Function of the Cloned Putative Neutral Sphingomyelinase as Lyso-platelet Activating Factor-Phospholipase C*

Sphingolipids such as ceramide and sphingosine have been regarded as novel signal mediators in cells. However, the mechanisms of generation of these lipids upon various stimulation remain to be elucidated. Neutral sphingomyelinase (N-SMase) is one of the key enzymes in the generation of ceramide, and recently the cloning of a putative N-SMase was reported. Because the function of the protein was unclear in the previous report, we investigated the role it plays in cells. N-SMase activity in cells overexpressing the protein with hexa-histidine tag was immunoprecipitated with anti-hexa-histidine antibody. The metabolism of ceramide and SM was not apparently affected in overexpressing cells. Radiolabeling experiments using [3H]palmitic acid or [3H]hexadecanol demonstrated an accumulation of 1-O-alkyl-sn-glycerol and a corresponding decrease of 1-alkyl-2-acyl-sn-glycero-3-phosphocholine in overexpressing cells. In vitro studies showed that both 1-acyl-2-lyso-sn-glycero-3-phosphocholine (lyso-PC) and 1-O-alkyl-2-lyso-sn-glycero-3-phosphocholine (lyso-platelet activating factor (lyso-PAF)) are good substrates of the protein. In further radiolabeling experiments, 1-acyl-lyso-PC was predominantly and equally metabolized into diacyl-PC in both vector and overexpressing cells. On the other hand, 1-O-alkyl-lyso-PC (lyso-PAF) was metabolized into both diradyl-PC and 1-O-alkyl-glycerol in overexpressing cells but only into diradyl-PC in vector cells. These results suggest that the protein acts as lyso-PAF-PLC rather than lyso-PC-PLC or N-SMase in cells.

A-SMase was purified from human urine, and its cDNA was cloned (11)(12)(13). Recently a knockout mouse model of A-SMase gene was established (19,20), and it was reported that radiation-induced apoptosis was suppressed in the A-SMase knockout mouse (21). However, another recent report showed that cells from a patient of Niemann-Pick disease, in which A-SMase is deficient, underwent apoptosis in a similar way to normal cells in response to Fas (22), suggesting that A-SMase is not involved in Fas-induced ceramide generation.
Advanced purification of N-SMase from rat brain was recently reported (14), but mammalian N-SMase has not been purified to homogeneity yet. However, cloning of a putative mouse and human N-SMase was recently reported by searching for sequences similar to those of bacterial N-SMase (23). Although N-SMase activity was greatly increased in cells overexpressing the cloned cDNA, it was not clear in the report whether the overexpressed protein itself has the N-SMase activity and whether SM is a physiological substrate or not.

Materials-[Choline-methyl-
Cell Culture-Human embryonic kidney 293 cells and human leukemia Molt-4 cells were purchased from ATCC. The cells were cultured in minimum essential medium (Life Technologies, Inc.) supplemented with 10% fetal calf serum at 37°C in a humidified 5% CO 2 incubator. Exponentially growing cells were used in experiments because an accumulation of monoalkylglycerol has previously been reported in confluent cells (24).
Construction of Expression Plasmid for Putative N-SMase-Two oligo-DNA primers (ATGAAGCTCAACTTCTCCCTGCGACTG and TTATTGTTCTTTAGTTCTGTCCCCCTCCTGCTG) were synthesized corresponding to the amino-terminal sense and carboxyl-terminal antisense sequences of the open reading frame from the reported putative human N-SMase cDNA (GenBank TM accession number AJ 222801), respectively. PCR was performed with these two primers using a human cDNA library generated from neuroblastoma cells as a template. A PCR product of expected molecular size (approximately 1.3 kilobase pairs) was obtained and subcloned into pT7Blue-3 (Novagen). The sequence of the PCR product was analyzed by ABI 377 DNA sequencer, and it demonstrated a 100% homology to the reported sequence of the putative human N-SMase. The insert was cut out between BamHI and XhoI sites and ligated into the same sites in pcDNA3.1/HisC and pcDNA3.1 (Invitrogen), and the construct was named pHisNSM and p3.1NSM, respectively.
