In a previous study we purified a novel lysoPLD (lysophospholipase D) which converts LPC (lysophosphatidylcholine) into a bioactive phospholipid, LPA (lysophosphatidic acid), from the rat brain. In the present study, we identified the purified 42 and 35 kDa proteins as the heterotrimeric G protein subunits Gαq and Gβ1 respectively. When FLAG-tagged Gαq or Gβ1 was expressed in cells and purified, significant lysoPLD activity was observed in the microsomal fractions. Levels of the hydrolysed product choline increased over time, and the Mg2+ dependency and substrate specificity of Gαq were similar to those of lysoPLD purified from the rat brain. Mutation of Gαq at amino acids Lys52, Thr186 or Asp205, residues that are predicted to interact with nucleotide phosphates or catalytic Mg2+, dramatically reduced lysoPLD activity. GTP does not compete with LPC for the lysoPLD activity, indicating that these substrate-binding sites are not identical. Whereas the enzyme activity of highly purified FLAG-tagged Gαq overexpressed in COS-7 cells was ~4 nmol/min per mg, the activity from Neuro2A cells was 137.4 nmol/min per mg. The calculated Km and Vmax values for lysoPAF (1-O-hexadecyl-sn-glycero-3-phosphocholine) obtained from Neuro2A cells were 21 μM and 0.16 μmol/min per mg respectively, similar to the enzyme purified from the rat brain. These results reveal a new function for Gαq and Gβ1 as an enzyme with lysoPLD activity. Tag-purified Gα11 also exhibited a high lysoPLD activity, but Gαi and Gαs did not. The lysoPLD activity of the Gα subunit is strictly dependent on its subfamily and might be important for cellular responses. However, treatment of Hepa-1 cells with Gαq and Gα11 siRNAs (small interfering RNAs) did not change lysoPLD activity in the microsomal fraction. Clarification of the physiological relevance of lysoPLD activity of these proteins will need further studies.
LPA (lysophosphatidic acid) is a potentially bioactive phospholipid that mediates a number of physiological processes, including cell adhesion, proliferation, differentiation, survival and migration. GPCRs (G-protein-coupled receptors) in the EDG (endothelial differentiation gene) and P2Y families are specific receptors for LPA in the plasma membrane, and intracellular signalling via these GPCRs has been well characterized . However, less is known about the physiological regulation of LPA production or how LPA-specific receptors in the plasma membrane are stimulated in vivo. At least two plausible enzymatic pathways for the production of LPA have been described. One of these is the deacylation of PA (phosphatidic acid) by activated PLA (phospholipase A) . In support of this idea, a PA-selective PLA1 (phospholipase A1) called mPA-PLA1α/LIPH (membrane-associated phosphatidic acid-selective phospholipase A1α/lipase member H) has been identified that hydrolyses acyl residues at the sn−1 position of PA and enhances release of LPA from the plasma membrane [3,4]. The other pathway for LPA production is direct production of equimolar amounts of LPA and choline from LPC (lysophosphatidylcholine), as catalysed by lysoPLD (lysophospholipase D) [5,6].
The physiological importance of LPA production in serum by lysoPLD was suggested by Tokumura et al.  approximately 25 years ago. In circulating blood, LPA modulates blood pressure [8,9], maintenance of pregnancy  and maturation of embryonic vessel systems . Previously, a secreted protein with lysoPLD activity was purified from plasma and serum, and was found to be identical with a molecule known as autotaxin [12,13]. Autotaxin [ENPP2 (ectonucleotide pyrophosphatase/phosphodiesterase 2)] was originally cloned as a tumour cell motility-stimulating factor and was recognized as a nucleotide pyrophosphatase on the basis of its primary structure and in vitro enzymatic characteristics . LPC, which can bind albumin and lipoproteins, is released into circulation from the liver and may be generated by lecithin/cholesterol acyltransferase. Autotaxin might hydrolyse these LPCs as substrates, and LPA accumulated to levels greater than 50 μM in plasma, presumably due to very low activity of LPA-degrading enzymes in that context .
When cells are stimulated by hormones and proliferation factors, equimolar amounts of free fatty acid and LPC are produced from PC (phosphatidylcholine) by the activated PLA2 (phospholipase A2) . The free fatty acids that are released, including arachidonic acid, which is present at position sn−2 of PC prior to cleavage, are converted into eicosanoids. Whereas the metabolism of arachidonic acid has been well characterized, the metabolism of LPC is less well understood. Because the accumulation of LPC in cells induces cell lysis , it has been thought that LPC was easily hydrolysed by lysophospholipases we cloned in a previous study , followed by conversion into a non-bioactive saturated free fatty acid and GPC (glycerophosphorylcholine). However, LPC itself was recognized recently as a bioactive phospholipid that can induce chemotaxis , signal-responsive kinase activities  and atherosclerosis .
