2-carba-cyclic phosphatidic acid modulates astrocyte-to-microglia communication and influences microglial polarization towards an anti-inflammatory phenotype

2-carba-cyclic phosphatidic acid (2ccPA) suppresses microglial and astrocyte inflammation for neuronal survival following traumatic brain injury. However, it remains unknown how 2ccPA regulates microglial activation. In this study, to elucidate the 2ccPA behavior in glial communication, we collected the astrocyte conditioned media (ACM) from primary astrocyte cultures that were treated by lipopolysaccharide (LPS) and 2ccPA and analyzed the alteration of microglial inflammation caused by the ACM treatment. The addition of the ACM derived from LPS- and 2ccPA-double treated astrocytes to microglia decreased the CD86 + pro-inflammatory M1 microglia, which were upregulated with the ACM collected from astrocytes treated by LPS without 2ccPA, while the direct addition of LPS and 2ccPA to microglia failed to decrease the CD86 + microglia to the basal level. We confirmed that the ACM from LPS- and 2ccPA-treated astrocytes increased the ratio of CD206 + anti-inflammatory M2 microglia to total microglia, whereas direct treatment of microglia with LPS and 2ccPA had no effect on the CD206 + microglia ratio, demonstrating the importance of astrocyte intervention in microglial polarization. In addition, we examined whether astrocytes modulate the 2ccPA-regulated proinflammatory cytokine production derived from microglia. The addition of the ACM from LPS- and 2ccPA-treated astrocytes to microglia remark- ably canceled the LPS-induced upregulation of IL-1 β , IL-6, and TNF- α secreted from microglia, while the direct addition of LPS and 2ccPA to microglia showed no affect. Therefore, our results indicate that astrocytes mediate the 2ccPA function to shift microglia towards the M2 phenotype by interfering with the polarization of M1 microglia and to suppress cytokine production.

Alzheimer's disease [2]. TGF-β produced by astrocytes, signals microglia, and reduces the expression of some inflammatory mediators [3]. In addition, it is also known that astrocytes express various chemokines such as, CCL2 and CXCL12, and microglia express receptors for these chemokines [4,5].
Cyclic phosphatidic acid (cPA), known to regulate the activity of astrocytes and microglia, is a bioactive substance with a similar structure to that of lysophosphatidic acid (LPA) [6]. It was first isolated from the slime mold Physarum polycephalum and has since been found to be ubiquitous in numerous organisms [7,8]. In particular, cPA is present in high concentrations in the animal brain and is thought to play an important role in nerves [9]. Alternatively, 2-carba-cyclic phosphatidic acid (2ccPA) is a more stable derivative of cPA, exhibiting neuroprotective effects equivalent to or better than cPA [10]. Our previous studies have shown that 2ccPA administration suppresses proinflammatory cytokine mRNA expression levels. In addition, the administration of 2ccPA decreases the number of microglia exhibiting the proinflammatory M1 phenotype and increases the number of microglia exhibiting the anti-inflammatory M2 phenotype around the stabwounded site; thus, indicating that 2ccPA modulates the microglial polarization and shifts them towards the M2 phenotype [11]. Furthermore, the administration of 2ccPA was shown to promote the expression of an extracellular matrix protein, tenascin-C, from astrocytes and suppress the neuronal death following a stab wound injury [12]. However, it remains unknown how 2ccPA affects astrocyte-to-microglia communication.
In this study, we aimed to clarify the effects of 2ccPA on astrocyte-tomicroglia communication. To address this issue, we examined the effect of the secretions expressed from the astrocytes treated with lipopolysaccharide (LPS) and 2ccPA on microglial condition and revealed that the 2ccPA functions to modulate microglial polarization toward an anti- Table 1 Primer sequences used for realtime RT-PCR Gene Primer Sequence Gapdh (C) Immunofluorescent images of S100A10 for primary astrocytes treated with LPS and 2ccPA. Astrocytes were stained with anti-S100A10 antibody (green), anti-GFAP antibody (red), and DAPI (blue). (D) Quantification of S100A10 + ;GFAP + /GFAP + cells ratio with 2ccPA, LPS, and vehicle. Scale bar: 150 μm. *: p < 0.05, oneway ANOVA. Data represent the mean ± SEM of three independent experiments. A total of 36 newborn mice (postnatal day 2) were subjected to this experiment. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) inflammatory phenotype were enhanced by astrocyte-to-microglia communication. We showed that the secretions derived from LPS-and 2ccPA-treated astrocytes suppressed the proinflammatory cytokine production in primary microglia and shifted them towards the M2 phenotype rather than the M1 phenotype, whereas LPS and 2ccPA did not induce microglial M2 phenotype activation without astrocytes. These results suggested that 2ccPA modulates astrocyte-to-microglia communication and thereby promotes M2 polarization as well as the release of inflammatory cytokines from microglia via astrocytes.

