A Novel Plasmal Conjugate to Glycerol and Psychosine (“Glyceroplasmalopsychosine”)
2001; Elsevier BV; Volume: 276; Issue: 25 Linguagem: Inglês
10.1074/jbc.m101288200
ISSN1083-351X
AutoresToshiyuki Hikita, Keiko Tadano‐Aritomi, Naoko Iida‐Tanaka, Jasbir K. Anand, Ineo Ishizuka, Sen‐itiroh Hakomori,
Tópico(s)Cellular transport and secretion
ResumoA novel plasmal conjugate of glycosphingolipid having cationic lipid properties was isolated from the white matter of bovine brain. Linkage analysis of galactosyl residue by methylation, liquid secondary ion, and electrospray ionization mass spectrometry of intact and methylated derivatives, and by 1H- and 13C-NMR spectroscopy, identified the structure unambiguously as an O-acetal conjugate of plasmal to the primary hydroxyl group of glycerol and to the 6-hydroxyl group of galactosyl residue of β-galactosyl 1→1 sphingosine (psychosine). This novel compound is hereby termed “glyceroplasmalopsychosine”; its structure is shown below. A novel plasmal conjugate of glycosphingolipid having cationic lipid properties was isolated from the white matter of bovine brain. Linkage analysis of galactosyl residue by methylation, liquid secondary ion, and electrospray ionization mass spectrometry of intact and methylated derivatives, and by 1H- and 13C-NMR spectroscopy, identified the structure unambiguously as an O-acetal conjugate of plasmal to the primary hydroxyl group of glycerol and to the 6-hydroxyl group of galactosyl residue of β-galactosyl 1→1 sphingosine (psychosine). This novel compound is hereby termed “glyceroplasmalopsychosine”; its structure is shown below. glycosphingolipid collision-induced dissociation carboxymethyl chloroform/methanol electrospray ionization mass spectrometry gas chromatography-mass spectrometry glycerol heteronuclear multiple bond connectivity liquid secondary ion mass spectrometry protein kinase C plasmalopsychosine plasmalopsychosine A plasmalopsychosine B major plasmal-containing unknown compound sphingosine double-quantum-filtered correlated spectroscopy two-dimensional homonuclear Hartmann-Hahn spectroscopy rotating frame Overhauser effect spectroscopy NeuAcα2→3Galβ1→4Glcβ1→1Cer (other gangliosides are abbreviated according to Svennerholm's list (44Holmgren J. Svennerholm L. Elwing H. Fredman P. Strannegard O. Proc. Natl. Acad. Sci. U. S. A. 1980; 77: 1947-1950Crossref PubMed Scopus (89) Google Scholar)). Cationic lipids are rare in animal cells and tissues (see “Discussion”). PLPS has the novel effect of inducing signal transduction in neuronal cells, particularly neuritogenesis in PC12 cells without addition of nerve growth factor; i.e. PLPS mimics the effect of nerve growth factor (7Sakakura C. Igarashi Y. Anand J.K. Sadozai K.K. Hakomori S. J. Biol. Chem. 1996; 271: 946-952Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). During previous studies, we noticed the presence of cationic lipids having slower TLC migration rate than PLPS in human brain (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar). We therefore undertook further systematic study of cationic lipids with plasmal conjugate in bovine brain.During isolation of cationic GSLs of bovine brain white matter by CM-Sephadex chromatography, we found a new compound different from previously known PLPS, characterized by much slower migration rate on TLC in various solvent systems. The new compound showed a chemical composition similar to that of PLPS but had an additional polar group, and is hereby termed PLPS x, or Px. In this report, we describe detailed structural analysis of this novel compound.DISCUSSIONThe present study indicates the presence of a significant quantity of a cationic GSL with plasmal conjugate, Px, having highly novel structure, in white matter of bovine brain. Px is characterized by properties similar to those of PLPS A and B, i.e. (i) Px and PLPS A and B are co-eluted with the same solvent (ionic strength 0.20–0.23 m triethylamine in methanol) and separated from psychosine; (ii) TLC behavior similar to that of PLPS in basic, neutral, and acidic solvents; (iii) both Px and PLPS yield psychosine and plasmal on weak acid hydrolysis. On the other hand, Px has certain properties distinct from those of PLPS: (i) Px has an additional polar group showing much slower TLC mobility than PLPS in all solvents tested. (ii) Preliminary negative-ion LSIMS of Px in comparison to PLPS A and B clearly indicated that Px has the higher molecular mass. Px had a molecular mass 92 Da higher in deprotonated molecules [M−H]− than that of PLPS; this difference in molecular mass corresponds to glycerol. (iii) Px has only one 6-O-substitution, whereas PLPS A has 3,4-di-O-substitution, and PLPS B has 4,6-di-O-substitution on methylation