Cyclic phosphatidic acid is produced by GDE7 in the ER lumen as a lysophospholipid mediator

doi:10.21203/rs.3.rs-2152597/v1

Download PDF

Article

Cyclic phosphatidic acid is produced by GDE7 in the ER lumen as a lysophospholipid mediator

https://doi.org/10.21203/rs.3.rs-2152597/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 16 May, 2023

Read the published version in Communications Biology →

Version 1

posted

You are reading this latest preprint version

Cyclic phosphatidic acid (cPA) is a lipid mediator, which regulates adipogenesis and glucose homeostasis by suppressing nuclear peroxisome proliferator-activated receptor γ (PPARγ). Glycerophosphodiesterase 7 (GDE7) is a Ca²⁺-dependent lysophospholipase D that localizes in the endoplasmic reticulum. Although mouse GDE7 catalyzes cPA production in a cell-free system, it is unknown whether GDE7 generates cPA in living cells. Here, we demonstrate that human GDE7 possesses cPA-producing activity in living cells as well as in a cell-free system. Furthermore, the active site of human GDE7 is directed towards the luminal side of the endoplasmic reticulum. Mutagenesis revealed that amino acid residues F227 and Y238 are important for catalytic activity. GDE7 deficiency derepresses the PPARγ pathway in human mammary MCF-7 cells, suggesting that cPA functions as an intracellular lipid mediator. These findings lead to a better understanding of the biological role of GDE7 and its product, cPA.

Biological sciences/Biochemistry/Enzymes

Biological sciences/Biochemistry/Lipids/Phospholipids

Biological sciences/Cell biology/Cell signalling/Lipid signalling

Lysophosphatidic acid (LPA) and cyclic phosphatidic acid (cPA) are lipid mediators in eukaryotic tissues and plasma^1,2. LPA regulates development, physiological functions, and pathological processes³ through at least eight types of G protein-coupled receptors⁴. LPA is mainly extracellularly produced from lysophosphatidylcholine (LPC) and other lysophospholipids by autotaxin (ATX) and is degraded by a class of lipid phosphatases⁴. Meanwhile, cPA and its synthetic analogs, carba-cPAs, inhibit cancer invasion and metastasis through ATX inhibition⁵, promote hyaluronic acid biosynthesis via LPA receptor activation⁶, and modulate adipogenesis and glucose homeostasis through the suppression of nuclear peroxisome proliferator-activated receptor γ (PPARγ)⁷. The in vivo biosynthetic mechanisms of cPA are not fully elucidated compared with those of LPA.

cPA is synthesized in mammalian cells by ATX⁸ and phospholipase D2 (PLD2)⁷. ATX is a lyso-PLD-type exoenzyme in the plasma, which hydrolyzes various lysophospholipids, including LPC, lysophosphatidylethanolamine, and lysophosphatidylserine to produce LPA⁸. ATX also possesses transphosphatidylation activity toward LPC to form cPA, although this activity is much lower than its hydrolysis activity to generate LPA⁹. Meanwhile, PLD2 mainly hydrolyzes phosphatidylcholine to phosphatidic acid in the cell membrane¹⁰ and also produces LPA and cPA from LPC⁷.

Recent evidence indicates that some members of the glycerophosphodiesterase (GDE) family have phospholipid-metabolizing activities. GDE4 and GDE7 are membrane-bound lyso-PLD-type enzymes which catalyze LPA generation using the same reaction as ATX^11–14. Recombinant mouse GDE7 (mGDE7), (but not mouse GDE4, mGDE4), produces cPA in a cell-free system¹⁴. GDE4 and GDE7 localize to the endoplasmic reticulum (ER) and require Mg²⁺ and Ca²⁺ for their enzymatic activity, respectively^11,13. However, it is unclear whether GDE7 produces cPA in living cells, and how and where GDE7 is activated by Ca²⁺. It is also unknown whether intracellularly generated cPA works as a lipid mediator. This study examined the cPA-producing activity of human GDE7 (hGDE7) in living cells as well as in a cell-free system. The topology of hGDE7 on the ER membrane was determined, and the effect of GDE7 deficiency on the PPARγ pathway was analyzed in GDE7-expressing cells. The results suggest that GDE7 is involved in the production of cPA which functions as an intracellular lipid mediator.

GDE7 produces cPA in a cell-free system

We examined whether hGDE7 produces cPA in a cell-free system. For this purpose, FLAG-tagged hGDE7 as well as mGDE7 were stably overexpressed in COS-7 cells and the membrane fractions were prepared. Activity of recombinant GDE7 enzymes was confirmed using a selective assay system with a fluorescent substrate, FS-3¹⁵. The FS-3-degrading activities of hGDE7- and mGDE7-expressing cells (0.65 ± 0.03 and 1.00 ± 0.06 µmol h^− 1 per mg protein, respectively) significantly increased compared with control cells (0.05 ± 0.02 µmol h^− 1 per mg protein) (Fig. 1a). Enzyme overexpression was confirmed by immunoblot analyses with anti-FLAG and anti-hGDE7 antibodies (Fig. 1b). Anti-FLAG antibody gave strong bands for mGDE7-expressing cells, while anti-hGDE7 antibody only presented faint bands for the same cells; this may reflect the fact that this antibody was raised against hGDE7. Recombinant hGDE7-containing membrane fractions generated cPA in addition to LPA following the addition of [¹⁴C]LPC substrate (Fig. 1c). The cPA-producing activity of hGDE7 (1.81 ± 0.58 µmol h^− 1 per mg protein) was 3.7-fold higher than the LPA-producing activity (0.50 ± 0.17 µmol h^− 1 per mg protein) (Fig. 1d). Recombinant mGDE7 also preferred to generate cPA rather than LPA, consistent with previous reports¹⁴. The cPA- and LPA-producing activities were 5.34 ± 1.98 and 0.83 ± 0.21 µmol h^− 1 per mg protein, respectively. In order to analyze endogenous GDE7, we also prepared the membrane fractions of wild-type human breast cancer MCF-7 cells (WT) endogenously expressing GDE7, as well as GDE7-knockout MCF-7 cells (KO)¹⁵. FS-3-degrading activity for KO was reduced to 37% of that for WT, and ectopic expression of mGDE7 in KO restored the activity to 242% (Fig. 1e). The loss of GDE7 in KO was confirmed by western blotting using an anti-GDE7 antibody, while mGDE7 overexpression was confirmed with an anti-FLAG antibody (Fig. 1f). cPA and LPA production rates of the KO membrane fractions were 16.7% and 48.2% of those of the WT, respectively. mGDE7 overexpression restored these rates to 292% and 434%, respectively (Fig. 1g, h). These results indicate that recombinant and endogenous proteins of hGDE7 produce not only LPA but also cPA in the cell-free system.