Mice
Female ICR wild-type mice were obtained from Charles River Laboratory (Kanagawa, Japan). All animal experiments were approved by the Institutional Animal Care and Use Committee of Ochanomizu University, Japan (animal study protocols 20005, 21010, and 22007), and were performed in accordance with the guidelines established by the Ministry of Education, Science, and Culture in Japan. All mice were provided with food and water ad libitum in a clean environment.

Cell cultures, astrocyte conditioned media (ACM), and treatment with LPS/2ccPA
Cortical glial cells were cultured using a modified version of the procedure described by Milner and Campbell [14,15]. Glial cells were collected from the cerebral cortices of ICR mice on postnatal day 2. Cells were cultured in T75 flasks in Dulbecco's Modified Eagle's Medium (Merck) with 10 % fetal bovine serum, 50 units/ml penicillin and 50 μg/ ml streptomycin for 10-14 days, before being shaken to yield microglial cells. Following shaking, the media were centrifuged at 1000 rpm for 2 min to collect microglia. After removal of microglia, the remaining astrocytes in T75 flasks were detached with trypsin, and the detached astrocytes were collected using a centrifuge.
To prepare ACM, the collected astrocytes were plated into 24-well plates (2.0 × 10 5 cells/well) and cultured for 24 h to adhere in Dulbecco's Modified Eagle's Medium (Merck) with 10 % fetal bovine serum, 50 units/ml penicillin, and 50 μg/ml streptomycin. Next, astrocytes were incubated in serum-free medium (50 % Neurobasal media and 50 % DMEM with 1 mM of sodium pyruvate, 2 mM of L-glutamine, 5 μg/ml N-acetyl-L-cysteine, 100 μg/ml BSA, 100 μg/ml transferrin, 16 μg/ml putrescine, 60 ng/ml progesterone, 40 ng/ml sodium selenite, and 5 ng/ ml HB-EGF) for 24 h. Subsequently, 1 μg/ml LPS was added to astrocytes and incubated for 6 h. PBS was used as the vehicle for LPS and 0.1 % BSA in PBS was used as the vehicle for 2ccPA. Next, the astrocytes were washed once with PBS to remove LPS and changed the media with fresh serum-free media supplemented with 10 μM 2ccPA and cultured for additional 12 h. Finally, the ACM were collected, filtered through a 0.2 um-pore-size filter, and stored at − 80 ℃ until use. For immunofluorescence staining, cells were fixed with 4 % paraformaldehyde (PFA).
Bromodeoxyuridine (BrdU) (20 μM) was added to the culture of cells at 2 h before the fixation.

Immunofluorescent staining
For immunofluorescence staining of microglial cells, the cells, which were fixed with 4 % PFA, were incubated with primary antibodies overnight at room temperature, followed by secondary antibodies for 1 h at room temperature. The primary antibodies in this study included anti-Iba1 were chosen. At least five areas on each coverslip were quantified.

Statistical analyses
The values are expressed as the mean ± standard error of the mean (SEM). The significance tests were calculated using Prism GraphPad (version 9). Changes were considered statistically significant if the p-value obtained from one-way ANOVA was lower than 0.05.