analysis. (iv) The sequence of the components and molecular species of aldehyde are characterized as a glycerol-O-aldehyde(16:0, 18:1, and 18:0)-O-Hex-Sph, by LSIMS and ESIMS of native and methylated compounds. (v) The structure was established unambiguously as a glycerol derivative of PLPS by 1H and 13C NMR.The novelty of the Px structure lies in the mode of plasmal conjugation, i.e. the way that O-plasmal conjugate is linked at C1 at two primary hydroxyl groups at glycerol and galactose. The glycerol residue may interact with galactosyl residue to achieve steric stability, such that axes of two aliphatic chains, sphingosine and plasmal, are oriented in parallel regardless of the C1 stereoisomer of plasmal. A tentative minimum energy conformational model is shown in Fig. 4. No similar structure has been observed previously in any plasmal conjugated compound. The most common plasmal conjugates are based on O-vinyl ether,i.e. O-alk-1-enyl group linked to one of the hydroxyls of glycerophosphoethanolamine or glycerophosphocholine, as observed typically in plasmalogen (for review see Refs. 27Hanahan D.J. Lipide Chemistry. J. Wiley & Sons, New York1960: 92-105Google Scholar, 28Debuch H. Seng P. Snyder F. Ether lipids: Chemistry and Biology. Academic Press, New York1972: 1-24Google Scholar). Another group of plasmal conjugates is based on cyclic acetal linked to a vicinal dihydroxyl group of galactosyl residue linked to Sph, ceramide, or diglyceride (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar, 4Levery S.B. Nudelman E.D. Hakomori S. Biochemistry. 1992; 31: 5335-5340Crossref PubMed Scopus (28) Google Scholar, 5Yachida Y. Kashiwaga M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1998; 39: 1039-1045Abstract Full Text Full Text PDF PubMed Google Scholar, 6Yachida Y. Kashiwagi M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1999; 40: 2271-2278Abstract Full Text Full Text PDF PubMed Google Scholar). We term this compound with novel structure “glyceroplasmalopsychosine.”The biosynthetic mechanism of either O-cyclic acetal or novel acetal, as now found in Px, is totally unknown at this time. In contrast, biosynthesis of O-alk-1-enyl structure as found in plasmalogen was well established through the 1-alkyl desaturase system, a microsomal mixed-function oxidase (29Snyder F. Lee T. Wykle R.L. Martonosi A.N. The Enzymes of Biological Membranes. Vol. 2, Biosynthesis and Metabolism. 2nd. Ed. Plenum Press, New York1985: 1-58Google Scholar). Release of free fatty aldehyde (plasmal) appears to be a prerequisite for synthesis ofO-acetal conjugates; however, the mechanism for this is still ambiguous. Three biochemical routes have been proposed: (i) action of acyl-CoA reductases in the presence of NADPH; (ii) oxidative cleavage of alkyl glyceride by alkyl mono-oxygenase (tetrahydropterine; Pte-H4-dependent); (iii) specific hydrolase for plasmalogen (“plasmalogenase”) capable of hydrolyzing O-alk-1-enyl group (30Snyder F. Vance D.E. Vance J.E. Biochemistry of Lipids, Lipoproteins and Membranes. Elsevier, Amsterdam1996: 183-210Google Scholar). However, released fatty aldehyde is transient and immediately converted to fatty alcohol or fatty acid, and it is perhaps utilized for O-acetal conjugation. The entire process may take place in the microsomal membrane.Although the mechanism of O-plasmal conjugate formation is unknown, it may be an enzymatic reaction coupled with either plasmal-forming reaction i, ii, or iii as above. This concept is supported by the following observations: (i) Specificity of plasmal conjugate always occurs at β-galactosyl but not other sugar residues (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar, 4Levery S.B. Nudelman E.D. Hakomori S. Biochemistry. 1992; 31: 5335-5340Crossref PubMed Scopus (28) Google Scholar, 5Yachida Y. Kashiwaga M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1998; 39: 1039-1045Abstract Full Text Full Text PDF PubMed Google Scholar, 6Yachida Y. Kashiwagi M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1999; 40: 2271-2278Abstract Full Text Full Text PDF PubMed Google Scholar). Such specificity cannot be associated with non-enzymatic reaction. (ii) Plasmal conjugate as plasmalocerebroside was detectable by MALDI-TOF of sphingolipid extract from as little as 200 mg of wet weight of Eker rat brain white matter but not from gray matter extract of the same brain. The level of plasmalocerebroside was many times higher in brain from Eker rats with hereditary renal carcinoma (31Taketomi T. Hara A. Uemura K. Sugiyama E. Acta Biochim. Pol. 1998; 45: 987-999Crossref PubMed Scopus (11) Google Scholar). This finding indicates that biosynthesis of acetal conjugates is enhanced in pathological processes and