The first enzyme with lysoPLD activity was isolated from rat brain microsomes by Wykle and Schremmer . The characteristics of this intracellular enzyme, such as substrate specificity, cation requirement and optimal pH, are different from those of autotaxin. We purified a novel lysoPLD from the rat brain that is a Mg2+-dependent enzyme and utilizes lysoPAF (1-O-hexadecyl-sn-glycero-3-phosphocholine) as a substrate , similar to the enzyme reported by Wykle and Schremmer . This enzyme might be involved in the production of LPA from LPC, which is produced from PC hydrolysed by PLA2 in cells. Previously, it has been reported that LPA receptors are expressed in perinuclear areas [22,23] and, furthermore, that the transcription factor PPARγ (peroxisome proliferation-activator receptor γ) is bound to LPA . Based on these observations, it seems reasonable to propose that intracellular lysoPLD is important for cellular responses.
In the present study, we have identified a purified protein with lysoPLD activity from the rat brain as the heterotrimeric G protein subunits Gαq and Gβ1. Mutations predicted to affect the binding of nucleotide phosphates or Mg2+ to the Gαq subunit resulted in a dramatic reduction of lysoPLD activity. Moreover, several types of Gα subunits, including Gαq and Gα11, exhibit lysoPLD activity when expressed and purified as tagged proteins.
LPC (1-palmitoyl, 16:0; stearoyl, 18:0; and oleoyl, 18:1) was obtained from Sigma–Aldrich. LysoPAF was from Alexis Biochemicals. 1-[14C]Palmitoyl-2-lyso-phosphatidylcholine was purchased from GE Healthcare. HPPA [3-(4-hydroxyphenyl) propionic acid], peroxidase and choline oxidase were purchased from Wako Chemicals. Silica Gel 60 plates and tag-purified G(β1 and γ2) complex overexpressed in Spodoptera frugiperda insect cells were purchased from Merck, and CHAPS was purchased from Dojindo. Antibodies against Gαq/11, Gβ1–4, and Gγ5 were obtained from Santa Cruz Biotechnology, and anti-FLAG antibody, anti-FLAG affinity gels, FLAG peptides, anti-actin antibody and GTP were purchased from Sigma. The monoclonal anti-autotaxin antibody was obtained from Dr Junken Aoki (Graduate School of Pharmaceutical Sciences, Tohoku University, Japan).
Identification of purified lysoPLD
Purification of lysoPLD from rat brain tissue was performed exactly as described previously . All steps in enzyme purification were carried out at 4°C. Briefly, rat brains were minced and homogenized with a Teflon homogenizer in 9 volumes of buffer A [0.3 M sucrose in 50 mM Tris/HCl (pH 7.4), 1 mM EDTA, 1 mM dithiothreitol and 0.1 mM PMSF). Nuclear fractions were obtained by centrifugation at 600 g for 10 min and resuspended in buffer A. To dissolve lysoPLD in the nuclear pellets, 10 mM CHAPS was added and the suspension was sonicated with a Microson Ultrasonic cell disruptor (Misonix). The suspensions were then centrifuged at 10000 g for 10 min and the supernatants were used as a source of enzyme. The supernatants were applied to a DEAE Cellulofine A-500 column equilibrated with buffer B [20 mM Tris/HCl (pH 7.5), 0.1 mM PMSF and 5 mM CHAPS), and the enzyme was eluted with a NaCl gradient. Active fractions were sequentially applied to three different types of columns. The enzyme fractions were concentrated using Microcon centrifugal filter devices (Millipore) and then applied to a Superdex 200 10/300 GL column. As a final step, the eluted active fractions were applied to a HiTrap DEAE FF column using a Waters 650 (Millipore). The final eluted fraction was then subjected to SDS/PAGE (10% gel) and the gel was stained with EZ Stain Silver (Atto). The major protein bands were excised for in-gel digestion with MS grade trypsin (Wako Chemicals). The mass spectra of extracted peptides were analysed by MALDI–TOF MS (matrix-assisted laser desorption ionization–time-of-flight MS) (Voyager Elite/STR Perspective, Life Technologies), and the proteins were subsequently identified using MS-Fit (http://prospector.ucsf.edu).
SDS/PAGE and immunoblotting
Proteins from rat brains and the FLAG tag affinity gel-purified fractions (10 μl) were separated on 10% acrylamide gels, and stained with a silver staining kit (Wako Chemicals) or transferred on to PVDF membranes using a semi-dry transfer apparatus (Nippon Eido) and blocked with 5% skimmed milk in TBS-T buffer [150 mM NaCl, 20 mM Tris/HCl (pH 8.0) and 0.05% Tween 20]. The membranes were incubated overnight at 4°C with primary antibody diluted 1:1000 (anti-Gαq/11, -Gβ1–4, -Gγ5, -actin or -autotaxin) or 1:5000 (anti-FLAG) in TBS-T and then for 1 h at room temperature (20°C) with a peroxide-conjugated secondary antibody diluted 1:5000 in TBS-T. Reactive bands were detected by chemiluminescence using the Bio-Rad Laboratories ChemiDoc XRS+ system.