2ccPA induces the polarization of astrocytes to A2 phenotype rather than A1 phenotype
Astrocytes activate to A1 phenotype, which represents proinflammatory functions, and A2 phenotype, which represents antiinflammatory functions, flexibly to regulate homeostasis in the central nervous system. We reported that 2ccPA enhances neuroprotective action in astrocytes [12]. Here, to determine whether 2ccPA modulates astrocyte polarization, we examined the effect of 2ccPA on the ratio of C3 + (A1 astrocyte marker) and S100A10 + (A2 astrocyte marker) to total GFAP + astrocytes in 2ccPA-treated astrocyte primary cultures. The LPSinduced upregulation of C3 + ratio to total GFAP + astrocytes was considerably suppressed by 2ccPA (Fig. 1A, B). In contrast, the ratio of S100A10 + astrocytes to total GFAP + astrocytes was not changed by LPS treatment, but LPS and 2ccPA treatment remarkably increased it (Fig. 1C, D). Moreover, we examined the astrocyte purity, which was quantified using immunofluorescent staining for GFAP and Iba1, indicating that the percentage of microglial cells to total cells in the astrocyte primary culture was 2.2 ± 2.7 % (Fig. S1).
Furthermore, we examined the effects of 2ccPA on astrocyte proliferation and cell death. For proliferation analysis, we performed immunofluorescent staining for Ki67 (a cell cycling marker) and 2 h-exposed BrdU (cell cycle S phase marker). In primary astrocyte culture, the astrocytes in cell cycle were rare, and LPS and 2ccPA treatment did not significantly affect the ratio of Ki67 + and BrdU + cells to total DAPI + cells ( Fig. S2A-C). For cell death analysis, we used immunofluorescent staining for cleaved caspase 3 (cell death marker). Therefore, there was no significant difference in the ratio of cleaved caspase 3 + cells to total GFAP + astrocytes by the treatment of LPS and 2ccPA (Fig. S2D, E). Moreover, we examined whether 2ccPA affects astrocyte activation by GFAP and S100B, which are expressed according to activation state. As a result, LPS and 2ccPA significanly decreased the Gfap mRNA level, however did not affect the S100b mRNA level compared with that of LPS treated astroytes (Fig. S3A, B).

Astrocyte enhances the 2ccPA functions for the microglial polarization towards M2 phenotype
To determine whether 2ccPA modulates astrocyte-to-microglia communication, we examined the effect of secreted factors from LPSand 2ccPA-treated primary astrocytes, which were prepared as ACM, on the ratios of CD86 + (M1 microglia marker) and CD206 + (M2 microglia marker) microglia to total Iba1 + microglia. The upregulation of the CD86 + ;Iba1 + ratio induced by ACM of LPS-treated primary astrocytes was significantly suppressed by the addition of ACM from LPS-and 2ccPA-treated primary astrocytes ( Fig. 2A, B). Furthermore, the addition of ACM from LPS-treated primary astrocytes did not affect the ratio of CD206 + ;Iba1 + cells, but the addition of ACM from LPS-and 2ccPAtreated primary astrocytes considerably increased it (Fig. 2 C, D). Next, we investigated the effect of direct treatment of LPS and 2ccPA to microglial primary cultures on the microglial polarization. The LPSinduced upregulation of the ratio of CD86 + ;Iba1 + to total Iba1 + microglia was decreased by the treatment of LPS and 2ccPA (Fig. 3A, B). Conversely, the direct addition of LPS and 2ccPA to microglia did not affect the ratio of CD206 + ;Iba1 + to total Iba1 + microglia (Fig. 3C, D). Therefore, it is suggested that astrocytes communicate with microglia via secretion factors and enhances the 2ccPA functions for the microglial polarization towards M2 anti-inflammatory phenotype. Then, we compared the effects between 2ccPA direct addition and ACM from 2ccPA-treated astrocytes on the mRNA levels of microglial polarization marker. The upregulation of the Cd86 mRNA levels, which was induced by the ACM from LPS-treated astrocytes was considerably suppressed by The expression level in microglia treated with vehicle and ACM from vehicletreated astrocytes was regarded as 1, respectively. Then, all other mRNA levels were normalized relative to this value. *: p < 0.05, one-way ANOVA. Data represent the mean ± SEM of at least five independent experiments. A total of 48 newborn mice (postnatal day 2) were subjected to this experiment. the treatment of ACM from LPS-and 2ccPA-treated astrocytes (Fig. 4A), and the direct treatment of LPS and 2ccPA also did (Fig. 4B). The ACM from LPS-and 2ccPA-treated primary astrocytes did not affect the Cd206 mRNA level, and direct 2ccPA treatment had no effects on the Cd206 mRNA level in microglial primary cultures (Fig. 4C, D). Furthermore, the treatment with ACM, LPS, and 2ccPA had no affect on the microglial cell death, which was quantified as the ratio of cleaved caspase 3 + microglia to total Iba1 + microglia (Fig. S4).