rules out the possibility that they are formed during the isolation procedure. (iii) The yield of Px from bovine brain white matter is consistent regardless of amount of material used or whether old (long-term frozen) versus fresh samples are used, i.e. 0.55–0.60 μmol per kg. (iv) The yield of PLPS from human brain is ∼10× higher than that from bovine brain, as found in both our previous and present studies. This indicates that a consistent level of plasmal conjugate is synthesized in brain and that the level varies depending on species. (v) No Px or PLPS was yielded when a mixture of plasmal and psychosine was incubated with or without glycerol for various durations, and the yield of PLPS was the same as that directly from the homogenate when human brain white matter homogenate was incubated with plasmal and psychosine or with phosphatidylethanolamine containing plasmalogen.NMR data indicate that two stereoisomers with regard to the asymmetric C1 carbon of plasmal in Px are detected in a ratio of ∼1:1, whereas those of PLPS are found exclusively in only one form (“endo type”). This may reflect a difference in stability of plasmal linkage in these two compounds, i.e. the linkage in Px is much more unstable than that in PLPS (see under “Experimental Procedures”: “Determination of Lipids Containing Aldehyde Using Schiff's Reagent”). Anomeric conversion of sugars as detected by mutarotation depends highly on stability of glycosidic linkage; e.g. N-glycosides are in general much more unstable thanO-glycosides, if the same sugar is linked to a similar aglycon. Susceptibility of mutarotation is higher in unstableN-linked structures than in O-linked structures. It is possible that the ratio of Px stereoisomers A and B is different in nascent product compared with those after isolation, although this is difficult to demonstrate at this time. In stereoisomers A and B, general conformational structure is very similar in terms of axes of two aliphatic chains and location of sugars.In general, lipids can be classified as acidic, neutral, zwitterionic, or cationic based on their ionic properties. Positive ionic properties of cationic lipids (which include psychosine and other lyso-GSLs, Sph, and dimethyl-Sph) are ascribable to the amino group of Sph. Cationic lipids having free Sph amino group, though a minor component, modulate activity of growth factor receptor kinase, protein kinase C, or other membrane-bound signal transducer molecules located at upstream regions of signal transduction pathways. This event leads to up- or down-regulation of key molecules, located downstream, which control transcription. For example, Sph and dimethyl-Sph inhibit PKC (32Igarashi Y. Hakomori S. Toyokuni T. Dean B. Fujita S. Sugimoto M. Ogawa T. El-Ghendy K. Racker E. Biochemistry. 1989; 28: 6796-6800Crossref PubMed Scopus (239) Google Scholar, 33Merrill A.H.J. Nimkar S. Menaldino D. Hannun Y.A. Loomis C.R. Bell R.M. Tyagi S.R. Lambeth J.D. Stevens V.L. Hunter R. Liotta D.C. Biochemistry. 1989; 28: 3138-3145Crossref PubMed Scopus (228) Google Scholar), activate epidermal growth factor receptor kinase (34Igarashi Y. Kitamura K. Toyokuni T. Dean B. Fenderson B.A. Ogawa T. Hakomori S. J. Biol. Chem. 1990; 265: 5385-5389Abstract Full Text PDF PubMed Google Scholar) and many other protein kinases that modulate chaperone effect, e.g.Sph-dependent kinase 1 for 14-3-3 protein (35Megidish T. Cooper J. Zhang L. Fu H. Hakomori S. J. Biol. Chem. 1998; 273: 21834-21845Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 36Megidish T. Takio K. Titani K. Iwabuchi K. Hamaguchi A. Igarashi Y. Hakomori S. Biochemistry. 1999; 38: 3369-3378Crossref PubMed Scopus (34) Google Scholar). Another example is activation by PLPS of Trk A tyrosine kinase in PC12 cells, leading to enhancement of mitogen-activated protein kinase. This mimics the effect of nerve growth factor, because withdrawal of PLPS induces apoptosis (7Sakakura C. Igarashi Y. Anand J.K. Sadozai K.K. Hakomori S. J. Biol. Chem. 1996; 271: 946-952Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). PLPS weakly inhibits PKC, whereas psychosine strongly inhibits PKC (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar).Systematic chemical analysis of cationic lipids from lower phase of Folch's partition of bovine brain revealed the following interesting points: (i) Although it was reported to be absent from human brain (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar), psychosine was clearly detectable in bovine brain white matter (Fig. 1; Table I); (ii) PLPS A and B, originally found in human brain white matter, were also detected in bovine brain, where their levels were, respectively, 12 and 18 times lower than in human (Table I); (iii) a