and enhance survival; it also induces marked cytoskeletal rearrangements and, as a consequence, regulates cell movement and invasion. These were just a few of the biological responses we studied in the laboratory. I vividly remember that my colleagues would tease my students when they presented their work by asking "What can S1P do today?" Although this was a very productive period, I felt like one of these blind men checking out an elephant in the dark (Fig. 6), as And so these men of Indostan Disputed loud and long Each in his own opinion Exceeding stiff and strong Though each was partly in the right, And all were in the wrong! We had suspected since 1995 that a G protein-coupled receptor (GPCR) might be targeted by S1P due to the sensitivity of many of its biological responses to pertussis toxin (34). Several postdoctoral fellows, particularly Dr. Jim Van Brocklyn, developed a S1P-binding assay for identifying the elusive receptor but did not succeed. After reading an important paper by Dr. Jerold Chun on the identification of a GPCR for lysophosphatidic acid (LPA, structurally similar to S1P), denoted vzg-1/ EDG-2, that belongs to the EDG receptor family (35), I called my friend Dr. Timothy Hla. I knew Tim had cloned the orphan receptor EDG-1 several years earlier (36). He told me that he had already submitted a paper to Nature showing that LPA is also the ligand for his orphan EDG-1 receptor, and his postdoctoral fellow Dr. Menq-Jer Lee had tested S1P, but it was not active in their morphogenesis assay. Since I was stubborn and we had a robust S1P-binding assay, I was able to convince Tim to send us their cells that overexpress EDG-1. The first binding assay that Jim did with these cells revealed very high specific binding of S1P to EDG-1, which told us that S1P, and not LPA, is the ligand for EDG-1. Convincing Tim of this idea required a great deal more effort, which demonstrated that S1P but not LPA could compete with binding of labeled S1P to EDG-1. Our experience adding S1P complexed with BSA enabled Hla's group to demonstrate that S1P was active in their cell-cell aggregation assay. Incidentally, Tim's original manuscript on LPA was not accepted by Nature, which allowed us to publish a joint paper in Science (37) and several others (38,39) with clear evidence that S1P is a ligand for EDG-1, now named S1PR1. In short order, we (40 -43) and others demonstrated that S1P and dihydro-S1P, which lacks the double bond in the sphingoid base, are ligands for a family of five closely related GPCRs that are now called S1PR1-5. Knowing that these S1PRs are ubiquitously expressed, it is no wonder that S1P regulates myriad physiological processes.
With the kids at home, most of my travel abroad to meetings was by myself. On the way back from a meeting in 1999 in Japan called New Frontier of Glyco-and Lipid-Biology toward the Twenty-First Century, I happened to sit next to Dr. Richard Proia from NIH. I had heard quite a bit about Rick at this meeting, since he was the expert in creating mouse models of sphingolipid storage diseases. I had 14 hours to try to convince him that he should also focus on the SphKs and S1P receptors, and this was the beginning of a long-term collaboration. Indeed, Rick's lab successfully generated these knockout mice, which are still heavily used by many investigators. The S1Pr1 knockout mice are inviable because of extensive embryonic hemorrhage due to incomplete vascular maturation. This work revealed that S1PR1 is required for blood vessel formation and that S1P signaling is essential during mammalian development (44). We also showed that S1PR1 is crucial for S1P-induced cell migration and suggested that S1PR1 is a master regulator of cell movement (44). Its physiological importance was elegantly established by the seminal work of Dr. Jason Cyster and col- leagues showing that S1P, which is high in blood (45), exerts a direct chemotactic effect on lymphocytes via the S1PR1 receptor that draws them to the blood (46) or lymph (reviewed in Ref. 47).

New mechanistic concept and additional challenges
While identification of the S1PRs was an exciting advance, it raised a fundamental question for me: What is the link between activation of SphK by growth factors and intracellular generation of S1P and the S1P receptors that are present on the cell surface? Because we observed that cell migration toward PDGF depended on SphK1 and EDG-1/S1PR1, and PDGF activated SphK1, we concluded that intracellularly produced S1P in response to PDGF must be able to leave the cell and activate its cell-surface receptors in an autocrine or paracrine manner, leading to downstream signaling pathways important for cell movement (48). We called this paradigm "inside-out signaling by S1P." Over the years, this concept has been shown to have important implications for the regulation of many physiological and pathological processes (49 -51). Yet, at the time, we were not able to detect S1P released from cells into the medium due to the lack of a suitable method with sufficient sensitivity. Only years later, when the LIPID Metabolites and Pathways Strategies (LIPID MAPS) initiative helped develop the field of sphingolipidomic MS, did it become possible to measure vanishing levels of S1P released from cells in response to stimuli (52,53).