Transfection-Plasmids were transfected into HEK293 cells by calcium phosphate methods as described (25). For the selection of stable transfectants, 100 g/ml of G418 (Life Technologies, Inc.) were added to the medium. Independent colonies were picked up and cultured in separate wells, and the cells from one colony expressing the highest N-SMase activity in in vitro N-SMase assays were used as pHisNSM transfectant cells.
N-SMase Assay-Cells were lysed in buffer containing 0.1% Triton X-100, 25 mM Tris (pH 7.4), 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 2 g/ml each of chymostatin, leupeptin, antipain, and pepstatin A, and centrifuged at 800 ϫ g for 5 min, and supernatant was used for the assay. 5-10 l of cell lysate was added to 100 l of reaction mixture containing 50 mM Tris (pH 7.4), 5 mM MgCl 2 , 0.05% Triton X-100, 5 mM dithiothreitol, 10 nmol [choline-methyl-14 C]SM (100000 cpm), and 10 nmol phosphatidylserine. After 30 or 60 min of incubation at 37°C 1.5 ml of chloroform/methanol (2:1) was added, the phases were separated by addition of 200 l of water, and 400 l of the upper phase was mixed with 4 ml of Safety Solve (Research Products International) for liquid scintillation counting.
PC-PLC Assay-The assay was performed as described in N-SMase assay except that reaction mixture contained no phosphatidylserine and 10 nmol [choline-methyl-14 C]PC (100,000 cpm) instead of SM.
Lyso-PAF-PLC Assay-Cell lysate was prepared as described in N-SMase assay. 5 l of cell lysate were added to 100 l of reaction mixture containing 50 mM Tris (pH 7.4), 5 mM MgCl 2 , 5 mM dithiothreitol, and 10 nmol [1-O-octadecyl-3 H]lyso-PAF (200,000 dpm). After 30 min of incubation at 37°C, the lipid was extracted by the method of Bligh and Dyer (26) and separated by TLC in solvent system A (chlorform/methanol/15 mM CaCl 2 60: 35:8) or B (chloroform/methanol/2N NH 4 OH 60: 35:5). The TLC plates were sprayed with EN 3 HANCE (NEN Life Science Products) and exposed to films at Ϫ80°C for 2 days. The band corresponding to monoalkylglycerol was scraped from the TLC plate for liquid scintillation counting.
Lyso-PC-PLC Assay-The assay was performed as described in lyso-PAF-PLC assay except that reaction mixture contained 10 nmol [1-palmitoyl- 14 C]lyso-PC (100,000 cpm) instead of lyso-PAF and the band corresponding to monoacylglycerol was scraped from the TLC plate for liquid scintillation counting.
SM Synthase Assay-Cells were suspended in the buffer containing 25 mM Tris (pH 7.4), 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 2 g/ml each of chymostatin, leupeptin, antipain, and pepstatin A, lysed by repeated sheering using 26 G needle, and centrifuged at 800 ϫ g for 5 min, and the supernatant was used for the assay as described (27) with modifications. Briefly, 10 l of cell lysate was added to 100 l of reaction mixture containing 50 mM Tris (pH 7.4), 25 mM KCl, 0.5 mM EDTA, and 10 nmol [acetyl-14 C]C 2 -ceramide (200,000 cpm). After 30 or 60 min of incubation at 37°C, 1.5 ml of chloroform/methanol (2:1) was added. The phases were separated by the addition of 0.2 ml of water, and the lipids in the lower phase were separated by TLC in solvent system A. The band corresponding to C 2 -SM was scraped from the TLC plate and transferred into 4 ml of Safety Solve for liquid scintillation counting.
Immunoprecipitation-Cells were lysed in 0.5 ml of lysis buffer (25 mM Tris (pH 7.4), 5 mM EDTA, 0.1% Triton X-100, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, and 2 g/ml each of chymostatin, leupeptin, antipain and pepstatin A), and centrifuged at 12,000 ϫ g for 15 min. 5 l of anti-His 6 monoclonal antibody (CLONTECH) was added to the supernatant and rocked at 4°C for 3 h. Then 30 l of protein G-agarose (Life Technologies, Inc.) was added, and after 2 h of incubation, the beads were centrifuged at 10,000 ϫ g for 1 min, washed twice with 1 ml of lysis buffer, and suspended in 0.5 ml of lysis buffer without protease inhibitors. For control experiments, mouse normal IgG (Santa-Cruz) was used instead of anti-His 6 antibody.