GDE7 produces cPA in living cells

We next examined whether hGDE7 produces cPA along with LPA in cells. Mass spectrometry revealed that hGDE7 overexpression in COS-7 cells produced a 2.2-fold increase in intracellular LPA levels compared to control cells (Fig. 2a), which correlated with our previous report¹³. The most abundant LPA molecular species in hGDE7-overexpressing cells was C18:1, followed by C16:0, C16:1, C18:0, and C18:2 (Fig. 2a). Polyunsaturated fatty acid-containing LPA was below the detection limit except for C18:2, even with hGDE7 overexpression. We could not detect any cPA molecular species in control COS-7 cells, while C18:0, C18:1, and C16:0 species were found in hGDE7-overexpressing cells with a total cPA amount of 7.5 ± 2.1 pmol per mg protein (Fig. 2b). The intracellular levels of LPA and cPA in WT were up to 91.2 ± 23.5 and 26.9 ± 6.9 pmol per mg protein, respectively. Their levels significantly decreased to 12.3% and 26.3%, respectively, in KO (Fig. 2c, d). Furthermore, mGDE7 overexpression in KO resulted in 2.3-fold and 1.4-fold increases in the LPA and cPA levels compared with the WT, respectively. Similar results were obtained with each molecular species of LPA and cPA. These results suggest that recombinant and endogenous proteins of hGDE7 produce cPA along with LPA in living cells.

The active site of hGDE7 is directed toward the ER lumen

We previously reported that overexpressed GDE7 localizes to the ER and its activity is stimulated by Ca^2 + 13. Analysis of hGDE7 topology on the ER membrane using three prediction programs (HMMTOP¹⁶, TMHMM¹⁷, and SPLIT¹⁸) suggested that it has up to three predicted transmembrane (TM) domains (around amino acid residues 3–23 (TM1), 200–219 (TM2), and 240–259 (TM3)) (Fig. 3a). A 3D structure model of hGDE7 was created using AlphaFold¹⁹ (Fig. 3b & Supplementary Fig. 1). The substrate and Ca²⁺ binding sites were then predicted using Firestar²⁰, a ligand-binding site prediction software (Fig. 3c). E71, D73, and E150 of hGDE7 were predicted as Ca²⁺ binding sites, while H44, R45, E71, D73, H86, E150, K152, E180, and W277 were predicted as binding sites for glycerol 3-phosphate, a component of LPC, cPA, and LPA (Fig. 3c). Most of these putative binding sites are between the predicted TM1 and TM2. The topology of hGDE7 on the ER membrane was evaluated using the proteinase K (proK) protection assay of particulate fractions from COS-7 cells overexpressing N- and C-terminally FLAG-tagged hGDE7 (GDE7-N and GDE7-C, respectively) and analyzed by western blotting using anti-FLAG and anti-GDE7 antibodies. The proK treatment without any detergent should degrade the cytoplasmic region of the protein but not the luminal or TM region of the protein. Meanwhile, proK treatment with the zwitterionic detergent CHAPS solubilizes the ER membrane and degrades all protein regions. This assay was validated based on the observed degradation of ER lumen localized protein disulfide isomerase (PDI) (around 57 kDa) following proK and CHAPS treatment, but not by the proK treatment alone (Fig. 4a). ProK treatment of the GDE7-N particulate fraction without detergent produced an immunoreactive band with a slightly lower molecular mass using anti-FLAG antibody (Fig. 4a). This suggests that the N-terminus of hGDE7 is located in the ER lumen. In contrast, the protein band of GDE7-C was lost by a similar proK treatment, suggesting that the C-terminus of hGDE7 localizes to the cytoplasmic side (Fig. 4a). Immunoblot analyses of proK-treated GDE7-N and GDE7-C particulate fractions with anti-GDE7 antibody indicated a protein band around 32 kDa, suggesting that the immunogenic sequence (aa 107–177) localizes to the ER luminal side (Fig. 4a). Furthermore, similar results were obtained with endogenous GDE7 in MCF-7 cells using anti-GDE7 antibody (Fig. 4b). These results indicate that the N-terminus of hGDE7 localizes to the ER luminal side, the C-terminus is on the cytoplasmic side, and the active site of hGDE7 is directed toward the ER lumen (Fig. 4c).

GDE7 topology was also examined by immunocytochemistry of COS-7 cells overexpressing GDE7-N and GDE7-C. A low digitonin concentration selectively permeabilizes the plasma membrane, whereas Triton X-100 permeabilizes all cellular membranes²¹. Thus, cytoplasmic antigens should be stained when permeabilized with either digitonin or Triton X-100, while antigens in the ER lumen are stained only when permeabilized with Triton X-100. We confirmed that PDI proteins in the ER lumen were immunodetected when permeabilized with Triton X-100 and that cytoplasmic tubulin β3 proteins were detected following treatment with either digitonin or Triton X-100 (Fig. 4d). GDE7-N was detected using an anti-FLAG antibody when permeabilized with Triton X-100, while GDE7-C was detected when permeabilized with either digitonin or Triton X-100 (Fig. 4d). These results support the proK protection assays indicating that the N- and C-terminus of hGDE7 are located on the ER lumen and cytoplasm, respectively (Fig. 4e). Together, these observations indicate that the hGDE7 active site is directed toward the ER lumen (Fig. 4f).