novel GSL, Px, was isolated and characterized as a major plasmal conjugate in bovine brain. Whether Px occurs in human brain remains to be determined.The large difference in PLPS level between human and bovine brain, and its association with neuritogenesis, raises the interesting question of whether this compound is involved in neuronal communication, which may be qualitatively much higher in human brain. Although the structure of Px is distinct from that of PLPS in that Px contains glycerol, they have in common the structure of psychosine with plasmal conjugate. An important topic for future study is how Px affects signal transduction in neuronal cells.Psychosine is neurotoxic and strongly inhibits brain PKC activity. Its PKC-inhibitory effect is much stronger than that of PLPS (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar). Although it was reported to be absent in normal human brain, psychosine is detectable in brains of patients with sphingolipidosis (e.g. Krabbe's disease) (37Hannun Y.A. Bell R.M. Science. 1989; 243: 500-507Crossref PubMed Scopus (1104) Google Scholar, 38Igarashi Y. Nojiri H. Hanai N. Hakomori S. Methods Enzymol. 1989; 179: 521-541Crossref PubMed Scopus (69) Google Scholar). An interesting question is whether conversion of psychosine to PLPS or to Px may result in entirely different neuronal signaling, because PLPS is neurotrophic whereas psychosine is neurotoxic. Cationic lipids are rare in animal cells and tissues (see “Discussion”). PLPS has the novel effect of inducing signal transduction in neuronal cells, particularly neuritogenesis in PC12 cells without addition of nerve growth factor; i.e. PLPS mimics the effect of nerve growth factor (7Sakakura C. Igarashi Y. Anand J.K. Sadozai K.K. Hakomori S. J. Biol. Chem. 1996; 271: 946-952Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). During previous studies, we noticed the presence of cationic lipids having slower TLC migration rate than PLPS in human brain (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar). We therefore undertook further systematic study of cationic lipids with plasmal conjugate in bovine brain. During isolation of cationic GSLs of bovine brain white matter by CM-Sephadex chromatography, we found a new compound different from previously known PLPS, characterized by much slower migration rate on TLC in various solvent systems. The new compound showed a chemical composition similar to that of PLPS but had an additional polar group, and is hereby termed PLPS x, or Px. In this report, we describe detailed structural analysis of this novel compound. DISCUSSIONThe present study indicates the presence of a significant quantity of a cationic GSL with plasmal conjugate, Px, having highly novel structure, in white matter of bovine brain. Px is characterized by properties similar to those of PLPS A and B, i.e. (i) Px and PLPS A and B are co-eluted with the same solvent (ionic strength 0.20–0.23 m triethylamine in methanol) and separated from psychosine; (ii) TLC behavior similar to that of PLPS in basic, neutral, and acidic solvents; (iii) both Px and PLPS yield psychosine and plasmal on weak acid hydrolysis. On the other hand, Px has certain properties distinct from those of PLPS: (i) Px has an additional polar group showing much slower TLC mobility than PLPS in all solvents tested. (ii) Preliminary negative-ion LSIMS of Px in comparison to PLPS A and B clearly indicated that Px has the higher molecular mass. Px had a molecular mass 92 Da higher in deprotonated molecules [M−H]− than that of PLPS; this difference in molecular mass corresponds to glycerol. (iii) Px has only one 6-O-substitution, whereas PLPS A has 3,4-di-O-substitution, and PLPS B has 4,6-di-O-substitution on methylation analysis. (iv) The sequence of the components and molecular species of aldehyde are characterized as a glycerol-O-aldehyde(16:0, 18:1, and 18:0)-O-Hex-Sph, by LSIMS and ESIMS of native and methylated compounds. (v) The structure was established unambiguously as a glycerol derivative of PLPS by 1H and 13C NMR.The novelty of the Px structure lies in the mode of plasmal conjugation, i.e. the way that O-plasmal conjugate is linked at C1 at two primary hydroxyl groups at glycerol and galactose. The glycerol residue may interact with galactosyl residue to achieve steric stability, such that axes of two aliphatic chains, sphingosine and plasmal, are oriented in parallel regardless of the C1 stereoisomer of plasmal. A tentative minimum energy conformational model is shown in Fig. 4. No similar structure has been observed previously in any plasmal conjugated compound. The most common plasmal conjugates are based on O-vinyl ether,i.e. O-alk-1-enyl group linked to one of the hydroxyls of glycerophosphoethanolamine or glycerophosphocholine, as observed typically in plasmalogen (for review see Refs. 27Hanahan D.J. Lipide Chemistry. J. Wiley & Sons, New York1960: 92-105Google Scholar, 28Debuch H. Seng P. Snyder F. Ether lipids: Chemistry and Biology. Academic Press, New York1972: 1-24Google Scholar). Another group of plasmal conjugates is based on cyclic acetal linked to a vicinal dihydroxyl group of galactosyl residue linked to Sph, ceramide, or diglyceride (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar, 4Levery S.B. Nudelman E.D. Hakomori S. Biochemistry. 