S1P/S1P2 Signaling Axis Regulates Both NLRP3 Upregulation and NLRP3 Inflammasome Activation in Macrophages Primed with Lipopolysaccharide

The activation of NLRP3 inflammasome is a key factor for various inflammatory diseases. Here, we provide experimental evidence supporting the regulatory role of sphingosine-1-phosphate (S1P) in NLRP3 inflammasome activation in mouse bone-marrow-derived macrophages (BMDMs), along with the S1P receptor subtype involved and underlying regulatory mechanisms. During the priming stage, S1P induced NLRP3 upregulation in BMDMs only when primed with lipopolysaccharide (LPS). In this event, S1P2, but not S1P1, was involved based on the attenuated NLRP3 upregulation with JTE013 (S1P2 antagonist) or S1P2 knockdown. During the activation stage, S1P induced NLRP3 inflammasome activation in LPS-primed BMDMs via caspase-1 activation, interleukin 1β maturation, apoptosis-associated speck-like protein containing a CARD (ASC) speck formation, and IL-1β secretion. Such NLRP3 inflammasome activation was blocked by either pharmacological inhibition or genetic knockdown of S1P2. NF-κB, PI3K/Akt, and ERK1/2 were identified as effector pathways underlying S1P/S1P2 signaling in the regulation of NLRP3 upregulation in LPS-primed BMDMs. Further, reactive oxygen species (ROS) production was dependent on the S1P/S1P2 signaling axis in these cells, and the ROS generated regulate NLRP3 inflammasome activation, but not NLRP3 priming. Collectively, our findings suggest that S1P promotes NLRP3 upregulation and NLRP3 inflammasome activation in LPS-primed BMDMs via S1P2 and subsequent effector pathways.

Sphingosine 1-phosphate (S1P), a bioactive lysophospholipid, regulates various biological activities via specific G protein-coupled receptors (S1P 1 -S1P 5 ) [5,6]. S1P is present in cells and tissues and its levels can be increased under disease conditions, including tissue fibrosis, cancer, and cerebral ischemia [7][8][9]. The latter finding indicates that increased levels of S1P may aggravate tissue damage following disease induction via S1P receptors. In the case of NLRP3 inflammasome activation, sphingosine, a precursor of S1P, induces NLRP3 inflammasome activation such as caspase-1 activation and IL-1β secretion in lipopolysaccharide (LPS)-primed peritoneal macrophages [10]. In addition to sphingolipid, lysophosphatidic acid (LPA), another lysophospholipid, induces NLRP3 inflammasome activation in LPS-primed bone marrow-derived macrophages (BMDMs), but not in normal BMDMs [11]. In addition, this finding was notable in psoriasis. in which LPA levels are increased in the lesion sites [11], further supporting the role of bioactive lysophospholipids in facilitating NLRP3 inflammasome activation in activated macrophages. However, it is still unclear whether S1P exerts similar biological effects, as is the role of specific S1P receptors.
In the current study, we found that S1P enhanced NLRP3 upregulation in BMDMs only in the presence of LPS. Further, we found that S1P 2 contributed to S1P-enhanced NLRP3 upregulation in LPS-primed BMDMs by employing either a pharmacological antagonist (JTE013) or a genetic knockdown with S1P 2 -specific siRNA. It was also found that the S1P/S1P 2 signaling axis contributed to NLRP3 inflammasome activation in LPS-primed BMDMs because the suppression of S1P 2 activity attenuated caspase-1 activation, IL-1β maturation, IL-1β secretion, and ASC speck formation. NF-κB, PI3K/Akt, ERK1/2, and reactive oxygen species (ROS) were identified as the key players in the underlying mechanism.
To induce NLRP3 inflammasome activation, BMDMs were incubated in an FBS-free growth medium overnight in the presence of the vehicle (0.1% fatty acid free bovine serum albumin, Sigma-Aldrich, St. Louis, MO, USA), primed with LPS (500 ng/mL, Sigma-Aldrich) for 4 h, and exposed to S1P (up to 1 µM, Avanti Polar Lipids, Birmingham, AL, USA) for an additional 1 h.
To antagonize S1P 1 or S1P 2 , BMDMs were serum-starved overnight in the presence of the vehicle, exposed to W146 (10 µM, Cayman, Ann Arbor, MI, USA) or JTE013 (10 µM, Cayman) for 30 min, and primed with LPS for 4 h. Cells were then exposed to S1P for 1 h. To induce the genetic knockdown of S1P 2 , BMDMs were subjected to transient transfection of S1P 2 siRNA (Dharmacon, Lafayette, CO, USA) or control siRNA (Dharmacon) with the Lipofectamine ® RNAiMAX reagent (Life Technologies) in a growth medium without serum and antibiotics for 6 h. BMDMs were recovered via incubation in a growth medium for 2 days. BMDMs were then serum-starved, primed with LPS, and exposed to S1P. S1P 2 knockdown by its siRNA was determined by quantitative real-time PCR (qPCR) analysis.