Mutagenesis reveals the importance of F227 and Y238 for hGDE7 catalytic activity

The Smith-Waterman algorithm²² revealed that the amino acid sequences of hGDE4 and hGDE7 are 42.1% identical and 64.1% similar (Fig. 5a). Since mGDE4 produces LPA but not cPA in cell-free assays¹⁴, we compared the amino acid sequences of GDE4 and GDE7 from humans and mice to determine which amino acid residues in hGDE7 are important for its ability to produce cPA. We considered that hydrophobic amino acid residues that prevent water molecules from approaching the active pocket are important for cPA production by transphosphatidylation. Therefore, F227E and Y238K substitutions in hGDE7 were selected based on the following three criteria: (i) hydrophobic amino acid residues in hGDE7 are replaced with hydrophilic amino acid residues in hGDE4, (ii) amino acid residues are located on the surface of the active pocket of the structural model, and (iii) amino acid residues are conserved between hGDE7 and mGDE7. The hGDE7 mutants F227E, Y238K, and F227E/Y238K were then expressed in COS-7 cells. The expression level of Y238K was higher compared to wild-type hGDE7, while F227E and F227E/Y238K had lower expression than the wild-type according to western blotting with anti-FLAG and anti-GDE7 antibodies using the same amount (10 µg) of the membrane fractions (Supplementary Fig. 2). We used the amounts of membrane fractions giving similar band intensities of hGDE7 by western blotting in the following enzyme assays (Fig. 5b). Membrane fractions from control cells were added to retain the same amounts of endogenous COS-7 proteins (10 µg). The FS-3-degrading activity of F227E (0.071 ± 0.005 µmol h^− 1 per mg protein) was significantly lower than that of the wild-type (0.136 ± 0.002 µmol h^− 1 per mg protein). The activity of Y238K (0.042 ± 0.004 µmol h^− 1 per mg protein) was almost the same as the control (0.035 ± 0.008 µmol h^− 1 per mg protein), while that of F227E/Y238K (0.028 ± 0.003 µmol h^− 1 per mg protein) was below the basal level (Fig. 5c). The LPA- and cPA-producing activities from [¹⁴C]LPC using 0.5 µg of protein sample (Fig. 5d, e) showed a similar tendency to those of FS-3-degrading activity, although the decrease in cPA-producing activity of F227E was not statistically significant. These mutation assays did not allow us to determine the critical amino acid residues of hGDE7 for cPA production; however, both F227 and Y238 are critical amino acid residues for the catalytic activity of hGDE7.

cPA produced by GDE7 suppresses PPARγ target genes

Finally, we examined whether intracellularly produced cPA by GDE7 acts as a lipid mediator. cPA acts as an antagonist of PPARγ⁷. Therefore, we examined the mRNA expression levels of PPARγ downstream factors: CD36, CYP27A1, and CSF1. All of these levels were derepressed in KO compared to WT, and their levels were significantly suppressed by the expression of mGDE7 in KO (Fig. 6). The expression levels of PPARG were similar between three types of MCF-7 cells. These results suggest that cPA produced by GDE7 in the ER lumen is an intracellular lipid mediator.

This study demonstrated that hGDE7 has intracellular cPA-producing properties and the active site of hGDE7 is directed toward the lumen of the ER. Moreover, loss of GDE7 derepressed the PPARγ pathway. These results suggest that intracellular cPA produced by GDE7 localized in the ER functions as a lipid mediator (Fig. 7).

ATX and PLD2 are involved in the production of cPA and LPA from LPC^7,9. Moreover, mouse GDE7 catalyzes these reactions in a cell-free system¹⁴. Consistent with these findings, we confirmed that hGDE7 produces cPA and LPA in cells as well as in a cell-free system. ATX preferentially generates LPA rather than cPA under near-physiological conditions in a cell-free system, with a 1:4 ratio of cPA:LPA production⁹. In contrast, we found that the ratio of cPA:LPA production by hGDE7 was approximately 4:1 to 5:1 in a cell-free system (Fig. 1). However, our assay system contains 5% ethanol, and this ratio should be examined under conditions similar to biological membranes. In contrast, the ratio of cPA:LPA production by hGDE7 was 1:3 to 1:9 in cells (Fig. 2). This difference may result from the involvement of other LPA-producing enzymes such as GDE4 and acylglycerol kinase²³ since MCF-7 cells have poor ATX expression²⁴. It is also possible that cPA was hydrolyzed to LPA non-enzymatically in the presence of water²⁵. In humans, GDE7 is abundant in the kidney, prostate, ovary, and placenta¹³, while PLD2 is widely expressed in various tissues²⁶. Thus, ATX, GDE7, and PLD2 produce cPA but differ in their sites of action, substrate specificity, and tissue distribution: ATX is mainly involved in blood cPA production, while GDE7 and PLD2 produce cPA intracellularly, and each enzyme may have a unique role. The LPC in plasma is involved in cell migration, cytokine production, oxidative stress, and apoptosis²⁷, while intracellular LPC is involved in ER stress²⁸. Therefore, GDE7 may be involved in LPC removal for ER homeostasis, in addition to the production of LPA and cPA as lipid mediators.

GDE family members GDE1 and GDE4 have two TM domains and their active sites are on the luminal side^29,30. Similarly, the present study revealed that the active site of GDE7 is on the luminal side of the ER membrane based on topological predictions, the proK protection assay, and immunocytochemistry (Figs. 3 and 4). GDE7 activity is Ca²⁺-dependent¹³; therefore, GDE7 should be activated by the abundant levels of Ca²⁺ in the ER lumen to constitutively produce LPA and cPA. In contrast, PLD2 localizes to the plasma membrane and generates cPA in a stimulus-coupled manner⁷. Thus, GDE7 and PLD2 may have distinct roles in terms of localization and stimulus responsiveness. The N-terminally FLAG-tagged GDE7 used in this study contains a FLAG region followed by a 5×Gly linker which is only two or three amino acid residues from the presumed TM1. AlphaFold predicts that adding a FLAG region to the N- or C-terminus of hGDE7 has no significant effect on the structural conformation. However, the negative charge of FLAG might alter translocon-mediated protein translocation and TM domain integration. In the present study, the intracellular localization of GDE7 was analyzed using cells overexpressing FLAG-tagged hGDE7. Thus, the localization of endogenous GDE7 remains unknown due to the unavailability of reliable antibodies applicable for immunocytochemistry. Further studies with more accurate structural information are required in the future.

