Cellular phosphatidic acid sensor, α-synuclein N-terminal domain, detects endogenous phosphatidic acid in macrophagic phagosomes and neuronal growth cones

Phosphatidic acid (PA) is the simplest phospholipid and is involved in the regulation of various cellular events. Recently, we developed a new PA sensor, the N-terminal region of α-synuclein (α-Syn-N). However, whether α-Syn-N can sense physiologically produced, endogenous PA remains unclear. We first established an inactive PA sensor (α-Syn-N-KQ) as a negative control by replacing all eleven lysine residues with glutamine residues. Using confocal microscopy, we next verified that α-Syn-N, but not α-Syn-N-KQ, detected PA in macrophagic phagosomes in which PA is known to be enriched, further indicating that α-Syn-N can be used as a reliable PA sensor in cells. Finally, because PA generated during neuronal differentiation is critical for neurite outgrowth, we investigated the subcellular distribution of PA using α-Syn-N. We found that α-Syn-N, but not α-Syn-N-KQ, accumulated at the peripheral regions (close to the plasma membrane) of neuronal growth cones. Experiments using a phospholipase D (PLD) inhibitor strongly suggested that PA in the peripheral regions of the growth cone was primarily produced by PLD. Our findings provide a reliable sensor of endogenous PA and novel insights into the distribution of PA during neuronal differentiation.


Introduction
Phosphatidic acid (PA) is a phospholipid that regulates many cellular events, including phagocytosis [1], adhesion/migration [2], proliferation [3], and differentiation [4], as a lipid second messenger. PA is produced by multiple pathways, such as the phosphorylation of diacylglycerol by diacylglycerol kinase (DGK) [5][6][7][8], the hydrolysis of phosphatidylcholine (PC) by phospholipase D (PLD) [9], and the transacylation of lysoPA (LPA) by LPA acyltransferase (LPAAT) [10]. DGK [11,12], PLD [9,13,14] and LPAAT [15,16] are involved in the pathogenesis of a wide variety of diseases, such as cancer, epilepsy, obsessive-compulsive disorder, bipolar disorder, neurodegenerative disorders (Parkinson's and Alzheimer's diseases), autoimmunity, cardiac hypertrophy, hypertension and type II diabetes. Consequently, sensing PA produced by these enzymes in cells is important for understanding diverse biological and pathological phenomena. However, most PAbinding domains (PABDs) reported to date exhibit their own subcellular localization to membranes, including the Golgi apparatus and plasma membrane, in a cell stimulation (PA generation)-independent manner [17,18]. The cell stimulation-independent localization to membranes diminishes their functions as PA sensors. Because a reliable and widely applicable PA sensor has not been established to date, developing an excellent PA sensor is an urgent issue.
Recently, we determined that α-synuclein (α-Syn) selectively and intensely interacted with PA in vitro [19] and that the N-terminal region of α-Syn (α-Syn-N) is a PABD [20]. Notably, α-Syn-N did not show significant membrane localization in quiescent cells [20]. In addition, α-Syn-N colocalized with various overexpressed PA-generating enzymes, such as DGKβ, phorbol ester-stimulated DGKγ, myristoylated (Myr)-DGKζ and PLD2, but not with a phosphatidylinositol 4,5-bisphosphate-producing enzyme, phosphatidylinositol 4-phosphate 5-kinase, in an activity-dependent manner [20]. These results indicate that α-Syn-N can bind to intracellular PA, but not the PA-producing enzyme proteins themselves. Therefore, α-Syn-N can be used as a reliable and widely applicable PA sensor in cells. However, whether α-Syn-N can sense physiologically produced, endogenous PA remains unclear.
In the present study, we first established an inactive PA sensor (α-Syn-N-KQ) as a negative control by replacing all (eleven) lysine residues with glutamine residues. We next confirmed that α-Syn-N, but not α-Syn-N-KQ, recognized PA in macrophagic phagosomes where PA is known to be enriched [1,17], further indicating that α-Syn-N can be used as a reliable PA sensor for endogenous PA in cells.
PA produced by PLD and DGK is linked to neurite outgrowth [21][22][23][24][25]. However, where PA exists in neurites is not clear. Therefore, we investigated the subcellular localization of PA during neuroblastoma cell differentiation using α-Syn-N.