Little did I know that our research and interaction with Dr. Suzanne Mandala would stimulate her interest in S1P and its receptors and that her research at Merck, together with Dr. Hugh Rosen, would lead to a seminal publication on the mechanism of action of the immunosuppressive agent FTY720 (fin- golimod) and the discovery that its phosphorylated form is a potent S1PR1 modulator that regulates lymphocyte circulation (54). This helped change the small S1P field, attracting many more scientists into what is now a very large community, and also led to the first orally available drug for multiple sclerosis that targets S1PRs (reviewed in Ref. 55).
Suddenly, it seemed, our children grew up and left us with an empty nest. My motherly instincts became unfulfilled and Shel convinced me that I was ready and now had time for new challenges in science. I already knew that I liked aspects of being a department chair, such as recruiting, nourishing, and helping new faculty members become established. I decided to accept a position as chair of the Department of Biochemistry and Molecular Biology at Virginia Commonwealth University (VCU) School of Medicine, also known as the Medical College of Virginia. However, fate had what seemed to be a different plan for me. After announcing that I was going to leave Georgetown University, where I had spent many wonderful, productive years, I suddenly noticed the appearance of a dark line in my vision. Within a day, I had surgery for a macular hole. Dr. Bert M. Glaser at the National Retina Institute removed the vitreous gel from my eye and placed a helium bubble there to hold the edges of the macular hole closed until it healed. I was in a facedown position for 7 weeks with my eyes covered. I talked with Dean Dr. Heber H. "Dickie" Newsome at VCU and explained why I could not accept the chair position, since I was not sure that I would regain my vision. He refused to accept my withdrawal. He was optimistic that everything would turn out fine and indicated that he was willing to wait for these two months. A lot of support from my family and many audible Agatha Christie tapes helped me get through this difficult period. I will always be thankful to Dr. Glaser, who took care of my retina for almost 30 years, for originating this treatment for macular holes and showed me the light in moments of darkness. I have never met any other physician who was so devoted to his patients, and I was heartbroken when I learned of his untimely death. Indeed, within a few months, I fully recovered and was ready for new challenges.
I had to learn new skills dealing with faculty and departmental budgets and re-organizing teaching assignments. Fortunately for me, the faculty was very welcoming. I also have had an amazing administrative staff that has been with me from almost the beginning. This, and working together with Shel, allowed me to continue putting my energy into research and leading a productive research laboratory, since I believed that department chairs should serve as a role model to their faculty. Pursuing scientific research was, and still is, the most joyful, fulfilling aspect of my career.
At the beginning, it was stimulating to have resources to recruit bright young faculty and to witness their development. Indeed, we were able to rejuvenate the Biochemistry and Molecular Biology Department and enhance research in the Massey Cancer Center (MCC) in the unique area of bioactive lipid signaling. The critical mass of investigators in the MCC Cancer Cell Signaling Program that I direct was the driving force for the creation of our Lipidomics Core, the only one of its kind in Virginia and one of the few in the entire country. This core facility is now being used by many faculty and has enhanced the collaborative spirit in the School of Medicine at VCU. However, I think that my most significant contribution to research at VCU was serving (as a very vocal member) on the search committee for recruitment of Dr. Jerome Strauss in 2005 as the Dean of the School of Medicine. Jerry, a nationally renowned researcher and a member of the National Academy of Medicine, initiated a period of visionary growth to lead our efforts over the next decade to re-establish the preeminence of the VCU School of Medicine. Jerry also cultivated talents and helped support programmatic growth in research and clinical care. He used every opportunity to promote and advance the careers of students, postdocs, faculty, and even chairs. He was probably the driving force behind my nomination as Virginia Outstanding Scientist of the Year in 2008 at the Science Museum in Richmond. I was humbled to hear the leadership of VCU and the governor of Virginia, Tim Kaine, recognize me for "pioneering work on a new lipid mediator that regulates vital physiological processes important for health and diseases." Although I was embarrassed to hear all of the praise, I was so thrilled that all of our kids, first granddaughter, and even my brother Yossi from Israel were there for the ceremony (Fig. 7,  A-C).
I was fortunate that most of my energetic group from Georgetown University moved with me to VCU, so, within 2 weeks, the lab was up and running. We decided it was important to establish the physiological relevance of "inside-out signaling by S1P." Puneet Jolly, a bright M.D./Ph.D. student, was the first to examine its role in allergy. He found that crosslinking of the high-affinity receptor for IgE on mast cells by antigen-activated SphK would lead not only to generation of S1P, but also to its extracellular secretion (56). Although activation of S1PR1 proved to be important for migration of mast cells toward antigen, it was dispensable for degranulation. However, activation of the repellant receptor S1PR2 inhibited migration toward antigen but was required for degranulation. Thus, we established that activation of SphKs, and consequently S1PRs, plays a crucial role in mast cell functions (56) and, years later, in mast cell-dependent allergic asthma (57). These yin-yang actions of S1PR1 and S1PR2 are also important for nerve growth factor-regulated antagonistic signaling pathways that modulate neurite extension development (58).