Although the active sites of GDE4 and GDE7 are highly conserved, their cPA-producing activities are quite different¹⁴. We constructed GDE7 mutants by replacing the hydrophobic amino acid residues in the active site (F227 and Y238) with hGDE4-like hydrophilic amino acid residues to investigate whether they are important for cPA-producing activity of hGDE7 (Fig. 5). LPA-producing activity was preferentially decreased in F227E, while both LPA- and cPA-producing activities were reduced in Y238K. Furthermore, both activities were completely lost in F227E/Y238K. Previous studies of Streptomyces PLD depicted that amino acid residues forming the active site entrance are involved in substrate recognition, while those forming the active site stabilize the conformation³¹. It was predicted that F227 forms the active site and Y238 forms the entrance of hGDE7 (Supplementary Fig. 1b). Thus, F227 and Y238 might be important for stability and substrate recognition, respectively. Although F227 and Y238 are important for hGDE7 activity, further structural biological studies are needed to clarify how GDE7 produces cPA.

Currently, no specific targets for cPA are reported. Extracellularly produced cPA acts on LPA receptors such as LPA1–3³², LPA5³³, LPA6³⁴, and P2Y5³⁵. However, it is unclear whether cPA acts in a similar or opposite manner to LPA. Meanwhile, cPA intracellularly produced by PLD2 acts as an antagonist of PPARγ⁷ and inhibits intimal thickening and proliferation of colon cancer cells by inducing apoptosis³⁶. The present study similarly showed that the loss of GDE7 derepressed the expression of PPARγ target genes (Fig. 6), indicating that intracellularly produced cPA by GDE7 inhibits the PPARγ pathway. These results suggest a role of cPA as an intracellular lipid mediator, and it is necessary to clarify whether and how intracellular cPA produced by GDE7 contributes to the pathogenesis of cancer and cardiovascular diseases via the PPARγ pathway. Furthermore, the possible involvement of another intracellular cPA target, adenine nucleotide translocase, which interacts with 2-carba-cPA³⁷ should be considered.

GDE7 is involved in cancer stem cell survival³⁸, hepatic lipidosis³⁹, and noise-induced hearing loss⁴⁰. Furthermore, GDPD3 encoding hGDE7 is located at chromosome 16p11.2, and copy number variation in this region is associated with increased risk of neuropsychiatric diseases⁴¹ and obesity-related syndromes⁴². However, it is unclear whether and how the lipid mediators intracellularly produced by GDE7 contribute to these pathologies. Protein carriers of lipid mediators are essential for the regulation of physiological functions and pathological processes⁸. LPA and cPA bind to albumin in the blood^43–45 and act as lipid mediators. Meanwhile, intracellularly produced LPA functions as a PPARγ agonist⁴⁶, and fatty acid binding protein 3 is an LPA carrier protein in PPARγ activation⁴⁷. Further studies are required to clarify the intracellular cPA-specific binding proteins and the transport mechanisms from the ER lumen to demonstrate that intracellular cPA also plays an important role in vivo. Future studies should focus on not only LPA but also cPA as the intracellular products of GDE7 when analyzing its association with various pathological conditions.

In conclusion, we demonstrated that GDE7 produces cPA in the ER lumen, which could function as an intracellular lipid mediator. These findings lead to a better understanding of the biological role of GDE7 as well as its products, LPA and cPA.

Materials

FS-3 was purchased from Echelon Biosciences (Salt Lake City, UT). Anti-FLAG M2 (F1804) and anti-PDI (P7372) antibodies were purchased from Sigma-Aldrich (St. Louis, MO), anti-GDE7 antibody (HPA041148) was from Atlas Antibodies (Stockholm, Sweden), anti-GAPDH antibody (M171-3) was from MBL (Tokyo, Japan), and anti-tubulin β-3 antibody (921001) was from BioLegend (San Diego, CA). [¹⁴C]LPC (1-[¹⁴C]oleoyl-sn-glycerophosphocholine) was purchased from American Radiolabeled Chemicals (ARC 3094; St. Louis, MO). LPC (1-oleoyl-sn-glycerophosphocholine) and cPA (1-oleoyl-sn-glycero-2,3-cyclic-phosphate) were from Avanti Polar Lipids (Alabaster, AL). Nonidet P-40 and proK were procured from Nacalai Tesque (Kyoto, Japan). CHAPS was purchased from Dojindo (Kumamoto, Japan).

Cell line

Transformed African green monkey kidney fibroblast COS-7 cells (RCB0539, Riken BioResource Research Center, Tsukuba, Japan) were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (FUJIFILM Wako, Osaka, Japan) supplemented with 10% (v/v) fetal bovine serum (FBS) (Biofluids, Fleming Island, FL) and nonessential amino acids (FUJIFILM Wako). Human breast cancer MCF-7 cells (Health Science Research Resources Bank, Osaka, Japan) were maintained in DMEM supplemented with 15% (v/v) FBS. A knockout cell line for human GDE7 gene (GDPD3) was previously established¹⁵. These cell lines were cultured at 37°C in a humidified atmosphere of 5% CO₂ and 95% air.

Plasmid construction and lentiviral preparation

cDNA of human or mouse GDE7 with a C-terminal FLAG tag was previously constructed ¹³. A cDNA sequence encoding a FLAG peptide was introduced by PCR using PrimeSTAR HS DNA Polymerase (TaKaRa Bio, Kusatsu, Japan) to construct cDNA for N-terminally FLAG-tagged GDE7. This was generated using a forward primer containing a XbaI site, Kozak sequence, FLAG, and 5×Gly linker, and a reverse primer with a BamHI site. Site-directed mutagenesis was performed by overlap extension PCR. Each DNA fragment was amplified using primer sets shown in Supplementary Table 1 and subcloned into the third generation lentiviral backbone vector containing a hygromycin resistant gene as previously described⁴⁸.