qPCR Analysis
Total RNA was extracted from BMDMs using the RNAiso plus reagent (Takara, Kusatsu, Japan) according to the manufacturer's instructions. Extracted total RNA (2 µg) was used for cDNA synthesis using a TransScript All-in-One First-Strand cDNA Synthesis SuperMix for qPCR kit (TransGen Biotech, Beijing, China). qPCR analysis was performed using the StepOnePlus TM Real-Time PCR system (Applied Biosystems, Foster city, CA, USA) and the FG Power SYBR Green PCR Master Mix (Life Technologies). GAPDH was used as a house-keeping gene to normalize the values of target genes. Table S1 presents the primer sequences used in this study.

ELISA for IL-1β Secretion
The levels of IL-1β protein in cell-free supernatants were determined using IL-1β ELISA kits (Cat#: DY401-05, R&D systems) according to the manufacturer's instructions.

Determination of Intracellular ROS
Intracellular ROS production was evaluated using a dichlorofluorescein diacetate (DCF-DA) probe (Cat#: D6883, Sigma-Aldrich). Briefly, BMDMs were washed with PBS three times and incubated with DCF-DA (5 µM) for 30 min at 37 • C. The fluorescence of cells was measured using a confocal microscope.

Statistical Analysis
All data analyses were conducted using GraphPad Prism Version 5.02 (GraphPad, La Jolla, CA, USA). Data are expressed as the mean ± S.E.M. Statistical significance was analyzed via Student's t-test between two groups and one-way ANOVA followed by the Newman-Keuls post hoc test for multiple comparisons. Statistical significance was set at a p-value less than 0.05.