Preparation of the Membrane Fraction-Cells were lysed as described in SM synthase assay. Rat brains were homogenized in the same buffer by using a Teflon pestle glass homogenizer. Lysate was centrifuged at 800 ϫ g for 5 min, and then the supernatant was centrifuged at 100,000 ϫ g for 1 h. The pellet was used as the membrane fraction.
Partial Purification of N-SMase from Rat Brain or Molt-4 Cells-N-SMase was partially purified from the membrane fraction of rat brain or Molt-4 cells as described (28) except that HiTrap Q column (1 ml, Amersham Pharmacia Biotech) was used instead of DEAE-Sepharose column.
Mass Measurement of SM and PC-Cellular Lipids were extracted by the method of Bligh and Dyer. For the measurement of SM, the lipids in chloroform were mixed with the same volume of 0.2 N NaOH in methanol and incubated at 37°C for 1 h. The phase was separated by the addition of 0.45 volume of 0.2 N HCl. Lipids were separated by TLC in solvent system A and visualized with iodine vapor, and the bands corresponding to SM and PC were scraped. Lipids were extracted from the silica gel by the method of Bligh and Dyer, and the level of lipid phosphate was determined as described before (30).
Ceramide Measurement-The level of ceramide was determined by using diacylglycerol kinase assay as described (29).
Preparation of [9, H]Hexadecanol-[9,10-3 H]Hexadecanol was made by a modification of the procedure described (31). Briefly, the methyl ester of [9,10-3 H]hexadecanoic acid was reduced by treatment with Vitride [sodium bis(2-methoxyethoxy)aluminum hydride] (Red-Al TM , Aldrich). Following reduction, the excess vitride was destroyed by addition of ice-cold 20% ethanol, and the radioactive alcohol was extracted with chloroform. The radioactive alcohol was localized by autoradiography following a preparative TLC in a silica gel-G plate (Analtech) (hexane:diethyl ether:acetic acid 70:30:1). The corresponding spot was eluted from the TLC plate by extraction of the silica gel scrapings with chloroform:methanol (2:1). The yield (99%) and purity (Ͼ98%) of the product was assessed by TLC in the same solvent system described above. The specific activity of the product was 0.2-0.3 Ci/mmol.
Radiolabeling Experiments of the Cells-For [ 3 H]choline labeling, cells plated in 100-mm culture dishes in 8 ml of medium were labeled with 2 Ci of [ 3 H]choline. For chase experiments, cells were labeled for 4 days, and then medium was replaced by new medium, and the cells were further incubated for the indicated hours. Lipids were extracted by the method of Bligh and Dyer. For the experiments with bacterial SMase treatment, the medium was replaced by new medium, the cells were rested for 2 h, 100 milliunits/ml of bacterial SMase from Staphylococcus aureus (Sigma) was added to the medium, and after 25 min of incubation the medium was replaced by new medium, and the cells were further incubated for the indicated times. Cells were lysed in 0.6 ml of water, and a portion was used for determination of protein concentration as described (27)

Overexpression of Putative N-SMase in HEK293 Cells-Two
primers corresponding to the amino-and carboxyl-terminal of the open reading frame were synthesized from the cDNA sequence of the putative human N-SMase, and PCR was performed using a human cDNA library generated from neuroblastoma cells as a template. The PCR product of the expected size (approximately 1.3 kilobase pairs) was subcloned into pT7Blue-3, and it was confirmed that its sequence was identical to that previously reported. The insert was cut out and ligated in BamHI-XhoI sites of pcDNA3.1/HisC for hexa-histidine-tagged expression of the protein (pHisNSM). When the pHisNSM was stably transfected into human embryonic kidney 293 cells, N-SMase activity in the cell lysate was increased by approximately 20-fold compared with vector transfectant cells ( Table I). The N-SMase activity was Mg 2ϩ -dependent and stimulated by reducing reagents such as dithiothreitol and ␤-mercaptoethanol when lysis buffer did not contain reducing reagents (data not shown). However, PC-PLC activity was not increased in pHisNSM transfectant cells (Table I) in contrast to the previous report, which described an increase of PC-PLC activity equivalent to about 30% of N-SMase activity. Also, SM synthase activity was not increased in pHisNSM transfectant cells (Table I). A-SMase activity was decreased in pHisNSM transfectant cells compared with wild type or vector transfectant cells (Table I).