N-terminally FLAG-tagged Gαq (GenBank® accession number NM_031036.1), Gα11 (GenBank® accession number NM_031033.1) or Gβ1 (GenBank® accession number NM_030987.2) cDNAs were amplified from cDNAs from a rat brain cDNA library using a forward primer containing the FLAG tag sequence, then subcloned into the pcDNA3.1 V5/His TOPO TA vector (Invitrogen). PCR was performed with the following pairs of primers: Gαq, 5′-GGAAGAATGGACTACAAGGACGACGATGACAAGACTCTGGAGTCCATCATGG-3′ and 5′-TCACACCAGATTGTACTCCTTCAGG-3′; Gα11, 5′-GCGACGATGGACTACAAGGACGACGATGACAAGACTCTGGAGTCCATGATG-3′ and 5′-TCACACCAGATTGTACTCCTTCAGG-3′; and Gβ1, 5′-GTGAAGATGGACTACAAGGACGACGATGACAAGAGTGAACTTGACCAGCTGC-3′ and 5′-GTTCCAGATCTTGAGGAAG-3′. The Gαi (GenBank® accession number NM_010305.1) and Gαs (GenBank® accession number AK168996.1) cDNAs were amplified from a mouse brain cDNA library with the primer pairs 5′-AAAGAATTCGCCACCATGGGCTGCAC-3′ and 5′-AAACTCGAGTTCGAAGAGACCACAGTCTTT-3′, and 5′-AAAGAATTCGCCGCCATGGGCTGCCTC-3′ and 5′-AAACTCGAGTTAGAGCAGCTCGTATTGGCG-3′ respectively. The amplicons were cloned into pcDNA3.1 His/V5 after digestion with EcoRI and XhoI. The FLAG tag was added via a site-directed mutagenesis approach using the following primers paired with their complimentary sequences: Gαi, 5′-GAATTCGCCACCATGGACTACAAGGACGACGATGACAAGGGCTGCACATTGAG-3′; and Gαs, 5′-GAATTCGCCGCCATGGACTACAAGGACGACGATGACAAGGGCTGCCTCGGC-3′. Mutations were introduced by site-directed mutagenesis using the following primers paired with complimentary sequences: Gαq (G48V), 5′-CTGCTGCTGGGGACAGTCGAGAGTGGCAAG-3′; Gαq (G48A), 5′-CTGCTGCTGGGGACAGCGGAGAGTGGCAAG-3′; Gαq (K52A), 5′-GACAGGGGAGAGTGGCGCGAGTACCTTCATAAG-3′; Gαq (T186A), 5′-GTTCGAGTCCCCGCCACAGGGATCATTG-3′; Gαq (D205A), 5′-TCTTCAGAATGGTCGCTGTAGGAGGCCAAAGG-3′; Gαq (Q209L), 5′-TGTAGGAGGCCTAAGGTCAGAGA-3′; and Gβ1 (H311A), 5′-GTCCTAGCTGGAGCTGACAACCGAGTCAGC-3′. The sequences of the constructs were verified by direct DNA sequencing (ABI PRISM 377-XL, Applied Biosystems).
The isotopic lysoPLD activity assay was performed as described previously . Briefly, 30 μl of the source of enzyme was incubated at 37°C for 6 h in a reaction mixture containing 0.15 mM 1-[14C]palmitoyl-GPC (6000 d.p.m./nmol), 20 mM Tris/HCl (pH 7.0), 1 mM Na3VO4, and 50 mM MgCl2 with or without 1 mM GTP in a final volume of 75 μl. The reactions were terminated by the addition of 15 μl of 2 M HCl and 187.5 μl of chloroform/methanol/HCl [100:200:1 (v/v)]. The lipids were extracted by the addition of 93.75 μl each of chloroform and 2 M KCl, followed by centrifugation at 500 g for 10 min at 20°C. The extracted lipids were subjected to two-dimensional TLC [chloroform/methanol/28% ammonia at 65:35:5 (v/v) for the first dimension and chloroform/acetone/methanol/acetic acid/water at 45:20:10:13:5 (v/v) for the second dimension]. The LPA spots were visualized and quantified using the Fuji BAS2000 system (Fujifilm).
For the colorimetric assay to detect lysoPLD activity, the amount of choline released from choline lysophospholipids was used as a measure of enzyme activity. Purified proteins (40 μl) were incubated at 37°C with 0.15 mM (0.0375–0.6 mM) choline lysophospholipids in the presence of 20 mM Tris/HCl (pH 7.0) and 1 mM Na3VO4 with or without 50 mM MgCl2 and/or 100 mM EDTA in a total reaction volume of 200 μl. After incubation for the given lengths of time, the reactions were terminated by boiling. To determine the amount of choline released, a second reaction was performed at 37°C for 15 min in a 500 μl reaction mixture containing the first assay mixture plus 50 mM Tris/HCl (pH 8.5), 0.5 mM HPPA, 0.033 unit/ml horseradish peroxidase and 1 unit/ml choline oxidase. The fluorescence intensity of each mixture was determined by excitation at 320 nm and collection at 404 nm using an RF5300 instrument (Shimadzu).