S1P Enhances NLRP3 Upregulation in LPS-Primed Macrophages
We investigated whether S1P influenced the expression of NLRP3 in LPS-primed BMDMs. Cells were treated with LPS (500 ng/mL, 4 h) and then exposed to S1P (1 µM) for an additional 1 h. When LPS-primed cells were exposed to S1P, the expression of the NLRP3 protein was markedly upregulated by approximately 3-fold compared to cells treated with LPS only ( Figure 1A). However, S1P alone did not induce NLRP3 upregulation in normal BMDMs ( Figure 1A). Even when cells were exposed to S1P for a longer duration (24 h, 1 µM) or at higher concentration (5 µM, 24 h), S1P itself did not induce NLRP3 upregulation in normal BMDMs ( Figure S1). S1P-mediated NLRP3 upregulation in LPS-primed BMDMs was concentration dependent: 0.1 or 1 µM S1P induced a significant upregulation of NLRP3 compared to cells treated with LPS alone ( Figure 1B). These data clearly demonstrated that S1P enhanced NLRP3 upregulation in LPS-primed macrophages. Figure 1. S1P enhances NLRP3 upregulation in LPS-primed BMDMs. Cells were treated with LPS (500 ng/mL) for 4 h and then exposed to S1P (1 µM) for an additional 1 h. In some cases, cells were exposed to S1P for 1 h without priming with LPS. (A) Effects of S1P on NLRP3 expression in BMDMs in the presence or absence of LPS were analyzed by Western blot. (B) Concentration-dependent effects of S1P (0.01, 0.1, and 1 µM) on NLRP3 expression in LPS-primed BMDMs were analyzed by Western blot. n = 4 per group. *** p < 0.001 versus control BMDMs (Veh). ## p < 0.01, and ### p < 0.001 versus LPS-primed BMDMs.
Notably, the current study demonstrated the regulatory role of S1P in NLRP3 priming in macrophages only in primed cells. In fact, S1P itself did not induce NLRP3 upregulation in mouse BMDMs under the experimental conditions (1 µM S1P exposure for 1 h) of the current study. Further, neither longer S1P exposure (24 h) or a higher concentration of S1P (5 µM) induced NLRP3 upregulation in unprimed BMDMs. Instead, S1P exposure enhanced NLRP3 upregulation in LPS-primed BMDMs. Therefore, findings from the current study indicate that S1P potentiates NLRP3 upregulation only in primed macrophages, as also reported by an independent group, albeit partially [22]. Inhibition of sphingosine kinases, enzymes for S1P synthesis, with SKII abrogated the NLRP3 upregulation at the transcription level in primary human macrophages [22], suggesting that endogenously produced S1P was involved in NLRP3 priming. In contrast, another group reported that S1P itself induced NLRP3 upregulation in bone-marrow monocytes/macrophages (BMMs) [23,24]. S1P exposure (1 µM) for 4 h was enough to induce NLRP3 upregulation based on the contrasting findings [23]. NLRP3 was upregulated by S1P itself in BMMs, which were differentiated from mouse bone-marrow cells cultured with the L929-conditioned medium [23,24], but not in BMDMs differentiated from the same bone-marrow cells with recombinant M-CSF. Although the L929-conditioned medium con-tains large amounts of M-CSF, it can also contain other molecules that may affect cellular function. In fact, L929-conditioned medium-differentiated macrophages released larger amounts of pro-inflammatory cytokines in response to LPS than M-CSF-differentiated macrophages [25].
In addition to NLRP3 priming, S1P activates the NLRP3 inflammasome in LPS-primed BMDMs. It induced ASC speck formation, caspase-1 activation, and IL-1β maturation in primed cells without affecting the expression levels of pro-caspase-1 and pro-IL-1β. In unprimed BMMs, S1P itself activated the NLRP3 inflammasome differently by inducing the upregulation of both the precursor and activated forms of caspase-1 and IL-1β [23,24]. In LPS-primed peritoneal macrophages, sphingosine, a precursor of S1P, can induce NLRP3 inflammasome activation such as IL-1β secretion [26]. Similarly, in LPS-primed primary human macrophages or mouse BMDMs that were differentiated with M-CSF and GM-CSF, the blockade of S1P production with SKII abrogated IL-1β secretion by aluminum hydroxide (AlOH), which is a known trigger of the NLRP3 inflammasome assembly [22]. However, these two independent groups [10,22] reported contrasting results involving S1P. In LPS-primed peritoneal macrophages, S1P induced IL-1β secretion only at a very high concentration (40 µM) [10]. In LPS-primed human macrophages, 1 µM of S1P did not induce IL-1β secretion [22]. Although the current study demonstrated that S1P activated the NLRP3 inflammasome in primed BMDMs, it may be noteworthy that S1P regulated NLRP3 inflammasome activation differently depending on the type of macrophages used experimentally.