Immunoprecipitation of N-SMase Activity-Because the overexpressed protein was not purified in the previous report, it was not clear whether the overexpressed protein itself has N-SMase activity or not. Therefore, the overexpressed protein was immunoprecipitated by anti-His 6 antibody (Fig. 1A), and N-SMase activity of the immunoprecipitant was measured. As shown in Fig. 1B, approximately 40% of the activity were recovered in the immunoprecipitant, and only 20% remained in the supernatant. However, the same amount of mouse IgG did not immunoprecipitate either the overexpressed protein ( Fig.  1A) or N-SMase activity (Fig. 1B). We also purified the overexpressed protein with hexa-histidine tag by using TALONspin metal affinity column (CLONTECH). The molecular mass of the purified protein was about 48 kDa, consistent with the expected molecular size of the hexa-histidine-tagged protein.
The 48-kDa protein was also detected by Western blot using anti-His 6 antibody. However, the purified protein had little N-SMase activity, possibly because of the inhibition by Co 2ϩ used for the affinity column (data not shown).
Comparison of N-SMase Activity of the Overexpressed Protein with That of Partially Purified Rat Brain N-SMase-In earlier experiments we found that a high concentration (1%) of Triton X-100 in the lysis buffer inhibited N-SMase activity of the overexpressed protein, whereas activity was rather stable in the presence of 0.1% Triton X-100 in the lysis buffer. Because it was previously reported that 1% Triton X-100 in the extraction buffer did not inhibit N-SMase activity in rat brain membrane fraction (14), the effect of Triton X-100 in the extraction buffer was examined on N-SMase activity in the membrane fraction of p3.1NSM transfectant cells. The final concentration of Triton X-100 in the N-SMase assay was adjusted to 0.1% in these experiments. As shown in Fig. 2, 1% Triton X-100 in the extraction buffer inhibited N-SMase activity of the transfectant cells by approximately 80%. However, rat brain N-SMase activity was not inhibited by 1% Triton X-100 in the extraction buffer (Fig. 2) as previously reported. The effects of other detergents on N-SMase activity were also examined. As shown in Fig. 2, both 1% Nonidet P-40 and 1.5% ␤-octylglucoside in the extraction buffer inhibited N-SMase activity of transfectant cells by approximately 90%, whereas 10 mM deoxycholic acid inhibited N-SMase activity of rat brain much more than that of transfectant cells. These results suggested that the overexpressed N-SMase is distinguishable from rat brain N-SMase.
Lack of Recognition of N-SMase Purified from Molt-4 Cells by an Antibody against the Cloned N-SMase-In this experiment, p3.1NSM, which does not contain the hexa-histidine tag, was used for the overexpression of the cloned protein in the native form. A band of approximately 45 kDa was detected in p3.1NSM transfectant cells by Western blot using an antibody against the cloned human N-SMase protein (Fig. 3). However, the antibody detected little, if any, 45-kDa protein in a partially purified N-SMase preparation from Molt-4 cells. In these experiments, equal N-SMase activities were loaded. These results  The level of SM and PC were slightly higher in vector transfectant, and the level of SM was slightly lower in pHisNSM transfectant cells compared with wild type cells (Fig. 5, A and  B). The level of ceramide was increased in pHisNSM transfectant cells by approximately 20% compared with vector transfectant cells and by only 10% compared with wild type cells (Fig. 5C). These changes in the levels of SM and ceramide in overexpressing cells seemed very small compared with the 20-fold increase in in vitro N-SMase activity.
Metabolism of SM and PC in Transfectant Cells-When vector or pHisNSM transfectant cells were continuously labeled with [ 3 H]choline, the time course of the increase in labeled PC or SM was similar in both transfectant cells (Fig. 6A). There was a slight decrease in the level of labeled SM at 3 days in pHisNSM transfectant cells compared with that in vector transfectant cells. When vector or pHisNSM transfectant cells were labeled with [ 3 H]choline in equilibrium and then chased after replacement of medium, labeled SM and PC in pHisNSM transfectant cells decreased in a similar time course with those in vector transfectant cells (Fig. 6B). These results argue against any significant effect of the enzyme on the catabolism of either SM or PC.