Purification of FLAG-tagged proteins
Mouse hepatocytoma cells (Hepa-1 cells), COS-7 cells or Neuro2A cells were seeded into 100 mm dishes in DMEM (Dulbecco's modified Eagle's medium; Wako Chemicals) with 10% fetal bovine serum (Gibco) 1 day prior to transfection. A total of 24 μg of empty vector or FLAG-tagged protein expression vector were transfected into cells using Lipofectamine™ 2000 (Invitrogen) according to the manufacturer's instructions. At 1 day (24 h) post-transfection, the cells in one or five dishes were collected in 800 μl/dish of buffer C [0.25 M sucrose in 50 mM Tris/HCl (pH 7.4), 1 mM EDTA and 1 mM PMSF] and disrupted by sonication (three to eight pulses of 10 s each) with a Microson Ultrasonic cell disruptor. The homogenates were then centrifuged at 105000 g for 1 h at 4°C and the supernatants were removed or used for FLAG-tagged protein purification as indicated below. The pellets were resuspended in 400 μl/dish of buffer C with 10 mM CHAPS, and homogenized with an Ultrasonic cell disruptor. The suspensions were diluted to 1600 μl/dish of buffer C, then centrifuged at 21500 g for 10 min at 4°C and the supernatants were mixed with 250 μl of anti-FLAG M2 affinity gel (Sigma–Aldrich) suspended in 1.6 ml/dish of TBS buffer [50 mM Tris/HCl (pH 7.4), 150 mM NaCl and 1mM PMSF]. After incubation with gentle agitation at 4°C for several hours, the gel was washed with 5 ml of wash buffer (TBS with 0.1% CHAPS), and then tagged proteins were eluted with 250 μl of 100 μg/ml FLAG peptide in wash buffer four times. The viability of cells transfected with control or FLAG-tagged protein expression vectors did not change over 24 h and nearly equal amounts of protein were obtained in each sample.
Effect of mouse Gαq and Gα11 siRNAs (small interfering RNAs) on lysoPLD activity in Hepa-1 cells
Stealth siRNA, a 25-base pair duplex oligoribonucleotide against mouse Gαq or Gα11, was purchased from Invitrogen. The target sense sequences were: Gαq, 5′-CCCUUUGACUUACAAAGUGUCAUUU-3′; and Gα11, 5′-CCAAGUUGGUGUACCAGAACAUCUU-3′. Hepa-1 cells (1×105) were cultured in 60 mm dishes for 24 h, and then transfected with both siRNAs by using Lipofectamine™ 2000 (Invitrogen) according to manufacturer's instructions. After 48 h, the cells were collected in 400 μl/dish of buffer C and disrupted by sonication (three to eight pulses of 10 s each) with a Microson Ultrasonic cell disruptor. The homogenates were then centrifuged at 105000 g for 1 h at 4°C and the supernatants were removed. The pellets were resuspended in 300 μl/dish of buffer C with 10 mM CHAPS, and homogenized with an Ultrasonic cell disruptor. The suspensions were centrifuged at 21500 g for 10 min at 4°C, and the supernatants were used as microsomal fractions for lysoPLD assay with 10 μg of the protein for a 30 min incubation.
All values are expressed as the means±S.D. (n≥3). Group means were compared using the Student's t test or ANOVA to determine the significance of the differences among the individual means. Statistical significance was assumed at P<0.05. Each experiment was repeated at least twice with similar results.
A protein with lysoPLD activity is highly associated with heterotrimeric G protein
In a previous study, we purified a novel protein with lysoPLD activity from rat brain tissue and analysed its enzymatic characteristics, which were different from those of the secreted lysoPLD autotaxin. Initially, a 35 kDa protein was the main band in purified lysoPLD from rat brain and did not react with an anti-autotaxin antibody . However, as shown in Figure 1(A), two bands (faint 42 kDa and 35 kDa bands, indicated by arrows on the right-hand side) were observed following repeated purifications as reported previously , separation by SDS/PAGE and silver staining. The 42 kDa protein was present at lower levels and was not recognized as potentially important at first. To identify their molecular species, the bands were analysed by PMF (peptide mass fingerprinting). The 42 kDa and 35 kDa proteins were identified as Gαq (Figure 1D) and Gβ1 (Figure 1E) respectively.
Identification of G protein subunits Gαq and Gβ1 in purified lysoPLD fractions from the rat brain
To confirm this result, the fractions obtained from the final step of enzyme purification with HiTrap DEAE FF (Figure 1C) were subjected to immunoblot analysis with antibodies specific against Gαq/11 and Gβ1–4. As shown in Figure 1(B), both Gαq and Gβ1 proteins were detected in those fractions and the levels of those proteins correlated well with lysoPLD activities in each fraction. On the basis of these results, we speculated that the heterotrimeric G protein complex is highly associated with lysoPLD activity.