2ccPA suppresses cytokines secretion from microglia by interaction with astrocytes
To clarify whether 2ccPA suppresses the expression of cytokines in microglia through astrocyte to microglia communication, we analyzed the effect of ACM from 2ccPA-treated astrocytes on the expression of pro-inflammatory cytokines in primary microglia. The upregulation of Il-1β, Tnf-α, and Il-6 mRNA levels, which was induced by the ACM from LPS-treated astrocytes was drastically suppressed by the addition of ACM from LPS-and 2ccPA-treated astrocytes at basal levels ( Fig. 5A-C). Interestingly, the direct treatment of LPS and 2ccPA to microglia suppressed the Il-1β mRNA level by approximately 40 % of LPS-induced level and did not affect the Tnf-α and Il-6 mRNA level (Fig. 5D-F). To confirm the the secretion of inflammatory cytokines at protein level, the concentrations of IL-1β, TNF-α, and IL-6 in MCM were quantified using ELISA. First, the upregulation of IL-1β, TNF-α, and IL-6 induced by the ACM from LPS-treated astrocytes was drastically suppressed by the ACM from LPS-and 2ccPA-treated astrocytes (Fig. 6A, C, E). In contrast, the direct treatment of LPS and 2ccPA did not suppress the LPS-induced upregulation of IL-1β, TNF-α, and IL-6 concentration in MCM (Fig. 6B, D, F). Therefore, it is suggested that the secretions from 2ccPA-treated astrocytes include factors that suppress the microglial cytokine secretion.