Protein expression and purification
The expression and purification (Ni 2+ -NTA affinity chromatography) of 6 × His-SUMO-tagged proteins were performed as described previously [20]. For the liposome co-sedimentation assay, the purified proteins were dialyzed in HEPES buffer (25 mM HEPES, pH 7.4, 100 mM NaCl). The protein concentration was measured with a bicinchoninic acid protein assay kit (Thermo Fisher Scientific).

RAW264 cell culture and phagocytosis assay
RAW264 macrophage cells were maintained in Dulbecco's modified Eagle's medium (DMEM) (Wako Pure Chemical Industries) supplemented with 10% fetal bovine serum (Biological Industries, Beit-Haemek, Israel), 100 units/ml penicillin, and 100 μg/ml streptomycin (Wako Pure Chemical Industries) at 37°C in an atmosphere containing 5% CO 2 . The cells were transfected using Lipofectamine 3000 (Thermo Fisher Scientific) as described by the manufacturer. Opsonized beads used for phagocytosis induction were prepared by incubating 3.0 μm polystyrene latex beads (Sigma-Aldrich) and 2 mg/ml normal rabbit IgG (Wako Pure Chemical Industries) at room temperature, and then incubating with goat anti-rabbit IgG-Alexa Fluor 594 (A11012, Thermo Fisher Scientific). The prepared opsonized beads were added to the dish containing RAW264 cells cultured for 24 h after transfection; the cells were incubated for 10 min at 37°C and fixed by 4% paraformaldehyde.

Neuro-2a cell culture and induction of neurite outgrowth
Neuro-2a neuroblastoma cells were maintained under the same conditions as RAW264 cells. The cells were transfected using Polyfect reagent (Qiagen, Venlo, the Netherlands) as described by the manufacturer. To induce neuronal differentiation, Neuro-2a cells were cultured in serum-free DMEM for 48 h and then fixed by 4% paraformaldehyde.

Confocal laser scanning microscopy
The cells were processed as described previously [27] and were observed using an Olympus FV1000-D (IX81) confocal laser scanning microscope (Olympus, Tokyo, Japan). Images were acquired using FV-10 ASW software (Olympus). The analysis of protein accumulation was performed using ImageJ software (National Institutes of Health, Bethesda, MD, USA) as previously described [20].

Statistical analysis
Data are represented as the means ± SD or ± SEM and were analyzed by Student's t-test, Kolmogorov-Smirnov test or Kruskal-Wallis followed by Dunn's multiple comparison test for multiple comparisons using GraphPad Prism 8 (GraphPad software, San Diego, CA, USA) to determine any significant differences. P < 0.05 was considered significant.

α-Syn-N detects PA generated by PLD in neuronal growth cones
We next determined what enzyme(s) contributed to the PA production. Neuro-2a cells were treated with a PLD inhibitor (5-fluoro-2indolyl deschlorohalopemide (FIPI)). Although AcGFP-α-Syn-N was strongly localized in the peripheral regions of growth cones in the absence of FIPI (plasma membrane/cytosol intensity ratio: 1.66 ± 0.06 (mean ± SEM)), the inhibitor (750 nM) almost completely inhibited the peripheral localization of AcGFP-α-Syn-N (plasma membrane/cytosol intensity ratio: 1.18 ± 0.04 (mean ± SEM)) ( Fig. 4A and B), indicating that PLD primarily contributed to PA production in the region. Moreover, the results demonstrated that the peripheral localization of AcGFP-α-Syn-N occurs in a PA production-dependent manner, strongly suggesting that α-Syn-N recognized endogenously generated PA. Because FIPI nearly completely inhibited the peripheral localization of α-Syn-N, the contribution of DGK to PA generation in the region is likely small.