COS-7 or MCF-7 cells were transduced with each obtained lentivirus using 8 µg mL^− 1 hexadimethrine bromide (Sigma-Aldrich) and then selected with hygromycin B (Nacalai Tesque).

Immunoblot analysis

Immunoblot analysis was performed as previously described⁴⁹.

Enzyme assay

Fluorescence-based enzyme assays were based on a previous method¹⁵. Briefly, membrane fractions (5 µg of protein) were incubated with 5 µM FS-3 for 3 h at 37°C in 60 µL of 50 mM Tris–HCl buffer (pH 7.4) in the presence of 2 mM CaCl₂ and 0.1% (w/v) Nonidet P-40.

Radioactivity-based enzyme assays were performed as previously described¹⁵ with modifications. The membrane fractions were incubated at 37°C for 30 min with 25 µM of [¹⁴C]LPC (20,000 cpm, dissolved in 5 µL of ethanol) in 100 µL of 50 mM Tris–HCl (pH 7.4) containing 2 mM of CaCl₂ and 100 µM of phosphatase inhibitor sodium orthovanadate to protect the product from degradation¹¹. The reactions were terminated by adding 0.32 mL chloroform/methanol/1 M citric acid (8:4:1, by vol.) with 5 mM butylated hydroxyanisole as an antioxidant. After centrifugation, the organic layer was transferred to a new tube, and the aqueous layer was further extracted with 250 µL of chloroform/methanol (17:3, v/v). The combined organic layer was dried under a gentle stream of nitrogen. The obtained lipids were dissolved in chloroform/methanol (2:1) and spotted on a silica gel-coated aluminum thin layer chromatography (TLC) sheet (20-cm height; Merck, Darmstadt, Germany), and developed at 4°C for 100 min using chloroform/methanol/28% (w/w) ammonium hydroxide (60:35:8, by vol.)¹⁴. The radioactivity of the substrate and product on the plate was quantified using an image analyzer (Typhoon 9400; Cytiva-GE Healthcare, Marlborough, MA), and the enzyme activity was then calculated.

Lipid analyses by liquid chromatography-tandem mass spectrometry (LC-MS/MS)

The cells were cultured to 100% confluence in a 100-mm dish, harvested by scraping, and stored at − 80°C until use. Lipid extraction was performed according to previous studies^13,50, and one hundred pmol of C17:0 LPA was mixed as an internal standard at this step. The amounts of LPA and cPA species were determined by the ratios of the peak areas in the following LC-MS/MS experiments.

LC-MS/MS was performed using a quadrupole-linear ion trap hybrid mass spectrometry system, 4000 Q TRAP (Applied Biosystems/MDS Sciex, Concord, Ontario, Canada) with an 1100 liquid chromatography system (Agilent Technologies, Wilmington, DE, USA) combined with an HTS-PAL autosampler (CTC Analytics AG, Zwingen, Switzerland) as previously stated¹³ with modifications. LPA and cPA species were separated using a Tosoh TSK-ODS-100Z column (150 mm × 2 mm, 5 µm particle size) with methanol/water (95:5, v/v) containing 0.05 M ammonium formate at a flow rate of 0.20 mL min^− 1 at 42°C. Five µL aliquots of the test solution were routinely applied using the autosampler. The negative ion mode of operation with multiple reaction monitoring was used, and deprotonated molecular ion and deprotonated cyclic glycerophosphate at m/z 153 were selected for Q1 and Q3, respectively. Values were calculated from the ratios of the ion peak areas of the analyzing target molecules to that of the internal standard, C17:0 LPA. The ionization efficiency of cPA species was found to be higher than that of LPA; therefore, the values for cPA species were multiplied by a correction factor of 0.239. This factor was determined by the slope of the calibration curve obtained from the plots of peak area ratios versus molar ratios between authentic molecules, C17:0 LPA and C18:1 cPA. Log transformed data were used for statistical analyses.

Structural Prediction

TM domains of hGDE7 were predicted by HMMTOP¹⁶, TMHMM-2.0¹⁷, and SPLIT4.0 ¹⁸. Ligand and calcium ion binding sites were predicted using Firestar²⁰. The spatial structure of GDE7 was modeled and visualized with AlphaFold¹⁹ and PyMOL 2.3 software.

Topology

ProK protection assays were performed as previously described³⁰ with modifications. Briefly, each cell suspension was prepared in buffer P (250 mM sucrose and 1 mM ethylenediaminetetraacetic acid (EDTA) in 50 mM Tris–HCl, pH 7.4) and homogenized by passing through a 27G needle 20 times on ice. Nuclei and unbroken cells were removed by centrifugation at 600 g for 5 min at 4°C, and the supernatants were further centrifuged using an Optima MAX-TL (Beckman Coulter, Brea, CA) at 105,000 g for 1 h at 4°C. The resulting pellets were resuspended in buffer P and used as the particulate fractions. The particulate fraction (30 µg of protein) was incubated with 150 µg mL^− 1 proK in the presence or absence of 1% (w/v) CHAPS for 60 min at 50°C. The reaction was terminated by the addition of 5 mM phenylmethylsulphonyl fluoride (PMSF) (Sigma-Aldrich). Samples were immediately added to sodium dodecyl sulfate (SDS) sample buffer containing 50 mM Tris–HCl (pH 6.8), 2% (w/v) SDS, 10% (v/v) glycerol, 0.01% (w/v) bromophenol blue, and 100 mM dithiothreitol (DTT), heated at 95°C for 10 min, and subjected to SDS-PAGE followed by immunoblotting.

Immunocytochemistry involved fixing COS-7 cells in a poly-l-lysine-coated 8-well Lab-Tek II chamber slide (Thermo Fisher Scientific, Waltham, MA) with 4% (w/v) paraformaldehyde in phosphate buffered saline (PBS) for 10 min at 25°C. Cells were rinsed with PBS, and then permeabilized with 0.1% (v/v) Triton X-100 (FUJIFILM Wako) in PBS or 0.001% (w/v) digitonin (Sigma-Aldrich) in PBS containing 10 mM HEPES (pH 7.5), 300 mM sucrose, 100 mM KCl, 2.5 mM MgCl₂, and 0.5% (w/v) BSA on ice for 15 min. After blocking with 1% (w/v) BSA/PBS for 60 min at 25°C, the samples were incubated with specific antibodies diluted in blocking buffer (FLAG, 1:1,000; PDI, 1:250; tubulin β-3, 1:1,000) for 16 h at 4°C, followed by incubation with Alexa Fluor-conjugated secondary antibodies (diluted 1:1,000 with the blocking buffer) for 1 h at 25°C and then with 10 µg mL^− 1 Hoechst 33342 solution for 5 min to stain the nuclei. Fluorescence was then observed using a laser-scanning confocal microscopy (LSM700; Carl Zeiss, Jena, Germany).