In another set of experiments, we employed bacterial SMase to evaluate the topology of SM and its resynthesis. [ 3 H]cholinelabeled cells were treated with exogenous bacterial SMase for 25 min and then chased (Fig. 6C). Treatment with bacterial SMase diminished the level of SM by approximately 80% in both vector and pHisNSM transfectant cells. These results show that the distribution of SM between bacterial SMasesensitive (outer leaflet) and resistant (internal) pools in pHisNSM transfectant cells is similar to that in vector transfectant cells. Moreover, the level of SM did not recover at least during 24 h in both transfectants, showing that the gene under investigation did not stimulate this pool of SM synthesis. This is consistent with the result of in vitro SM synthase assay, which showed no increase of SM synthase activity (Table I), and with the result of the pulse choline labeling experiment, which showed no stimulation of SM synthesis in pHisNSM transfectant cells compared with vector transfectant cells (Fig. 6A).
Stress-induced Ceramide Generation in Transfectant Cells-Because it has been reported that N-SMase may be involved in stress-induced ceramide generation, we investigated whether overexpression of the protein might stimulate ceramide generation induced by hydrogen peroxide, a known inducer of ceramide (9,32). The level of ceramide was increased by treatment with 1 mM hydrogen peroxide in a similar time course in both vector and pHisNSM transfectant cells (Fig. 7). These results showed that the overexpressed protein is not involved in stressinduced ceramide generation.

Increase of 1-Alkyl-glycerol in pHisNSM Transfectant
Cells-To investigate whether the levels of lipids other than SM, PC, or ceramide might be changed in pHisNSM transfectant compared with vector transfectant cells, both transfectants were labeled with [ 3 H]palmitic acid, and the levels of labeled lipids were compared between vector and pHisNSM transfectant cells. The majority of labeled lipids showed no apparent difference in various TLC solvent systems. However, a band was increased in pHisNSM cells compared with vector transfectant cells (Fig. 8A). This band was scraped from the TLC plate, and the lipid was extracted by the method of Bligh and Dyer and subjected to mild base hydrolysis. The extracted lipid was resistant to mild base hydrolysis, indicating that the lipid does not contain ester linkages (data not shown). This band was not labeled with [ 3 H]dihydrosphingosine or [ 3 H]sphingosine (data not shown). These results suggested that this band is not a sphingolipid and might be a glycerolipid containing alkyl or alkenyl groups. In fact, the standard of 1-alkyl-glycerol comigrated with the band on TLC. To confirm that the lipid is 1-alkyl-glycerol, the cells were labeled with [ 3 H]hexadecanol, which can be directly metabolized into1-alkyl-glycerol. It was found that the band was more prominently labeled with [ 3 H]hexadecanol than with [ 3 H]palmitic acid in pHisNSM transfectant cells (Fig. 8B). These results further supported that the band was 1-alkyl-glycerol. When lipids from cells labeled with [ 3 H]hexadecanol were subjected to mild base hydrolysis, a decrease of 1-alkyl-lyso-PC (lyso-PAF) was detected in pHisNSM transfectant cells compared with vector transfectant cells (Fig. 8C), suggesting that 1-alkyl-2-acyl-PC was used as the source of 1-alkyl-glycerol through deacylation at sn-2 position followed by the action of the overexpressed protein (discussed below).