Tag-purified G protein exhibits lysoPLD activity
FLAG-tagged Gαq or Gβ1 proteins were overexpressed in the mouse hepatocytoma cell line Hepa-1 and then purified via immune-affinity chromatography with an anti-FLAG antibody. When overexpressed, FLAG–Gαq was purified with FLAG peptides, endogenous Gβ co-purified and, similarly, endogenous Gαq/11 co-purified with FLAG-tagged Gβ1 (Figure 2A). Gγ5 also co-purified with FLAG–Gαq (Figure 2A). These results suggest that heterotrimeric G protein complexes can form with overexpressed FLAG–Gαq. The density of the immunoreactive bands with anti-Gαq/11 or anti-Gβ1–4 antibody and lysoPLD activities coincided well (Figures 2A and 2B); thus we used fraction 4 for subsequent analyses. The significant increase in lysoPLD activity was not observed in fractions prepared from cells transfected with empty vector. The lysoPLD activity was confirmed by choline production from lysoPAF (Figure 2C) and LPA production from [14C]palmitoyl-labelled LPC, (Figures 2D and 2E). Significant increases in both choline and/or LPA production via lysoPLD activity were detected in the eluates prepared from cells overexpressing FLAG–Gαq or FLAG–Gβ1. Autotaxin was not detected in the tag-purified fraction by immunoblot analysis (Figure 2A, indicated by the arrow), and lower bands were non-specific bands. LysoPLD activity of autotaxin is not dependent on Mg2+ supplementation . These results suggest that the lysoPLD activities were attributable to Gαq or Gβ1 themselves, or to a protein(s) in a complex with these proteins.
LysoPLD activity of tag-purified G protein subunits Gαq and Gβ1
Time course, substrate specificity and Mg2+ cation requirement of lysoPLD activity of Gαq
To examine the time course of choline production from lysoPAF hydrolysed by purified FLAG-tagged Gαq subunit, the enzyme was incubated with 0.15 mM lysoPAF for 0–6 h and activities were tested using a colorimetric assay. As shown in Figure 3(A), the levels of the hydrolysed product, choline, significantly increased in the presence of purified FLAG-tagged Gαq for up to 6 h of incubation.
Time dependence, substrate specificity and requirement of Mg2+ cations of the lysoPLD activity of tag-purified Gαq
To identify a preferred substrate, purified FLAG-tagged Gαq was incubated for 6 h with several different choline lysophospholipid substrates (Figure 3B). The preferred substrate identified in this assay is lysoPAF (P<0.01). The enzyme was more efficient at hydrolysis of 1-palmitoyl-2-lyso-GPC (16:0) than hydrolysis of 1-oleoyl-2-lyso-GPC (18:1) (P<0.01). The substrate specificity of Gαq for lysophospholipids is very similar to what was observed for the enzyme activity previously purified from the rat brain . Choline production from lysoPAF by the purified FLAG–Gαq was dependent on incubation with Mg2+ (Figure 3C), similar to what was observed for the enzyme purified from rat brain tissue . Although after centrifugation at 105000 g high levels of FLAG–Gαq could be purified from the supernatants of homogenates from cells overexpressing the protein, enzymatic activity was not detected (Figure 3D).
The lysoPLD activity of Gαq and Gβ1 mutant proteins
As FLAG–Gαq purified from Hepa-1 cells overexpressing the protein co-purified with endogenous Gβ and purified FLAG-Gβ1 co-purified with endogenous Gαq/11 (Figure 2A), we next tried to clarify which subunit is responsible for the lysoPLD activity we observed. Autotaxin had been identified previously as a member of the ecto-nucleotide PDE (pyrophosphatase/phosphodiesterase) family due to similarity in its primary structure [14,26]. Later, it became clear that the ester bond between phosphate and choline in LPC was also hydrolysed by autotaxin [12,13]. Therefore we speculated that Gαq subunit might exhibit lysoPLD activity because it hydrolyses GTP to GDP via its intrinsic GTPase activity. The Gβ1 subunit would be the candidate for the protein with lysoPLD activity, as a histidine acid phosphatase consensus motif is present in its C-terminus (V305GILSGHDNAVSCLGV320, analysed by GENETYX). To test this, we introduced several mutations at amino acid residues thought to be important for Gαq or Gβ1 enzyme activities. The mutant proteins were then purified and their lysoPLD activities were assayed. Tag-purified Gβ1–Gγ2 complex overexpressed in S. frugiperda insect cells was also obtained, and its lysoPLD activity was determined.