Puneet was so excited about his research that he delayed his return to his last two years of medical school. Maybe finding his sweetheart Meryem Bektas, a postdoc in my group, who later became his wife, in the laboratory also had something to do with it (59). I never considered myself a matchmaker, but there have been three couples that passed through my laboratory and later became married. It came to light that inside-out signaling by S1P in cancer has high relevance to human health. We showed that cross-talk between S1PRs and growth and angiogenic factor receptors can lead to the amplification of signals that regulate tumor invasion, vascular remodeling, and angiogenesis (60 -62). However, translating basic biochemical findings on SphKs (63,64) or S1PRs (65) into new drug-discovery programs for treatment of cancer is still much more difficult than I originally imagined.
As a co-director of the NIH-sponsored Building Interdisciplinary Research Careers in Women's Health (BIRCWH) Program at VCU, I had the opportunity to mentor the research career development of junior faculty. I was especially impressed with one of them, Kazuaki Takabe, M.D./Ph.D., and I was delighted that he joined our creative young group in 2007 and focused his research interests on the role of S1P in breast cancer. Kazu was an amazingly energetic and dedicated clinician/ scientist. In addition to spending many hours doing breast cancer surgery, he could always be found in the laboratory. This was the start of a very fruitful long-term collaboration and development of a strong friendship. Kazu brought an uncanny ability to link clinical cancer information to our basic research. He also recruited numerous outstanding clinical fellows from Japan who helped make the laboratory more vibrant and diverse (Fig. 8). We published many papers together, and Kazu developed an outstanding reputation that led to his recruitment as the clinical chief of breast surgery at Roswell Park Comprehensive Cancer Center.
Having established that inside-out signaling by S1P has important and widespread physiological and pathological functions, we wondered how S1P is exported out of cells. Mast cells initially provided us with a convenient model to tackle this question, as they readily secrete S1P. We had identified ABCC1, a member of the ATP-binding cassette (ABC) transporters, as an S1P transporter. In our first publication together, Kazu extended our previous studies and showed that estradiol induces S1P export from breast cancer cells via both ABCC1 and ABCG2 transporters (53). However, we were not sure that we had the complete answer, since the ABC transporters were known to be promiscuous transporters of lipids and drugs, and we noted that there was basal secretion of S1P unrelated to expression of these transporters. The identification of spinster 2 (Spns2), a member of the major facilitator superfamily of non-ATP-dependent transporters, as a S1P transporter in zebrafish (66) prompted us to investigate its physiological role in mammals. We found that Spns2 is indeed a bona fide transporter of phosphorylated sphingoid bases in vivo and that it regulates blood and lymph S1P levels and, consequently, influences lymphocyte trafficking (67). Subsequently, we found that Spns2 also plays a critical role in inflammatory and autoimmune diseases (68) and also contributed to a study describing endothelium Spns2 as a novel regulator of the host immune response and metastasis (69). Even after these discoveries, I believe that we still do not completely understand how Spns2 regulates S1P levels in vivo, and knowledge of its physiological functions remains fragmentary. This represents a challenging area for us and provides impetus to our future research.

Challenging dogma
By now, life was very comfortable without many worries about the kids: Sarah, the oldest, a paralegal married to Fred with a beautiful daughter, Ellis; Ben, a VP for business development in health fitness with a brilliant daughter, Annika; Shlo- mit, a psychiatrist in Israel, married to Adam with two amazing, adorable children, Keren and Nadav; and Michael, the youngest, a Ph.D. computer engineer, married to Rachel with an energetic, charming daughter, Ada. Thus, I found myself thinking more about new directions for our S1P research.
Most of the research in our field has been focused on S1P signaling through S1PRs, which accounts for both the diversity and, at times, the opposing effects of S1P on cells (70). However, for many years, it was my suspicion that S1P also possesses intracellular functions (38,71), although its targets remained elusive. This notion stemmed from the observations that although organisms such as yeast, worms, flies, and plants express the evolutionarily conserved SphKs, S1P phosphatases, and S1P lyase and although phosphorylated sphingoid bases regulate important biological functions in these organisms, they do not have S1P receptors (49). Moreover, sphingolipid metabolic enzymes are present in distinct subcellular compartments, suggesting that the location(s) where sphingosine is converted to S1P helps to dictate its functions (55). Together with my colleague Professor Tomasz Kordula, we discovered that S1P formed by SphK1 in response to TNF or IL-1 binds to TNF receptor-associated factor 2 (TRAF2) and cellular inhibitor of apoptosis 2 (cIAP2), respectively, and enhances their lysine 63-linked polyubiquitylation activities (72,73) involved in the regulation of signaling via cytokines and chemokines. Our studies suggested a new paradigm linking SphK1 and S1P to Lys-63-linked polyubiquitylation and provided a mechanistic explanation for the numerous observations of the importance of SphK1 in inflammatory, anti-apoptotic, and immune responses. Although several reports reproduced and extended this paradigm, others have challenged its validity. Clearly, resolving these discrepancies requires further investigations and exploration of possibilities to reconcile seemingly contradictory observations. These future endeavors would surely shed new light on the inflammatory roles of S1P.