Lyso-PAF and Lyso-PC-PLC Activity of the Overexpressed Protein in Vitro-Assuming that the overexpressed enzyme has PLC activity, we suspected that lyso-PAF might be another substrate of the enzyme. Therefore, we conducted in vitro experiments using [1-alkyl-3 H]lyso-PAF as a substrate. As expected, lyso-PAF was a good substrate for this activity, and we observed an increased generation of 1-alkyl-glycerol (Table II). Lysate of pHisNSM transfectant cells had 20 times as high PLC activity on lyso-PAF as that of vector transfectant cells. Lyso-PAF-PLC activity was dependent on Mg 2ϩ and was stimulated by dithiothreitol (data not shown). In contrast to N-SMase activity, which requires detergent such as Triton X-100, lyso-PAF-PLC activity was detected in the absence of Triton X-100. In fact, addition of Triton X-100 in the assay inhibited lyso-PAF-PLC activity (data not shown). The protein immunoprecipitated with anti-His 6 antibody also demonstrated high lyso-PAF-PLC activity (Fig. 9). Approximately 30% of the lyso-PAF-PLC activity was recovered in the immunoprecipitant, whereas about 20% of the activity remained in the supernatant. Lyso-PC was also a good substrate of the enzyme in vitro (Table II). The K m and V max values of the immunoprecipitated protein for lyso-PAF and lyso-PC were similar (Table III). The K m and V max for SM could not be determined because N-SMase activity in the absence of Triton X-100 in the assay was very low (less than 2% compared with the activity in the presence of 0.05% Triton X-100 in the assay). The K m for SM in the presence of 0.05% Triton X-100 was similar to and V max was higher than that for lyso-PAF or lyso-PC. No PAF-PLC activity was detected in either vector or pHisNSM transfectant cells (Table  II), suggesting the requirement of 2-lyso structure of the glycerolipid substrates.
Comparison of Lyso-PAF-PLC Activity in the Overexpressed Protein and Rat Brain N-SMase-Lyso-PAF-PLC activity in partially purified rat brain N-SMase was also examined. Lyso-PAF-PLC activity was only 1% of N-SMase activity in partially purified rat brain protein (Table IV), whereas lyso-PAF-PLC activity was almost the same as N-SMase activity in the overexpressed protein (Tables I and II). These results further suggest that rat brain N-SMase is unrelated to the overexpressed protein.

DISCUSSION
In this paper we demonstrated that the cloned putative N-SMase protein does not function as N-SMase in cells and that the protein exerted lyso-PAF-PLC activity both in vitro and in cells. Because the overexpressed protein has high N-SMase activity in vitro, there seems to be two possibilities to explain why it does not act as N-SMase in cells. One possibility is the localization of the protein in cells. Approximately 80% of SM is localized in the outer leaflet of plasma membrane judging from the results using exogenous bacterial SMase, and SM involved in stress-induced responses is supposed to be located in the inner leaflet of plasma membrane or caveolae (33). It is possible that this protein is not localized in the plasma membrane so that it cannot gain access to most of SM in cells. In fact, overexpression of a green fluorescent protein fusion protein showed that this protein mainly exists in the ER. 2 Another possibility is the requirement of specific detergents for N-SMase activity. The protein has little N-SMase activity in the absence of detergents in the assay (Table III). Among the detergents tested, Triton X-100 and Nonidet P-40 could stimulate N-SMase activity in vitro, whereas CHAPS, ␤-octylglucoside and deoxycholic acid did not fully support N-SMase activity. 2 These results suggest that this protein might not exert N-SMase activity under prevailing conditions in cells. In contrast, lyso-PAF-(or lyso-PC-)PLC activity did not require detergents in vitro, and this protein also exerted lyso-PAF-PLC activity in cells.
Several criteria also suggest that the overexpressed protein is distinct from N-SMases partially purified from rat brain and Molt-4 cells. First, N-SMase activity in the overexpressed protein was inhibited by Triton X-100, Nonidet P-40, and ␤-octyl-  a K m and V max for SM could not be determined because the activity was too low in the absence of detergent in the assay (less than 2% compared with the activity in the presence of 0.05% Triton X-100 in the assay).
b The values in parentheses were determined in the presence of 0.05% Triton X-100 in the assay.  glucoside, whereas that in partially purified rat brain was not inhibited by these detergents but by deoxycholic acid. Second, the partially purified protein from Molt-4 cells was not detected by an antibody against the overexpressed protein. Third, lyso-PAF-PLC activity of the partially purified protein from rat brain was only 1% of N-SMase activity, whereas lyso-PAF-PLC activity was almost the same as N-SMase activity in the overexpressed protein. It was also reported that lyso-PC is not a good substrate for bacterial N-SMase (34). This discrepancy is now explained by the observation that the cloned enzyme is not the previously studied N-SMase. Finally, the overexpressed protein was not involved in stress-induced ceramide generation. These results suggest that the overexpressed protein is not the major N-SMase in rat brain and Molt-4 cells. In support of this is the evidence that the mRNA expression of the mouse homologue of this protein was the most abundant in kidney, whereas N-SMase activity was the highest in brain among various mouse tissues (23). It will be of great significance to identify the protein for N-SMase activity involved in stress responses.