As shown in Figure 4(A), the K52A, T186A, D205A, G48A and Q209L mutant forms of Gαq showed a significant reduction of lysoPLD activity, whereas G48V did not. The decrease in lysoPLD activity of G48A and Q209L was low. Based on amino acid sequence similarity to p21ras, the amino acid sequence motif G46TGESGKS53 is predicted to form a loop in which main-chain amide hydrogens of several amino acids and the ϵ-amino group of Lys52 might form bonds with the β- and γ-phosphate of GTP . This site also corresponds to the GTP-binding site. Therefore we speculated that the side chain of Lys52 might be important for the binding of phosphate at the sn−3 position of LPC, whereas the side chain of Gly48 is not likely to be important for this binding. This suggests that the GTP- and LPC-binding sites or interacting amino acids of Gαq seem to be not completely the same. Moreover, the side chain of Thr186 may point away from the bound nucleotide in the GDP-bound form, but flip towards the nucleotide in the GTP-bound form, where it might directly interact with a Mg2+ ion coordinated with oxygen molecules in the β- and γ-phosphates of GTP. On the basis of crystal structure analysis, it has been suggested that the side chain of Asp205 in Gαq might bind catalytic Mg2+ via an intervening water molecule [27,28]. The well-known dominant negative Gαq (Q209L) construct was prepared and its activity was assayed. Only slight reduction of lysoPLD activity of Q209L was detected. The Gβ1 mutant H311A, which disrupts a conserved amino acid residue in the histidine acid phosphatase consensus, did not show a significant change in lysoPLD activity as compared with the wild-type Gβ1 subunit (Figure 4B). Tag-purified Gβ1–Gγ2 complex overexpressed in insect cells did not exhibit lysoPLD activity. These results strongly suggest that Gαq exhibits lysoPLD activity and, furthermore, that the predicted nucleotide phosphate and Mg2+-binding sites of the protein are important for lysoPLD enzymatic activity. When we purified lysoPLD from rat brain or culture cells, Gαq and Gβ were always co-purified. It is possible that lysoPLD activity of Gαq is dependent on the presence of Gβ protein.
LysoPLD activity of mutant forms of the G protein subunits Gαq and Gβ1, and tag-purified Gβ1–Gγ2 complex
LysoPLD activity of Gαq protein and a mutant form, T186A, highly purified from COS-7 cells
To eliminate the possibility that the detection of lysoPLD activity from purified tagged Gαq protein is cell-line dependent, we performed the same experiment with COS-7 cells, which do not exhibit autotaxin activity and express relatively low endogenous levels of G proteins. As shown in Figure 5(A), highly purified wild-type FLAG–Gαq protein and a mutant form, T186A, could be detected following ectopic expression in COS-7 cells. The levels of the hydrolysed product, choline, significantly increased over time in the presence of purified FLAG-tagged Gαq (Figure 5B). The activity was dependent on incubation with Mg2+ cations (Figure 5C). The presence of lysoPLD activity was confirmed by the significant production of choline from lysoPAF (Figures 5B–5D) and LPA from LPC (Figures 5E and 5F). The T186A mutant form of Gαq showed significantly less lysoPLD activity than the wild-type FLAG–Gαq protein. These results clearly show that Gαq has lysoPLD activity. As shown in Figure 5(A) (bottom panel), a small amount of endogenous Gβ in COS-7 cells was also co-purified with intact or mutated FLAG-tagged Gαq.
LysoPLD activity of highly purified Gαq protein and a mutant form, T186A, from COS-7 cells
LysoPLD activity of Gβ1 protein highly purified from COS-7 cells
To examine the possibility that purified FLAG–Gβ1 obtained from COS-7 cells also has lysoPLD activity, we repeated the experiment described above with these cells. A small amount of endogenous Gαq/11 in COS-7 cells was co-purified with FLAG–Gβ1 (Figure 6A). Almost the same lysoPLD activity as for purified FLAG–Gαq protein was observed, in FLAG–Gβ1 with the production of choline from lysoPAF (Figure 6B) and LPA from LPC (Figure 6C). When purified FLAG–Gαq and FLAG–Gβ1 proteins were incubated together in vitro, an additive effect of lysoPLD activity was observed, as specific activity did not change (Figure 6B).
LysoPLD activity of highly purified Gβ1 protein from COS-7 cells
Effect of substrate concentration and GTP on the activity of FLAG-purified lysoPLD from Neuro2A cells overexpressing FLAG–Gαq
Because the lysoPLD enzyme that we have purified previously was from the brain, we also tested overexpression and tag-affinity purification of FLAG–Gαq using Neuro2A cells. As shown in Figures 7(A) and 7(B), the level of lysoPLD activity of purified FLAG-Gαq was dependent on cell type (COS-7<Hepa-1<Neuro2A cells). Notably, the enzyme activity of purified FLAG-tagged Gαq expressed in COS-7 cells was approximately 4 nmol/min per mg, whereas the activity from Neuro2A cells was much higher, i.e. 137.4 nmol/min per mg. The calculated Km and Vmax values for lysoPAF using Neuro2A cells were 21 μM and 0.16 μmol/min per mg respectively (Figures 7B and 7C). The Km and Vmax values of lysoPLD obtained from rat brain for lysoPAF were 26.7 μM and 0.29 μmol/min per mg respectively. These results using Neuro2A cells and rat brain were almost equivalent with each other. The slightly lower Vmax value in Neuro2A cells is probably due to the purity of the enzyme or the disturbance of enzyme activity by N-terminal FLAG peptides.