I was particularly puzzled that SphK2 is also localized in the nucleus in many cell types and was motivated to uncover its functions there. Together with Dr. Nitai Hait, who worked with me for many years, we showed that S1P made in the nucleus by SphK2 inhibits histone deacetylases HDAC1/2, thus increasing histone acetylation to regulate the transcription of specific target genes (74). This was the first observation linking nuclear S1P to epigenetic regulation in response to environmental cues. Nitai later demonstrated that the prodrug FTY720 (fingolimod), an analog of sphingosine, is also phosphorylated by nuclear SphK2 in mice and accumulates in the brain (75). Like S1P, it inhibits HDACs and enhances histone acetylation and gene-expression programs associated with memory and learning, and it also represses fearful aversive memories independently of its immunosuppressive actions on S1PRs (75). It remains to be determined whether these nuclear actions of FTY720-P contribute to its beneficial effects on cognitive impairment in multiple sclerosis (76). Moreover, although independent studies have confirmed the stimulatory effects of nuclear S1P and FTY720-P on histone acetylation, it is not clear that this solely stems from the inhibition of HDACs.
It should be noted that the mechanism of action of FTY720, the gold standard for S1P-centric drugs, is not as straightfor-ward as was initially envisioned. For example, in animal models, FTY720 has broad anti-cancer activities and simultaneously prevents formation and actions of S1P. We found that S1P produced by up-regulation of SphK1 links chronic intestinal inflammation to colitis-associated cancer by increasing the NF-B-regulated cytokine IL-6, leading to persistent activation of the master transcription factor STAT3 and consequent up-regulation of its target gene S1pr1. FTY720 decreased SphK1 and S1PR1 expression and eliminated this amplification cascade and development of colitis-associated cancer (65). Moreover, in mouse breast cancer models, treatment with FTY720 reduced tumor growth and metastasis and enhanced sensitivity of hormonal refractory and triple-negative breast cancers to conventional therapies (77)(78)(79). Intriguingly, in collaboration with Professor Daniela Salvemini at Saint Louis University, we reported that neuropathic pain induced by chemotherapy (80) and cancer-induced bone pain and neuroinflammation (81) are also greatly reduced by administration of clinically relevant doses of FTY720, likely by targeting S1PR1 on astrocytes. Thus, I believe that targeting the S1P/S1PR1 signaling axis by administration of FTY720 could be an effective approach to enhance chemotherapy efficacy for treatment of triple-negative breast cancer and at the same time suppress chemotherapy-induced peripheral neuropathy (82). I hope future studies will pave the way for "fast tracking" of FTY720/fingolimod as a cancer treatment regimen. I especially recognize and greatly appreciate my enthusiastic and talented 16 graduate students and 65 postdoctoral fellows, who have made my career productive and enjoyable. I regret that, due to space limitations, I could not name each one individually and cite their contributions. Nothing has been more gratifying to me than seeing that many now have successful careers in biomedical science and are running their own laboratories in multiple countries. I know that they have shared my inspiration and fascination with the field of sphingolipid signaling. Educating, motivating, and helping with the career development of young scientists has always been my passion, and nothing has been more meaningful and rewarding in my scientific life. I was delighted to learn that they appreciated my contributions when I received the VCU Distinguished Mentor Award in 2018, and it is a great feeling to pass the torch to an REFLECTIONS: My journey with sphingosine-1-phosphate ever-expanding group of investigators in the sphingolipid field. I also found that this satisfaction outweighed the irritations of being a departmental chair. It is especially gratifying to know that the faculty that I recruited have continued to excel, and some have left for other challenges, including a chairmanship and a deanship. Moreover, the dedication of Jodi Humpage and Dr. Michael Maceyka with many budget and research administrative burdens greatly helped me maintain my own research laboratory activities.

Gratitude
It is with great satisfaction that I have watched S1P emerge as an important lipid mediator and to have been a part of that process. I am still being constantly surprised by reports of new additional physiological or pathophysiological functions that are regulated by this simple sphingolipid mediator. This is way beyond anything I could have imagined nearly 40 years ago, and I suspect that the S1P road is still long and has many twists and surprises ahead, and the enigma of the Sphinx associated with its name, despite extensive research, remains fitting. As always in science, the more we know, the more we know that we do not know.