RNA isolation and quantitative PCR (qPCR)

MCF-7 cells were cultured in serum-free media containing 0.2% (w/v) fatty acid-free BSA (FUJIFILM Wako) for 20 h. RNA isolation, cDNA synthesis, and qPCR were then performed as previously described⁵¹. The primers used are listed in Supplementary Table 1.

Statistics and Reproducibility

All data are expressed as the mean ± SD. Statistical analyses were performed using Prism ver. 8.42 (GraphPad Software; San Diego, CA, USA). The sample sizes and number of replicates for each experiment are indicated in the Figure legends. The unpaired two-tailed t-test was used to compare two groups, while ANOVA and Dunnett’s test were used to compare three or more groups. P-values < 0.05 were considered statistically significant.

Data availability

Uncropped immunoblots are shown in Supplementary Fig. 3–5. The original data of the graphs and charts are shown in Supplementary Data 1. All other data provided in the article and Supplementary files are available from the corresponding author upon reasonable request.

Acknowledgments

The authors would like to thank Rie Sugimoto for technical assistance and also acknowledge technical support from the Units of Molecular and Cellular Biology Research, Medical Bioresource Research, and Radioisotope Research, Central Research Institute, Kawasaki Medical School. The authors thank Editage (www.editage.com) for English language editing. This work was supported by the Japan Society for the Promotion of Science KAKENHI (grant numbers JP20K19732 to K.K. and JP20K11571 to K.T.); the Ichiro Kanehara Foundation (grant number 2020-23 to K.K.); Teraoka Scholarship Foundation (to K.K.); Teijin Pharma Limited (grant numbers TJNS20200413014, TJNS20210412018, and TJNS20220411014 to K.K. and Y.O.); Kyowa Kirin Co., Ltd. (grant number KKCS20210402007 and KKCS20220401021 to K.K. and Y.O.); Bayer Academic Support (grant number BASJ20210402010 and BASJ20220401001 to K.K); Teijin Nakashima Medical Co., Ltd. (to K.K. and Y.O.); Otsuka Pharmaceutical Co., Ltd. (grant number AS2022A000441768 to K.K. and Y.O.); Asahi Kasei Pharma Corporation (grant number APJS20220406011 to K.K. and Y.O.) and Research Project Grants from Kawasaki Medical School (grant numbers R01S-002, R02S-002, and R04Y-001 to K.K.). The funders had no role in study design, data collection, decision to publish, or preparation of the manuscript.

Author contributions

Conceptualization, K.K., N.U., and K.T.; Formal analysis, K.K.; Investigation, K.K., H.A., and R.K.; Data curation, K.K., and H.A.; Writing - Original Draft, K.K., Y.O., and K.T.; Writing - Review & Editing, all authors; Visualization, K.K.; Project administration; K.K. and K.T.; Funding acquisition, K.K., Y.O., and K.T. All authors approved the manuscript for publication.

Ethics declarations

Conflict of interest

K.K. and Y.O. received a grant from Teijin Pharma Limited, Bayer Yakuhin, Ltd., Kyowa Kirin Co., Ltd., Teijin Nakashima Medical Co., Ltd., Otsuka Pharmaceutical Co., Ltd., and Asahi Kasei Pharma Corporation. H.I. is an employee of Maruho Co., Ltd. All other authors declare no conflicts of interest.