1992; 31: 5335-5340Crossref PubMed Scopus (28) Google Scholar, 5Yachida Y. Kashiwaga M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1998; 39: 1039-1045Abstract Full Text Full Text PDF PubMed Google Scholar, 6Yachida Y. Kashiwagi M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1999; 40: 2271-2278Abstract Full Text Full Text PDF PubMed Google Scholar). We term this compound with novel structure “glyceroplasmalopsychosine.”The biosynthetic mechanism of either O-cyclic acetal or novel acetal, as now found in Px, is totally unknown at this time. In contrast, biosynthesis of O-alk-1-enyl structure as found in plasmalogen was well established through the 1-alkyl desaturase system, a microsomal mixed-function oxidase (29Snyder F. Lee T. Wykle R.L. Martonosi A.N. The Enzymes of Biological Membranes. Vol. 2, Biosynthesis and Metabolism. 2nd. Ed. Plenum Press, New York1985: 1-58Google Scholar). Release of free fatty aldehyde (plasmal) appears to be a prerequisite for synthesis ofO-acetal conjugates; however, the mechanism for this is still ambiguous. Three biochemical routes have been proposed: (i) action of acyl-CoA reductases in the presence of NADPH; (ii) oxidative cleavage of alkyl glyceride by alkyl mono-oxygenase (tetrahydropterine; Pte-H4-dependent); (iii) specific hydrolase for plasmalogen (“plasmalogenase”) capable of hydrolyzing O-alk-1-enyl group (30Snyder F. Vance D.E. Vance J.E. Biochemistry of Lipids, Lipoproteins and Membranes. Elsevier, Amsterdam1996: 183-210Google Scholar). However, released fatty aldehyde is transient and immediately converted to fatty alcohol or fatty acid, and it is perhaps utilized for O-acetal conjugation. The entire process may take place in the microsomal membrane.Although the mechanism of O-plasmal conjugate formation is unknown, it may be an enzymatic reaction coupled with either plasmal-forming reaction i, ii, or iii as above. This concept is supported by the following observations: (i) Specificity of plasmal conjugate always occurs at β-galactosyl but not other sugar residues (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar, 4Levery S.B. Nudelman E.D. Hakomori S. Biochemistry. 1992; 31: 5335-5340Crossref PubMed Scopus (28) Google Scholar, 5Yachida Y. Kashiwaga M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1998; 39: 1039-1045Abstract Full Text Full Text PDF PubMed Google Scholar, 6Yachida Y. Kashiwagi M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1999; 40: 2271-2278Abstract Full Text Full Text PDF PubMed Google Scholar). Such specificity cannot be associated with non-enzymatic reaction. (ii) Plasmal conjugate as plasmalocerebroside was detectable by MALDI-TOF of sphingolipid extract from as little as 200 mg of wet weight of Eker rat brain white matter but not from gray matter extract of the same brain. The level of plasmalocerebroside was many times higher in brain from Eker rats with hereditary renal carcinoma (31Taketomi T. Hara A. Uemura K. Sugiyama E. Acta Biochim. Pol. 1998; 45: 987-999Crossref PubMed Scopus (11) Google Scholar). This finding indicates that biosynthesis of acetal conjugates is enhanced in pathological processes and rules out the possibility that they are formed during the isolation procedure. (iii) The yield of Px from bovine brain white matter is consistent regardless of amount of material used or whether old (long-term frozen) versus fresh samples are used, i.e. 0.55–0.60 μmol per kg. (iv) The yield of PLPS from human brain is ∼10× higher than that from bovine brain, as found in both our previous and present studies. This indicates that a consistent level of plasmal conjugate is synthesized in brain and that the level varies depending on species. (v) No Px or PLPS was yielded when a mixture of plasmal and psychosine was incubated with or without glycerol for various durations, and the yield of PLPS was the same as that directly from the homogenate when human brain white matter homogenate was incubated with plasmal and psychosine or with phosphatidylethanolamine containing plasmalogen.NMR data indicate that two stereoisomers with regard to the asymmetric C1 carbon of plasmal in Px are