Effect of substrate concentration and GTP on the activity of FLAG-purified lysoPLD obtained from Neuro2A cells overexpressing FLAG–Gαq
As shown in Figure 7(D), 1 mM GTP did not disturb the lysoPLD activity (0.15 mM LPC) of FLAG–Gαq obtained from Neuro2A cells. As the mutation at the GTP-binding region (Gly48) did not decrease lysoPLD activity of Gαq much (Figure 4A), the GTP- and LPC-binding sites seem to not be completely the same. Moreover, the exchange of GDP to GTP and activation of GTPase activity of Gαq possibly rarely occurred in this experiment. As for most of the GTPases, the rate of exchange of GDP to GTP and GTPase activity of Gαq are quite low without ligand-stimulated receptor and GAPs (GTPase-activating proteins) [29–31]. LysoPLD activity of autotaxin may be inhibited by incubation with GTP because autotaxin catalyses the hydrolysis of GTP as well as ATP.
LysoPLD activity of other Gα subfamily members and the effect of Gαq and Gα11 siRNAs
In the human genome, 16 different Gα subunits have been identified and these can be divided into four subfamilies based on their amino acid similarities and the machinery for their signal transduction. Thus we were interested in the possibility that other Gα protein family members might also exhibit lysoPLD activity. To test this, we first constructed plasmids encoding FLAG-tagged Gα11, Gαi and Gαs. The plasmids were introduced into Hepa-1 cells and the activities of the purified overexpressed proteins were analysed. As shown in Figure 8(A), purified FLAG–Gα11, which has 89% sequence similarity with Gαq and belongs to the same subfamily, exhibited a high lysoPLD activity (P<0.01), whereas FLAG–Gαi and FLAG–Gαs did not. However, a slight increase in lysoPLD activity of FLAG–Gαs was observed.
Testing LysoPLD activity of other Gα subfamily members and effect of Gαq and Gα11 siRNAs
As shown in Figure 8(B), pretreatment of Gαq and Gα11 siRNAs clearly decreased the endogenous Gαq and Gα11 in Hepa-1 cells. However, lysoPLD activity in the crude microsomal fraction did not changed compared with the control. In the crude microsome, we speculated that other undetermined G proteins or contaminated autotaxin may be responsible for the high endogenous lysoPLD activity and that the lysoPLD activity of the interested protein may be masked.
We were interested to molecularly identify a lysoPLD activity that had been reported previously. Two proteins from the rat brain exhibiting lysoPLD activity (faint 42 kDa and 35 kDa bands by gel electrophoresis) can be observed following a method reported previously for purification . As shown in Figure 1(A), the 42 kDa protein was present at lower levels and, at first, was not recognized as potentially important. The enzymatic properties of the fractions containing both proteins, including optimum pH, substrate specificity and cation requirement, were different from those of a secreted protein with lysoPLD activity, autotaxin. In the present study, we analysed the 42 kDa and 35 kDa proteins by MS and identified them as the heterotrimeric G protein subunits Gαq and Gβ1 respectively. When FLAG-tagged Gαq or Gβ1 was overexpressed in Hepa-1 cells and purified, both purified subunits exhibited lysoPLD activity (Figure 2). Mutations in conserved amino acid residues of Gαq known to be important for GTPase activity and interaction with nucleotide phosphates or Mg2+ significantly reduced lysoPLD activity (Figure 4). Furthermore, highly purified FLAG–Gαq protein ectopically expressed in COS-7 cells clearly hydrolysed LPC to LPA, whereas a mutant form, T186A, showed significantly less lysoPLD activity (Figure 5). The calculated Km and Vmax values for lysoPAF of the FLAG-purified enzyme obtained from Neuro2A cells were 21 μM and 0.16 μmol/min per mg respectively (Figures 7B and 7C). These values were similar to the enzyme purified from the rat brain (Km=26.7 μM and Vmax=0.29 μmol/min per mg) (21). Although Gβ1 detected by silver staining was the main protein of purified lysoPLD from rat brain, Gβ1 is not likely to be the main component for lysoPLD activity because of following four reasons: (i) FLAG–Gβ1 mutant (H311A) exhibited significant lysoPLD activity (Figure 4B); (ii) FLAG–Gα11 co-purified with lower amounts of Gβ1 exhibited lysoPLD activity comparable with Gαq (Figure 8A); (iii) FLAG–Gαi and FLAG–Gαs did not exhibit as high a lysoPLD activity as FLAG–Gαq when they were co-purified with Gβ (Figure 8A); and (iv) tag-purified Gβ1 and Gγ2 overexpressed in insect cells did not exhibit lysoPLD activity (Figure 4A). However, the potential importance of Gβ1 for lysoPLD activity of Gαq may be possible because they were always co-purified with respect to lysoPLD activity. These results reveal a new function for Gαq and Gβ1, which are well known as multifunctional signal transduction proteins, as an enzyme with lysoPLD activity.