Tokumura, A. Metabolic pathways and physiological and pathological significances of lysolipid phosphate mediators. Journal of Cellular Biochemistry 92, 869–881, doi:10.1002/jcb.20147 (2004).
Geraldo, L. H. M. et al. Role of lysophosphatidic acid and its receptors in health and disease: Novel therapeutic strategies. Signal Transduction and Targeted Therapy 6, 45, doi:10.1038/s41392-020-00367-5 (2021).
Fujiwara, Y. Cyclic phosphatidic acid—a unique bioactive phospholipid. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1781, 519-524, doi:10.1016/j.bbalip.2008.05.002 (2008).
Kano, K., Aoki, J. & Hla, T. Lysophospholipid mediators in health and disease. Annual Review of Pathology: Mechanisms of Disease 17, 459-483, doi:10.1146/annurev-pathol-050420-025929 (2022).
Baker, D. L. et al. Carba analogs of cyclic phosphatidic acid are selective inhibitors of autotaxin and cancer cell invasion and metastasis. Journal of Biological Chemistry 281, 22786–22793, doi:10.1074/jbc.M512486200 (2006).
Maeda-Sano, K. et al. Cyclic phosphatidic acid and lysophosphatidic acid induce hyaluronic acid synthesis via CREB transcription factor regulation in human skin fibroblasts. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1841, 1256-1263, doi:10.1016/j.bbalip.2014.05.004 (2014).
Tsukahara, T. et al. Phospholipase D2-dependent inhibition of the nuclear hormone receptor PPARγ by cyclic phosphatidic acid. Molecular Cell 39, 421–432, doi:10.1016/j.molcel.2010.07.022 (2010).
Tan, S. T., Ramesh, T., Toh, X. R. & Nguyen, L. N. Emerging roles of lysophospholipids in health and disease. Progress in Lipid Research 80, 101068, doi:10.1016/j.plipres.2020.101068 (2020).
Tsuda, S. et al. Cyclic phosphatidic acid is produced by autotaxin in blood. Journal of Biological Chemistry 281, 26081–26088, doi:10.1074/jbc.M602925200 (2006).
Bowling, F. Z., Frohman, M. A. & Airola, M. V. Structure and regulation of human phospholipase D. Advances in Biological Regulation 79, 100783, doi:10.1016/j.jbior.2020.100783 (2021).
Tsuboi, K. et al. Glycerophosphodiesterase GDE4 as a novel lysophospholipase D: a possible involvement in bioactive N-acylethanolamine biosynthesis. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1851, 537–548, doi:10.1016/j.bbalip.2015.01.002 (2015).
Ohshima, N. et al. New members of the mammalian glycerophosphodiester phosphodiesterase family: GDE4 and GDE7 produce lysophosphatidic acid by lysophospholipase D activity. Journal of Biological Chemistry 290, 4260–4271, doi:10.1074/jbc.M114.614537 (2015).
Rahman, I. A. S. et al. Calcium-dependent generation of N-acylethanolamines and lysophosphatidic acids by glycerophosphodiesterase GDE7. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids 1861, 1881–1892, doi:10.1016/j.bbalip.2016.09.008 (2016).
Tserendavga, B. et al. Characterization of recombinant murine GDE4 and GDE7, enzymes producing lysophosphatidic acid and/or cyclic phosphatidic acid. The Journal of Biochemistry 170, 713–727, doi:10.1093/jb/mvab091 (2021).
Kitakaze, K. et al. Development of a selective fluorescence-based enzyme assay for glycerophosphodiesterase family members GDE4 and GDE7. Journal of Lipid Research 62, 100141, doi:10.1016/j.jlr.2021.100141 (2021).
Tusnády, G. E. & Simon, I. The HMMTOP transmembrane topology prediction server. Bioinformatics 17, 849–850, doi:10.1093/bioinformatics/17.9.849 (2001).
Krogh, A., Larsson, B., Von Heijne, G. & Sonnhammer, E. L. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. Journal of Molecular Biology 305, 567–580, doi:10.1006/jmbi.2000.4315 (2001).
Juretić, D., Zoranić, L. & Zucić, D. Basic charge clusters and predictions of membrane protein topology. Journal of Chemical Information and Computer Sciences 42, 620–632, doi:10.1021/ci010263s (2002).
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589, doi:10.1038/s41586-021-03819-2 (2021).
López, G., Valencia, A. & Tress, M. L. firestar—prediction of functionally important residues using structural templates and alignment reliability. Nucleic Acids Research 35, W573–W577, doi:10.1093/nar/gkm297 (2007).
Yasuda, S., Nishijima, M. & Hanada, K. Localization, topology, and function of the LCB1 subunit of serine palmitoyltransferase in mammalian cells. Journal of Biological Chemistry 278, 4176–4183, doi:10.1074/jbc.M209602200 (2003).
Smith, T. F. & Waterman, M. S. Identification of common molecular subsequences. Journal of Molecular Biology 147, 195–197, doi:10.1016/0022-2836(81)90087-5 (1981).
Bektas, M. et al. A novel acylglycerol kinase that produces lysophosphatidic acid modulates cross talk with EGFR in prostate cancer cells. The Journal of Cell Biology 169, 801-811, doi:10.1083/jcb.200407123 (2005).
Samadi, N., Gaetano, C., Goping, I. & Brindley, D. Autotaxin protects MCF-7 breast cancer and MDA-MB-435 melanoma cells against Taxol-induced apoptosis. Oncogene 28, 1028–1039, doi:10.1038/onc.2008.442 (2009).
Ortuno, V. E., Pulletikurti, S., Veena, K. S. & Krishnamurthy, R. Synthesis and hydrolytic stability of cyclic phosphatidic acids: implications for synthetic-and proto-cell studies. Chemical Communications 58, 6231-6234, doi:10.1039/D2CC00292B (2022).
Steed, P. M., Clark, K. L., Boyar, W. C. & Lasala, D. J. Characterization of human PLD2 and the analysis of PLD isoform splice variants. The FASEB Journal 12, 1309-1317, doi:10.1096/fasebj.12.13.1309 (1998).
Liu, P. et al. The mechanisms of lysophosphatidylcholine in the development of diseases. Life Sciences 247, 117443, doi:10.1016/j.lfs.2020.117443 (2020).
Kakisaka, K. et al. Mechanisms of lysophosphatidylcholine-induced hepatocyte lipoapoptosis. American Journal of Physiology-Gastrointestinal and Liver Physiology 302, G77-G84, doi:10.1152/ajpgi.00301.2011 (2012).
Zheng, B., Berrie, C. P., Corda, D. & Farquhar, M. G. GDE1/MIR16 is a glycerophosphoinositol phosphodiesterase regulated by stimulation of G protein-coupled receptors. Proceedings of the National Academy of Sciences of the United States of America 100, 1745–1750, doi:10.1073/pnas.0337605100 (2003).
Aoyama, C. et al. Characterization of glycerophosphodiesterase 4-interacting molecules G_αq/11 and G_β, which mediate cellular lysophospholipase D activity. Biochemical Journal 476, 3721–3736, doi:10.1042/BCJ20190666 (2019).
Uesugi, Y., Arima, J., Iwabuchi, M. & Hatanaka, T. C‐terminal loop of Streptomyces phospholipase D has multiple functional roles. Protein Science 16, 197-207 (2007).
Fujiwara, Y. et al. Identification of residues responsible for ligand recognition and regioisomeric selectivity of lysophosphatidic acid receptors expressed in mammalian cells. Journal of Biological Chemistry 280, 35038–35050, doi:10.1074/jbc.M504351200 (2005).
Williams, J. R. et al. Unique ligand selectivity of the GPR92/LPA₅ lysophosphatidate receptor indicates role in human platelet activation. Journal of Biological Chemistry 284, 17304–17319, doi:10.1074/jbc.M109.003194 (2009).
Okuyama, K., Mizuno, K., Nittami, K., Sakaue, H. & Sato, T. Molecular mechanisms of cyclic phosphatidic acid-induced lymphangiogenic actions in vitro. Microvascular Research 139, 104273, doi:10.1016/j.mvr.2021.104273 (2022).
Konakazawa, M., Gotoh, M., Murakami-Murofushi, K., Hamano, A. & Miyamoto, Y. The effect of cyclic phosphatidic acid on the proliferation and differentiation of mouse cerebellar granule precursor cells during cerebellar development. Brain Research 1614, 28–37, doi:10.1016/j.brainres.2015.04.013 (2015).
Tsukahara, T., Hanazawa, S., Kobayashi, T., Iwamoto, Y. & Murakami-Murofushi, K. Cyclic phosphatidic acid decreases proliferation and survival of colon cancer cells by inhibiting peroxisome proliferator-activated receptor γ. Prostaglandins & Other Lipid Mediators 93, 126–133, doi:10.1016/j.prostaglandins.2010.09.002 (2010).
Tsukahara, T. et al. Adenine nucleotide translocase 2, a putative target protein for 2-carba cyclic phosphatidic acid in microglial cells. Cellular Signalling 82, 109951, doi:10.1016/j.cellsig.2021.109951 (2021).
Naka, K. et al. The lysophospholipase D enzyme Gdpd3 is required to maintain chronic myelogenous leukaemia stem cells. Nature Communications 11, 4681, doi:10.1038/s41467-020-18491-9 (2020).
Key, C.-C. C. et al. Human GDPD3 overexpression promotes liver steatosis by increasing lysophosphatidic acid production and fatty acid uptake. Journal of Lipid Research 61, 1075–1086, doi:10.1194/jlr.RA120000760 (2020).
Beaulac, H. J., Gilels, F., Zhang, J., Jeoung, S. & White, P. M. Primed to die: an investigation of the genetic mechanisms underlying noise-induced hearing loss and cochlear damage in homozygous Foxo3-knockout mice. Cell Death & Disease 12, 682, doi:10.1038/s41419-021-03972-6 (2021).
Lew, A. R., Kellermayer, T. R., Sule, B. P. & Szigeti, K. Copy number variations in adult-onset neuropsychiatric diseases. Current Genomics 19, 420-430, doi:10.2174/1389202919666180330153842 (2018).
D'Angelo, C. S. & Koiffmann, C. P. Copy number variants in obesity-related syndromes: review and perspectives on novel molecular approaches. Journal of Obesity 2012, 845480, doi:10.1155/2012/845480 (2012).
Thumser, A., Voysey, J. E. & Wilton, D. C. The binding of lysophospholipids to rat liver fatty acid-binding protein and albumin. Biochemical Journal 301, 801-806, doi:10.1042/bj3010801 (1994).
Kobayashi, T., Tanaka-Ishii, R., Taguchi, R., Ikezawa, H. & Murakami-Murofushi, K. Existence of a bioactive lipid, cyclic phosphatidic acid, bound to human serum albumin. Life Sciences 65, 2185–2191, doi:10.1016/s0024-3205(99)00483-x (1999).
Sekula, B., Ciesielska, A., Rytczak, P., Koziołkiewicz, M. & Bujacz, A. Structural evidence of the species-dependent albumin binding of the modified cyclic phosphatidic acid with cytotoxic properties. Bioscience Reports 36, doi:10.1042/BSR20160089 (2016).
Stapleton, C. M. et al. Lysophosphatidic acid activates peroxisome proliferator activated receptor-γ in CHO cells that over-express glycerol 3-phosphate acyltransferase-1. PLoS One 6, e18932, doi:10.1371/journal.pone.0018932 (2011).
Tsukahara, R., Haniu, H., Matsuda, Y. & Tsukahara, T. Heart-type fatty-acid-binding protein (FABP3) is a lysophosphatidic acid-binding protein in human coronary artery endothelial cells. FEBS Open Bio 4, 947-951, doi:10.1016/j.fob.2014.10.014 (2014).
Kitakaze, K. et al. ATF4-mediated transcriptional regulation protects against β-cell loss during endoplasmic reticulum stress in a mouse model. Molecular Metabolism 54, 101338, doi:10.1016/j.molmet.2021.101338 (2021).
Okamoto, Y. et al. Sphingosine 1-phosphate receptor type 2 positively regulates interleukin (IL)-4/IL-13-induced STAT6 phosphorylation. Cellular Signalling 88, 110156, doi:10.1016/j.cellsig.2021.110156 (2021).
Tokumura, A. et al. Elevated serum levels of arachidonoyl-lysophosphatidic acid and sphingosine 1-phosphate in systemic sclerosis. International Journal of Medical Sciences 6, 168-176, doi:10.7150/ijms.6.168 (2009).
Takenouchi, Y., Kitakaze, K., Tsuboi, K. & Okamoto, Y. Growth differentiation factor 15 facilitates lung fibrosis by activating macrophages and fibroblasts. Experimental Cell Research391, 112010, doi:10.1016/j.yexcr.2020.112010 (2020).