detected in a ratio of ∼1:1, whereas those of PLPS are found exclusively in only one form (“endo type”). This may reflect a difference in stability of plasmal linkage in these two compounds, i.e. the linkage in Px is much more unstable than that in PLPS (see under “Experimental Procedures”: “Determination of Lipids Containing Aldehyde Using Schiff's Reagent”). Anomeric conversion of sugars as detected by mutarotation depends highly on stability of glycosidic linkage; e.g. N-glycosides are in general much more unstable thanO-glycosides, if the same sugar is linked to a similar aglycon. Susceptibility of mutarotation is higher in unstableN-linked structures than in O-linked structures. It is possible that the ratio of Px stereoisomers A and B is different in nascent product compared with those after isolation, although this is difficult to demonstrate at this time. In stereoisomers A and B, general conformational structure is very similar in terms of axes of two aliphatic chains and location of sugars.In general, lipids can be classified as acidic, neutral, zwitterionic, or cationic based on their ionic properties. Positive ionic properties of cationic lipids (which include psychosine and other lyso-GSLs, Sph, and dimethyl-Sph) are ascribable to the amino group of Sph. Cationic lipids having free Sph amino group, though a minor component, modulate activity of growth factor receptor kinase, protein kinase C, or other membrane-bound signal transducer molecules located at upstream regions of signal transduction pathways. This event leads to up- or down-regulation of key molecules, located downstream, which control transcription. For example, Sph and dimethyl-Sph inhibit PKC (32Igarashi Y. Hakomori S. Toyokuni T. Dean B. Fujita S. Sugimoto M. Ogawa T. El-Ghendy K. Racker E. Biochemistry. 1989; 28: 6796-6800Crossref PubMed Scopus (239) Google Scholar, 33Merrill A.H.J. Nimkar S. Menaldino D. Hannun Y.A. Loomis C.R. Bell R.M. Tyagi S.R. Lambeth J.D. Stevens V.L. Hunter R. Liotta D.C. Biochemistry. 1989; 28: 3138-3145Crossref PubMed Scopus (228) Google Scholar), activate epidermal growth factor receptor kinase (34Igarashi Y. Kitamura K. Toyokuni T. Dean B. Fenderson B.A. Ogawa T. Hakomori S. J. Biol. Chem. 1990; 265: 5385-5389Abstract Full Text PDF PubMed Google Scholar) and many other protein kinases that modulate chaperone effect, e.g.Sph-dependent kinase 1 for 14-3-3 protein (35Megidish T. Cooper J. Zhang L. Fu H. Hakomori S. J. Biol. Chem. 1998; 273: 21834-21845Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 36Megidish T. Takio K. Titani K. Iwabuchi K. Hamaguchi A. Igarashi Y. Hakomori S. Biochemistry. 1999; 38: 3369-3378Crossref PubMed Scopus (34) Google Scholar). Another example is activation by PLPS of Trk A tyrosine kinase in PC12 cells, leading to enhancement of mitogen-activated protein kinase. This mimics the effect of nerve growth factor, because withdrawal of PLPS induces apoptosis (7Sakakura C. Igarashi Y. Anand J.K. Sadozai K.K. Hakomori S. J. Biol. Chem. 1996; 271: 946-952Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). PLPS weakly inhibits PKC, whereas psychosine strongly inhibits PKC (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar).Systematic chemical analysis of cationic lipids from lower phase of Folch's partition of bovine brain revealed the following interesting points: (i) Although it was reported to be absent from human brain (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar), psychosine was clearly detectable in bovine brain white matter (Fig. 1; Table I); (ii) PLPS A and B, originally found in human brain white matter, were also detected in bovine brain, where their levels were, respectively, 12 and 18 times lower than in human (Table I); (iii) a novel GSL, Px, was isolated and characterized as a major plasmal conjugate in bovine brain. Whether Px occurs in human brain remains to be determined.The large difference in PLPS level between human and bovine brain, and its association with neuritogenesis, raises the interesting question of whether this compound is involved in neuronal communication, which may be qualitatively much higher in human brain. Although the structure of Px is distinct from that of PLPS in that Px contains glycerol, they have in common the structure of psychosine with plasmal conjugate. An important topic for future study is how Px affects signal transduction in neuronal cells.Psychosine is neurotoxic and strongly inhibits brain PKC activity. Its PKC-inhibitory effect is much stronger than that of PLPS (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar). Although it was reported to be absent in normal human brain, psychosine is detectable in brains of patients with sphingolipidosis (e.g. Krabbe's disease) (37Hannun Y.A. Bell R.M. Science. 