Discussion
In the present study, we first established an inactive PA sensor (α-Syn-N-KQ) as a negative control by changing all (eleven) lysine residues, which are predicted to be important for PA binding [28], to glutamine residues (Fig. 1A). α-Syn-N-KQ indeed lost PA-binding activity in vitro ( Fig. 1B and C) and in cells (Fig. 1D and E). Because the predicted secondary structures of α-Syn-N and α-Syn-N-KQ were almost the same (Fig. 1A), α-Syn-N-KQ can be a useful negative control of α-Syn-N. We next determined that α-Syn-N, but not α-Syn-N-KQ, detected PA in macrophagic phagosomes (Fig. 2), in which PA is well known to be enriched [1,17]. Therefore, α-Syn-N is a reliable sensor for endogenous PA in cells.
Axon growth is driven by the forward movement of a growth cone, which consists of a central domain rich in microtubules and a peripheral domain enriched in actin filaments [29]. The distribution of PA in the neuronal growth cone remains unclear to date. In the present study, we found for the first time that α-Syn-N, but not α-Syn-N-KQ, detected PA in the F-actin-rich peripheral regions (close to the plasma membrane) in the neurite growth cones (Fig. 3).
FIPI almost completely inhibited the localization of α-Syn-N at the peripheral regions in neurite growth cones (Fig. 4), indicating that PLD mainly contributes to PA production at the growth cone. Remodeling the cytoskeleton is important for growth cone functions and has been suggested to be regulated by PA. For example, Ammar et al. demonstrated that PLD correlates with F-actin formation and neurite outgrowth [21]. Our findings further demonstrated the functional linkages among PA, PLD, cytoskeleton (F-actin) remodeling and neurite outgrowth/growth cone formation. Although 16:0/16:0-PA was highly generated in Neuro-2a neuroblastoma cells during neuronal differentiation, FIPI did not substantially reduce the PA amount [24]. Therefore, PLD probably generated a relatively small amount of PA in restricted regions (growth cones).
Even in the presence of FIPI, some α-Syn-N in the peripheral regions of the growth cone remained (plasma membrane/cytosol intensity ratio: 1.15 (not 1.00)) (Fig. 4). Thus, other PA-producing enzymes may produce PA in the peripheral regions. Previous studies reported that DGKζ regulated neurite outgrowth and co-localized with F-actin [30] and that the isozyme controlled the maintenance of the actin-rich spine [31]. We reported strong expression of DGKζ in Neuro-2a cells [24]. Quantitative image analysis of AcGFP alone (n = 28), AcGFP-α-Syn-N-KQ (n = 27) and AcGFP-α-Syn-N (n = 30) accumulation at the plasma membrane. Protein localization was quantified by ImageJ software. Each dot shows the plasma membrane:cytosol intensity ratio. Bars, mean ± SEM. ***P < 0.001, Kruskal-Wallis followed by Dunn's multiple comparison tests.
Moreover, overexpressed DGKζ was partly co-localized with α-Syn-N in the growth cone (Suppl Fig. 1). However, because long-term treatment with a DGKζ-specific siRNA also attenuated neurite formation [24] and a DGKζ-specific inhibitor is not available, it was difficult to analyze the contribution of DGKζ. In addition to DGKζ, DGKδ is abundant in the cells [24]. However, DGKδ did not produce considerable PA during the neuronal differentiation [24]. DGKβ also plays an important role in neurite outgrowth and spinogenesis through F-actin cytoskeleton remodeling [23,25]. However, because the expression level of DGKβ was low in Neuro-2a cells [24], DGKβ would not contribute substantially to PA generation in the cells. Therefore, among the DGK isozymes, DGKζ may generate PA at the growth cone during neuronal differentiation in Neuro-2a cells. However, further studies are needed to address this question.

Declaration of competing interest
The authors declare no conflicts of interest associated with the contents of this article.