Yes there is potential Competing Interest. K.K. and Y.O. received a grant from Teijin Pharma Limited, Bayer Yakuhin, Ltd., Kyowa Kirin Co., Ltd., Teijin Nakashima Medical Co., Ltd., Otsuka Pharmaceutical Co., Ltd., and Asahi Kasei Pharma Corporation. H.I. is an employee of Maruho Co., Ltd. All other authors declare no conflicts of interest.

SupplementaryData1.xlsx
Supplementary Data
Supplementaryinformation.docx

Download PDF

Journal Publication

published 16 May, 2023

Read the published version in Communications Biology →

Version 1

posted

You are reading this latest preprint version

Cyclic phosphatidic acid is produced by GDE7 in the ER lumen as a lysophospholipid mediator

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

GDE7 produces cPA in a cell-free system

GDE7 produces cPA in living cells

The active site of hGDE7 is directed toward the ER lumen

Mutagenesis reveals the importance of F227 and Y238 for hGDE7 catalytic activity

cPA produced by GDE7 suppresses PPARγ target genes

Discussion

Methods

Materials

Cell line

Plasmid construction and lentiviral preparation

Immunoblot analysis

Enzyme assay

Lipid analyses by liquid chromatography-tandem mass spectrometry (LC-MS/MS)

Structural Prediction

Topology

RNA isolation and quantitative PCR (qPCR)

Statistics and Reproducibility

Declarations

Data availability

Acknowledgments

Author contributions

Ethics declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1