The receptor-mediated S1P signaling regulates NLRP3 inflammasome activation [23,24,27,28]. The current expression profiling of S1P receptors in BMDMs demonstrated that S1P 1 and S1P 2 were abundant. Importantly, the current study identified S1P 2 as the S1P receptor subtype contributing to S1P-enhanced NLRP3 priming in LPS-primed BMDMs using an antagonist (JTE013) and a specific siRNA. In addition to the priming event, the current study demonstrated that S1P 2 was responsible for S1P-driven NLRP3 inflammasome activation in LPS-primed BMDMs. The currently identified role of S1P 2 in the regulation of NLRP3 activity is supported by recent studies [23,24]. Suppressing S1P 2 activity with JTE013 treatment attenuated NLRP3 priming and NLPR3 inflammasome activation in S1P-treated BMMs in vitro and bile duct ligation-induced cholestatic liver injury in vivo [23,24]. In addition to S1P 2 , other receptor subtypes regulate NLRP3 activity. However, in the current study, S1P 1 , which is another highly expressed receptor subtype in BMDMs, was not associated with NLRP3 priming because its antagonist (W146) did not attenuate S1P-enhanced NLRP3 upregulation in LPS-primed BMDMs. Similarly, S1P 1 was highly expressed on BMMs, but it did not affect S1P-induced NLRP3 priming and NLPR3 inflammasome activation [24]. However, the role of S1P 1 in NLRP3 inflammasome activity is disrupted in certain cases [27,28] compared with previous [24] and current findings. In fact, genetic deletion of S1P 1 in macrophages prevented pulmonary metastasis and lymphangiogenesis via NLRP3 upregulation and IL-1β production [28]. W146 also suppressed both events in BMDMs exposed to LPS/AlOH [23]. The impact of S1P 1 on NLRP3 inflammasome activation was also reported in spinal cord injury. Intrathecal injection of SEW2871, a selective agonist of S1P 1 , resulted in mechanoallodynia in the dorsal horn of the spinal cord via NLRP3 upregulation and NLRP3 inflammasome activation [27]. In the case of S1P 3 , a previous study demonstrated that it was also highly expressed on BMMs but was not involved in S1P-induced NLRP3 priming and NLPR3 inflammasome activation [24]. Although the role of other S1P receptor subtypes (i.e., S1P 4 and S1P 5 ) in NLRP3 activity has yet to be investigated, it remains possible that such receptors participate in NLRP3 priming and activation of its inflammasome.
It is well-known that ROS serve as signal 2 in NLRP3 inflammasome activation by triggering the NLRP3 inflammasome assembly [2,29,30]. ROS also serve as signal 1 to induce NLRP3 priming [21,[31][32][33]. In addition to such a regulatory role in NLRP3 priming and activation of its inflammasome, ROS mediate the biological function of S1P 2 . Indeed, suppressing S1P 2 activity with either JTE013 or S1P 2 -specific siRNA attenuates ROS production [34][35][36], suggesting that S1P 2 triggers ROS generation. In the current study, JTE013 treatment attenuated ROS production from LPS/S1P-treated BMDMs, indicating that ROS production represents an underlying mechanism for NLRP3 priming and/or NLRP3 inflammasome activation in these cells. Results from the current study using NAC indicated that ROS participated solely in NLRP3 inflammasome activation of LPS/S1Ptreated BMDMs via IL-1β maturation.

Conclusions
In summary, the current study demonstrated that S1P enhanced both NLRP3 priming and NLRP3 inflammasome activation in LPS-stimulated macrophages. Further, S1P 2 played a pivotal role in the regulation of such events via activating pathways of NF-κB, PI3K, ERK1/2, and ROS. Based on the established roles of the NLRP3 inflammasome and S1P 2 demonstrated in independent studies involving several disease types, such as liver fibrosis [37,38], cerebral ischemia [39,40], and psoriasis [41,42], the current findings suggest a possible clue for the pathogenic mechanism underlying the role of the S1P/S1P 2 signaling axis in tissue injuries. It might also be interesting to pursue the role of this signaling axis in the regulation of macrophage polarization, which is a critical event for immune responses under various tissue injuries since NLRP3 is associated with M1 polarization of macrophages [43][44][45].