Discussion
In this study, we revealed that the mechanisms of 2ccPA for microglial polarization toward the anti-inflammatory M2 phenotype are supported by astrocyte-to-microglia communication. First, we clarified that 2ccPA induced astrocyte activation into an anti-inflammatory phenotype, while inhibiting the activation into a pro-inflammatory phenotype. Moreover, we discovered that the ACM from LPS-and 2ccPA-treated astrocytes suppressed the primary microglial activation into pro-inflammatory M1 phenotype and increased the microglial polarization into M2 phenotype. Interestingly, the direct treatment of 2ccPA to primary microglia did not induce the M2 polarization. Consistent with these results, the ACM from LPS-and 2ccPA-treated astrocytes remarkably suppressed the secretion of proinflammatory cytokines such as IL-1β, TNF-α, and IL-6; however, the direct treatment of LPS and 2ccPA to primary microglia did not inhibit the proinflammatory cytokines secretion. These results indicated that 2ccPA regulates microglial polarization toward an anti-inflammatory phenotype and pro-inflammatory cytokines production by interacting with astrocytes.
In our previous study, to investigate the effect of 2ccPA on neuroprotection following traumatic brain injury, we prepared traumatic brain injury model mice using a stab wound injury in the cortices and intraperitoneally administered 2ccPA into the mice [11]. We found that the 2ccPA has pharmacological functions to promote the microglial polarization toward anti-inflammatory M2 phenotype and suppress inflammatory cytokine mRNA expression [11]. However, it is unknown how 2ccPA administration regulates microglial polarization.
We showed that the function of 2ccPA to promote M2 microglial polarization is enhanced by interacting with astrocytes. In this study, we indicated that the secretions produced from LPS-and 2ccPA-treated astrocytes include factors to promote microglial polarization towards the CD206 + anti-inflammatory M2 phenotype rather than CD86 + proinflammatory M1 phenotype (Fig. 2), whereas the treatment of 2ccPA did not contribute to M2 phenotype polarization without astrocytes (Fig. 3D). These results support the idea that astrocytes are required for exercising the anti-inflammatory functions of 2ccPA in primary microglia. The result shows that the secretions from 2ccPA-treated primary astrocytes do not promote Cd206 mRNA expression (Fig. 4C), although CD206 + cells are upregulated by the treatment of ACM from LPS-and 2ccPA-treated astrocytes, which may have occurred because Cd206 mRNA expression could be peaked out within 12 h incubation after LPS removal, and CD206 protein remains at high levels. In addition, it is possible that the classification of microglia into M1 and M2 phenotype is oversimple for identifying 2ccPA functions on microglial inflammation. In this study, the direct treatment of LPS and2ccPA to microglia decreased CD86 + M1 microglia ratio without reducing the proinflammatory cytokine secretion (Fig. 3B, Fig. 6B, D, F), suggesting that microglia other than M1 phenotype produce actively proinflammatory cytokines. In addition, the ACM derived from LPS-and 2ccPA-treated astrocytes promoted CD206 + M2 microglia polarization and reduced the pro-inflammatory cytokine secretion (Fig. 2D, Fig. 6A, C, D), whereas LPS and 2ccPA itself did not change the M2 ratio (Fig. 3D). Thus, it is possible that M2 microglia regulate the production of pro-inflammatory cytokines in a significant way.
We newly reveal the 2ccPA effects on astrocyte polarization. It has been proposed that astrocytes are activated in a disease state and Fig. 6. Intervention of astrocytes for 2ccPA-regulated proinflammatory cytokines expression in primary microglia cultures. (A, C, E) Concentration of IL-1β, TNF-α, and IL-6 in MCM from microglia stimulated by the ACM derived from LPS-and 2ccPA-treated astrocytes. (B, D, F) Concentration of IL-1β, TNF-α, and IL-6 in MCM from microglia stimulated by direct treatment of LPS and 2ccPA. *: p < 0.05, one-way ANOVA. Data represent the mean ± SEM of four independent experiments. A total of 36 newborn mice (postnatal day 2) were subjected to this experiment. converted to either a neurotoxic A1 phenotype or neuroprotective A2 phenotype [24]. Our results show that 2ccPA significantly decreased the ratio of C3 + A1 astrocytes to astrocytes and increased the ratio of S100A10 + A2 astrocytes to astrocytes (Fig. 1). Therefore, it is suggested that 2ccPA regulates astrocyte towards to A2 phenotype polarization rather than A1 phenotype, which may promote astrocytes to secrete anti-inflammatory factors. In addition, we analyzed the effects of LPS and 2ccPA on astrocyte activation using Gfap and S100b mRNA expression, indicating that LPS and 2ccPA treatment does not affect it (Fig. S2, S3). Various studies show that LPS activates astrocytes [25,26]. This discrepancy may be caused by the fact that the mRNA would peak out within 12 h after removing LPS. This notion is supported by our previous study, which reported that the astrocyte activation marker, Tnc mRNA, expression is peaked out at 6 h and decreased at 12 h after the LPS treatment [12]. Furthermore, our results demonstrate that 2ccPA contributes to astrocyte polarization into the neuroprotective A2 phenotype without affecting proliferation and apoptosis, suggesting that 2ccPA has few adverse effects on primary astrocytes.
What is the molecular mechanism of 2ccPA on neuroprotective astrocyte polarization? Present study does not identify the receptor of 2ccPA in primary astrocytes. It has been known that 2ccPA binds to lysophosphatidic acid (LPA) receptors as a ligand [16], since chemical formula of 2ccPA is similar to that of LPA [13]. Interestingly, 2ccPA has an inhibitory effect on the LPA producing enzyme, autotaxin [6], even though the chemical structures of 2ccPA and LPA are similar. Moreover, the LPA expression is is known to upregulate in injured brain and contributes to inflamatory response [27], suggesting that 2ccPA may suppress the effect of LPA in astrocytes by inhibiting LPA productions. In addition, it has been reported that S100A10 has the function of suppressing the expression of various inflammatory cytokines [28,29]. We observed that the upregulation of S100A10 in astrocytes was induced by 2ccPA treatment (Fig. 1D), suggesting that 2ccPA may decrease the production of pro-inflammatory cytokines by increasing S100A10 expression.
What are the candidate factors in the ACM derived from 2ccPAtreated astrocytes, that contribute to the M2 microglial polarization of microglia? Our results showed that 2ccPA promotes astrocyte polarization toward the A2 phenotype. It is known that A2 astrocytes secrete various anti-inflammatory factors such as TGF-β, thrombospondin, and estrogen [28], but it has not been fully revealed how the secreted factors from astrocytes alter the phenotype of microglia. A1 astrocytes secrete neurotoxic factors such as complement component C3 and proinflammatory cytokine IL-1α, while A2 astrocytes secrete neurotrophic factors [28]. The results of this study showed that the addition of 2ccPA inhibits the expression of C3 (Fig. 1B). It is reported that C3 secreted from astrocytes induces microglial activation via C3 receptor [30], suggesting that lower C3 secretion from astrocytes, which is caused by 2ccPA, might involve microglial M2 polarization. In addition, other candidates that serve as ligands promoting the polarization of M1 microglia include IFN-γ, and TNF-α [31], while for M2 polarization include TGF-β, IL-3 [32], IL-4, IL-13 [33,34], IL-21 [35], IL-33 [36], as well as chemokines CCL2 and CXCL4 [37,38]. It is necessary to perform further studies for identification of specific mechanisms by which 2ccPA regulate microglial state through astrocyte activation.

Conclusion
In conclusion, we demonstrated that astrocyte-to-microglia communication contributes to the mechanical system of 2ccPA for microglial polarization toward the anti-inflammatory M2 phenotype. Considering the dynamics of glial communication, our study indicates that the therapeutic effects of potential drugs, including 2ccPA, must be investigated.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
No data was used for the research described in the article.