Although FLAG-purified Gαq obtained from the microsomal fraction exhibited significant lysoPLD activity, the purified enzyme from supernatants after 105000 g centrifugation did not have lysoPLD activity (Figure 3D). The level of lysoPLD activity of purified FLAG–Gαq from the microsomal fraction was dependent on cell type (COS-7<Hepa-1<Neuro2A cells) (Figure 7A). These results suggest that post-translational modification(s), or some kind of activator(s) present in the microsome, is important for lysoPLD activity of Gαq. However, it is also possible that Gαq proteins in supernatants were misfolded after translation.
Agonist-activated GPCRs enhance GDP/GTP exchange on the G protein α subunit, thereby generating an active GTPbound Gα subunit, followed by dissociation from the receptor and Gβγ subunit. The onset and termination of G protein signalling is determined by the length of time spent in the active GTP-bound state. GTP is hydrolysed to GDP by the intrinsic GTPase activity of the Gα subunit, resulting in re-association of the Gα subunit with the Gβγ subunit and termination of signalling . As for most of GTPases, the rates of exchange of GDP to GTP and GTPase activity of Gαq are quite low. However, the GTPase activity of Gαq can be dramatically increased by in vitro incubation with ligand-stimulated cholinergic receptor and GAPs, such as the RGS (regulator of G protein signalling) proteins  or phospholipase C . In the present study, we found evidence that the lysoPLD activity of FLAG-tagged Gαq is exhibited in the presence of Gβ and Gγ5, as endogenous Gβ and Gγ5 co-purified with FLAG–Gαq (Figure 2). We speculate that the heterotrimeric form of G protein subunits may exhibit lysoPLD activity.
Previously, GTPγS was found to be bound to the purified activators  and, additionally, that purified PI-PLC (phosphoinositide phospholipase C) from the bovine brain could be activated by purified Gαq in the presence of GTPγS (0.1–1 mM) . On the basis of our results, we propose that a possible role of the novel lysoPLD identified in the present study for Gαq and Gβ1 is the production of LPA inside cells in response to ligand stimulation, which is then followed by sequential activation of PI-PLC (Ca2+ influx) and PLA2s that hydrolyse PC to free fatty acid and LPC . It was reported previously that LPA receptors can be detected on the nuclear membrane [22,23] and that the transcription factor PPARγ acts as a LPA receptor . It seems reasonable to speculate that the Gαq and Gβ1 subunits mediate the signal to these intracellular receptors via lysoPLD activity.
After testing several family members as shown in Figure 8(A), Gαq and Gα11 exhibit a high lysoPLD activity, but Gαi and Gαs do not. Therefore we conclude that the lysoPLD activity of the Gα subunit is strictly dependent on its subfamily. However, treatment of Hepa-1 cells with Gαq and Gα11 siRNAs did not affect lysoPLD activity in the microsomal fraction (Figure 8B). It may be possible that other undetermined G proteins and/or different kinds of lysoPLD, such as autotaxin, are responsible for the high endogenous lysoPLD activity and that they masked the reduced activity. Furthermore, our preliminary experiments showed that the stimulation of radio-labelled cells by carbachol that activates Gαq/11 through the muscarinic receptor did not affect the basal amount of LPA (results not shown). We speculate that the degradation of LPA in cells was rapid or that the stimulation system was inadequate. Further investigation is required to help elucidate the importance of LPA production by Gαq in signal transduction.
heterotrimeric G protein α subunit q
heterotrimeric GTP-binding protein β subunit 1
3-(4-hydroxyphenyl) propionic acid
matrix-assisted laser desorption ionization–time-of-flight MS
phosphoinositide phospholipase C
peptide mass fingerprinting
peroxisome proliferation-activator receptor γ
small interfering RNA
The project strategy was devised by, and most of the experiments were performed by, Chieko Aoyama and Hiroyuki Sugimoto. Purification of proteins was helped by Hiromi Ando, Satoko Yamashita and Sayaka Sugimoto. siRNA experiments and PMF analysis were performed by Yasuhiro Horibata and Motoyasu Satou respectively.
We thank Dr Takashi Namatame of the Medical Research Center for help with DNA sequencing and the Research Support Center for allowing us to use the facilities at Dokkyo Medical University School of Medicine. We also thank Dr Junken Aoki (Graduate School of Pharmaceutical Science, Tohoku University, Japan) for providing the monoclonal anti-autotaxin antibody.
This work was supported by a Grant-in-Aid for Young Scientists from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
These authors contributed equally to this work.