1989; 243: 500-507Crossref PubMed Scopus (1104) Google Scholar, 38Igarashi Y. Nojiri H. Hanai N. Hakomori S. Methods Enzymol. 1989; 179: 521-541Crossref PubMed Scopus (69) Google Scholar). An interesting question is whether conversion of psychosine to PLPS or to Px may result in entirely different neuronal signaling, because PLPS is neurotrophic whereas psychosine is neurotoxic. The present study indicates the presence of a significant quantity of a cationic GSL with plasmal conjugate, Px, having highly novel structure, in white matter of bovine brain. Px is characterized by properties similar to those of PLPS A and B, i.e. (i) Px and PLPS A and B are co-eluted with the same solvent (ionic strength 0.20–0.23 m triethylamine in methanol) and separated from psychosine; (ii) TLC behavior similar to that of PLPS in basic, neutral, and acidic solvents; (iii) both Px and PLPS yield psychosine and plasmal on weak acid hydrolysis. On the other hand, Px has certain properties distinct from those of PLPS: (i) Px has an additional polar group showing much slower TLC mobility than PLPS in all solvents tested. (ii) Preliminary negative-ion LSIMS of Px in comparison to PLPS A and B clearly indicated that Px has the higher molecular mass. Px had a molecular mass 92 Da higher in deprotonated molecules [M−H]− than that of PLPS; this difference in molecular mass corresponds to glycerol. (iii) Px has only one 6-O-substitution, whereas PLPS A has 3,4-di-O-substitution, and PLPS B has 4,6-di-O-substitution on methylation analysis. (iv) The sequence of the components and molecular species of aldehyde are characterized as a glycerol-O-aldehyde(16:0, 18:1, and 18:0)-O-Hex-Sph, by LSIMS and ESIMS of native and methylated compounds. (v) The structure was established unambiguously as a glycerol derivative of PLPS by 1H and 13C NMR. The novelty of the Px structure lies in the mode of plasmal conjugation, i.e. the way that O-plasmal conjugate is linked at C1 at two primary hydroxyl groups at glycerol and galactose. The glycerol residue may interact with galactosyl residue to achieve steric stability, such that axes of two aliphatic chains, sphingosine and plasmal, are oriented in parallel regardless of the C1 stereoisomer of plasmal. A tentative minimum energy conformational model is shown in Fig. 4. No similar structure has been observed previously in any plasmal conjugated compound. The most common plasmal conjugates are based on O-vinyl ether,i.e. O-alk-1-enyl group linked to one of the hydroxyls of glycerophosphoethanolamine or glycerophosphocholine, as observed typically in plasmalogen (for review see Refs. 27Hanahan D.J. Lipide Chemistry. J. Wiley & Sons, New York1960: 92-105Google Scholar, 28Debuch H. Seng P. Snyder F. Ether lipids: Chemistry and Biology. Academic Press, New York1972: 1-24Google Scholar). Another group of plasmal conjugates is based on cyclic acetal linked to a vicinal dihydroxyl group of galactosyl residue linked to Sph, ceramide, or diglyceride (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar, 4Levery S.B. Nudelman E.D. Hakomori S. Biochemistry. 1992; 31: 5335-5340Crossref PubMed Scopus (28) Google Scholar, 5Yachida Y. Kashiwaga M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1998; 39: 1039-1045Abstract Full Text Full Text PDF PubMed Google Scholar, 6Yachida Y. Kashiwagi M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1999; 40: 2271-2278Abstract Full Text Full Text PDF PubMed Google Scholar). We term this compound with novel structure “glyceroplasmalopsychosine.” The biosynthetic mechanism of either O-cyclic acetal or novel acetal, as now found in Px, is totally unknown at this time. In contrast, biosynthesis of O-alk-1-enyl structure as found in plasmalogen was well established through the 1-alkyl desaturase system, a microsomal mixed-function oxidase (29Snyder F. Lee T. Wykle R.L. Martonosi A.N. The Enzymes of Biological Membranes. Vol. 2, Biosynthesis and Metabolism. 2nd. Ed. Plenum Press, New York1985: 1-58Google Scholar). Release of free fatty aldehyde (plasmal) appears to be a prerequisite for synthesis ofO-acetal conjugates; however, the mechanism for this is still ambiguous. Three biochemical routes have been proposed: (i) action of acyl-CoA reductases in the presence of NADPH; (ii) oxidative cleavage of alkyl glyceride by alkyl mono-oxygenase (tetrahydropterine; Pte-H4-dependent); (iii) specific hydrolase for plasmalogen (“plasmalogenase”) capable of hydrolyzing O-alk-1-enyl group (30Snyder F. Vance D.E. Vance J.E. Biochemistry of Lipids, Lipoproteins and Membranes. Elsevier, Amsterdam1996: 183-210Google Scholar). However, released fatty aldehyde is transient and immediately converted to fatty alcohol or fatty acid, and it is perhaps utilized for O-acetal conjugation. The entire process may take place in the microsomal membrane. Although the mechanism of O-plasmal conjugate formation is unknown, it may be an enzymatic reaction coupled with either plasmal-forming reaction i, ii, or iii as above. This concept is supported by the following observations: (i) Specificity of plasmal conjugate always occurs at β-galactosyl but not other sugar residues (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar, 4Levery S.B. Nudelman E.D. Hakomori S. Biochemistry. 