Although this protein demonstrated PLC activity toward both lyso-PAF and lyso-PC in vitro, little 1-acyl-glycerol was accumulated when cells were labeled with lyso-PC, whereas accumulation of 1-alkyl-glycerol was detected by labeling with lyso-PAF or PAF (Fig. 10). This can be explained in several ways. One possibility is that lyso-PC is much more preferably converted into diradyl-PC by acyltransferases or transacylases than into 1-acyl-glycerol by this protein. In this case, it should be also considered that acyltransferases or transacylases prefer lyso-PC rather than lyso-PAF because lyso-PAF was almost equally converted into both diradyl-PC and 1-alkyl-glycerol. It is also possible that accumulation of 1-acyl-glycerol could not be detected because 1-acyl-glycerol was rapidly degraded by lipases, whereas 1-alkyl-glycerol accumulated because of the slower metabolism of 1-alkyl-glycerol. Another explanation is the distribution of lyso-PC in cells where this protein cannot access it as a substrate. However, this appears less probable because the structure of lyso-PC is very close to that of lyso-PAF, and one would expect similar physical properties including solubility, uptake, and partitioning.
Several reports have shown the existence of a lyso-PLD and the generation of 1-alkyl-glycerol from lyso-PAF by lyso-PLD followed by the action of phosphohydrolase (35)(36)(37). Here we demonstrated that the overexpressed protein is not lyso-PLD but lyso-PLC because not only cell lysates but also the immunoprecipitated protein produced 1-alkyl-glycerol from lyso-PAF in vitro. In fact, this protein did not generate lysophosphatidic acid even in the presence of 20 mM sodium fluoride, an inhibitor of phosphohydrolase, in the assay. 2 Because of these considerations, we propose the name lyso-PAF-PLC for this enzyme.
The significance of lyso-PLC in the metabolism of PAF has not been described before. Although the existence of lyso-PLC involved in the metabolism of PAF was reported in one study, the possibility of the involvement of lyso-PLD instead of lyso-PLC was not considered there (38). Another report described a lyso-PLC activity that hydrolyzed ether-linked lysophosphoglycerides and was involved in the synthesis of choline plasmalogens from ethanolamine plasmalogens (39). It remains to be elucidated whether this lyso-PLC activity is identical to that described here.
Because the overexpressed protein produced 1-alkyl-glycerol from lyso-PAF generated from PAF by acetylhydrolase in cells, it is highly probable that the protein is involved in the metabolic pathway of PAF in cells. PAF is a biologically very potent molecule that induces platelet aggregation and granule secretion at nanomolar concentrations (40). It is generally consid-ered that PAF is metabolized into inactive lyso-PAF by acetylhydrolase (41). Lyso-PAF can be reacylated into 1-alkyl-2acyl-PC by transacylases or acyltransferases. Lyso-PAF can be further catabolized into 1-alkyl-glycerol by lyso-PLD and phosphohydrolase. 1-Alkyl-glycerol can be degraded into fatty aldehyde by alkyl monooxygenase. The protein described in this paper can directly generate 1-alkyl-glycerol from lyso-PAF. Several possibilities for the function of this protein can be considered here. First, this enzyme can decrease the level of lyso-PAF. The role of lyso-PAF in cells is not clear, but lyso-PAF may be cytotoxic because an analogue of lyso-PAF, 1-Ooctadecyl-2-O-methyl-PC, which cannot be acylated at sn-2 position because of methylation and is therefore more stable than lyso-PAF, has a cytotoxic function (42). Second, the protein can generate 1-alkyl-glycerol, which is reported to have an inhibitory effect on protein kinase C (24). Third, the protein can decrease the level of 1-alkyl-2-acyl-PC, from which PAF is produced in the remodeling pathway upon stimulation (41). In other words, generation of PAF upon stimulation might be decreased in cells overexpressing this protein because of a decrease in the levels of 1-alkyl-2-acyl-PC and lyso-PAF. Because the remodeling pathway is involved in the generation of PAF in inflammation and hypersensitivity responses, overexpression of this enzyme will diminish these responses by decreasing the generation of PAF. These possibilities for the function of this protein will be further elucidated in the near future by experiments of gene knockout and transgenic overexpressors.