We previously purified lysophospholipase D (lysoPLD), which hydrolyzes lysophosphatidylcholine (lysoPC) to lysophosphatidic acid (LPA), from rat brain and identified the heterotrimeric G protein subunits G αq and G β1 in the lysoPLD active fractions. Tag-affinity purified G αq exhibits lysoPLD activity but a mutant that affected cellular localization or interaction with the G β subunit reduced lysoPLD activity. Size exclusion chromatography revealed that active lysoPLD is a much higher molecular weight complex than is heterotrimeric G protein, suggesting the presence of other components. Liquid chromatography-tandem mass spectrometry of lysoPLD purified from rat brain identified glycerophosphodiesterase 4 (GDE4), recently reported as lysoPLD, in the same fraction as G proteins. The over-expressed and tag-purified G αq fractions, which exhibit lysoPLD activity, contained GDE4. Exogenously expressed GDE4 was co-immunoprecipitated with endogenous G αq and G β and exhibited high lysoPLD activity. The results of confocal microscopy and cell fractionation experiments indicated that exogenously expressed GDE4 in cells mainly localized at the ER and partially co-localized with G αq protein at the plasma membrane. Proteinase K protection assay results suggested that the catalytic domain of GDE4 faces the lumen/extracellular space. Mutations at the conserved amino acids in the C-terminus cytoplasmic regions amongst GDE1, 4 and 7, dramatically suppressed GDE4 enzyme activities. When both the G αq and G α11 genes in Neuro2A cells were disrupted using the CRISPR-Cas9 system, endogenous lysoPLD activity was partially reduced but rescued by over-expression of G αq . These results suggest that GDE4 is a new effector of G protein signaling that produces bioactive phospholipid LPA and/or modulates membrane homeostasis.

Pcyt2 (CTP:phosphoethanolamine cytidylyltransferase) is the rate-limiting enzyme in mammalian PE (phosphatidylethanolamine) biosynthesis. Previously, we reported that Pcyt2 mRNA levels increased in several types of cells after serum starvation, an effect that could be suppressed by supplementation with low-density lipoprotein or 25-HC (25-hydroxycholesterol). Transcription of Hmgcr , which encodes 3-hydroxy-3-methylglutaryl-CoA reductase, is also suppressed by 25-HC in the same dose-dependent manner. Nevertheless, a sterol-regulatory element was not detected in the Pcyt2 promoter region. The important element for transcriptional control of Pcyt2 by 25-HC (1.25 μM) was determined to reside between −56 and −36 on the basis of analysis with several Pcyt2 promoter deletion–luciferase reporters in NIH 3T3 cells. Using the yeast one-hybrid system, we found that NF-Y (nuclear factor-Y) binds at C −37 CAAT −41 and YY1 (Yin Yang1) binds at C −42 AT −40 in the Pcyt2 promoter. Endogenous NF-Y and YY1 bind clearly and competitively to these sites and are important for basal Pcyt2 transcription. Moreover, NF-Y binds to the Hmgcr promoter at C −14 CA −12 in gel-shift analysis, and suppression of the basal luciferase activity of the Hmgcr promoter–reporter construct (−30/+61) by 25-HC was abolished when C −14 CA −12 was mutated. Furthermore, transcriptional suppression of Pcyt2 by 25-HC was reduced following knockdown targeting of NF-YA or YY1. ChIP analysis revealed that 25-HC inhibited the interaction between NF-Y and RNA polymerase II on the Pcyt2 and Hmgcr promoters. On the basis of these results, we conclude that NF-Y and YY1 are important for the basal transcription of Pcyt2 and that NF-Y is involved in the inhibitory effects of 25-HC on Pcyt2 transcription.

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 Gα q and Gβ 1 respectively. When FLAG-tagged Gα q or Gβ 1 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 Mg 2+ dependency and substrate specificity of Gα q were similar to those of lysoPLD purified from the rat brain. Mutation of Gα q at amino acids Lys 52 , Thr 186 or Asp 205 , residues that are predicted to interact with nucleotide phosphates or catalytic Mg 2+ , 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 Gα q 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 K m and V max 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 Gα q and Gβ 1 as an enzyme with lysoPLD activity. Tag-purified Gα 11 also exhibited a high lysoPLD activity, but Gα i and Gα s 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 Gα q and Gα 11 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.

In order to search for novel components of lipid membrane microdomains involved in neural signalling pathways, mAbs (monoclonal antibodies) were raised against the detergent-insoluble membrane fraction of PC12 (pheochromocytoma) cells. Among the 22 hybrid clones, mAb PR#1 specifically detected a fucoganglioside Fuc(Gal)-G M1 [α-fucosyl(α-galactosyl)-G M1 ], a ganglioside homologous with G M1a (II 3 NeuAc,GgOse 4 Cer), as a novel member of microdomain components with biological functions. In the presence of mAb PR#1 in the culture medium, the outgrowth of neurites was induced in PC12 cells in a dose-dependent manner, with no effects on cell proliferation, suggesting that Fuc(Gal)-G M1 is preferentially involved in PC12 cell neuritogenesis. Effects through Fuc(Gal)-G M1 were different from those through G M1a during differentiation, e.g. under PR#1 treatment on Fuc(Gal)-G M1 , round cell bodies with thinner cell processes were induced, whereas treatment with CTB (cholera toxin B subunit), a specific probe for G M1a , produced flattened cell bodies with thicker pro-cesses. Molecular analysis demonstrated that the PR#1–Fuc(Gal)-G M1 pathway was associated with Fyn and Yes of the Src family of kinases, although Src itself was not involved. No association was found with TrkA (tropomyosin receptor kinase A) and ERKs (extracellular-signal-regulated kinases), which are responsible for G M1a -induced differentiation. From these findings, it is suggested that a fucoganglioside Fuc(Gal)-G M1 provides a functional platform distinct from that of G M1a for signal transduction in PC12 cell differentiation.