1992; 31: 5335-5340Crossref PubMed Scopus (28) Google Scholar, 5Yachida Y. Kashiwaga M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1998; 39: 1039-1045Abstract Full Text Full Text PDF PubMed Google Scholar, 6Yachida Y. Kashiwagi M. Mikami T. Tsuchihashi K. Daino T. Akino T. Gasa S. J. Lipid Res. 1999; 40: 2271-2278Abstract Full Text Full Text PDF PubMed Google Scholar). Such specificity cannot be associated with non-enzymatic reaction. (ii) Plasmal conjugate as plasmalocerebroside was detectable by MALDI-TOF of sphingolipid extract from as little as 200 mg of wet weight of Eker rat brain white matter but not from gray matter extract of the same brain. The level of plasmalocerebroside was many times higher in brain from Eker rats with hereditary renal carcinoma (31Taketomi T. Hara A. Uemura K. Sugiyama E. Acta Biochim. Pol. 1998; 45: 987-999Crossref PubMed Scopus (11) Google Scholar). This finding indicates that biosynthesis of acetal conjugates is enhanced in pathological processes and rules out the possibility that they are formed during the isolation procedure. (iii) The yield of Px from bovine brain white matter is consistent regardless of amount of material used or whether old (long-term frozen) versus fresh samples are used, i.e. 0.55–0.60 μmol per kg. (iv) The yield of PLPS from human brain is ∼10× higher than that from bovine brain, as found in both our previous and present studies. This indicates that a consistent level of plasmal conjugate is synthesized in brain and that the level varies depending on species. (v) No Px or PLPS was yielded when a mixture of plasmal and psychosine was incubated with or without glycerol for various durations, and the yield of PLPS was the same as that directly from the homogenate when human brain white matter homogenate was incubated with plasmal and psychosine or with phosphatidylethanolamine containing plasmalogen. NMR data indicate that two stereoisomers with regard to the asymmetric C1 carbon of plasmal in Px are detected in a ratio of ∼1:1, whereas those of PLPS are found exclusively in only one form (“endo type”). This may reflect a difference in stability of plasmal linkage in these two compounds, i.e. the linkage in Px is much more unstable than that in PLPS (see under “Experimental Procedures”: “Determination of Lipids Containing Aldehyde Using Schiff's Reagent”). Anomeric conversion of sugars as detected by mutarotation depends highly on stability of glycosidic linkage; e.g. N-glycosides are in general much more unstable thanO-glycosides, if the same sugar is linked to a similar aglycon. Susceptibility of mutarotation is higher in unstableN-linked structures than in O-linked structures. It is possible that the ratio of Px stereoisomers A and B is different in nascent product compared with those after isolation, although this is difficult to demonstrate at this time. In stereoisomers A and B, general conformational structure is very similar in terms of axes of two aliphatic chains and location of sugars. In general, lipids can be classified as acidic, neutral, zwitterionic, or cationic based on their ionic properties. Positive ionic properties of cationic lipids (which include psychosine and other lyso-GSLs, Sph, and dimethyl-Sph) are ascribable to the amino group of Sph. Cationic lipids having free Sph amino group, though a minor component, modulate activity of growth factor receptor kinase, protein kinase C, or other membrane-bound signal transducer molecules located at upstream regions of signal transduction pathways. This event leads to up- or down-regulation of key molecules, located downstream, which control transcription. For example, Sph and dimethyl-Sph inhibit PKC (32Igarashi Y. Hakomori S. Toyokuni T. Dean B. Fujita S. Sugimoto M. Ogawa T. El-Ghendy K. Racker E. Biochemistry. 1989; 28: 6796-6800Crossref PubMed Scopus (239) Google Scholar, 33Merrill A.H.J. Nimkar S. Menaldino D. Hannun Y.A. Loomis C.R. Bell R.M. Tyagi S.R. Lambeth J.D. Stevens V.L. Hunter R. Liotta D.C. Biochemistry. 1989; 28: 3138-3145Crossref PubMed Scopus (228) Google Scholar), activate epidermal growth factor receptor kinase (34Igarashi Y. Kitamura K. Toyokuni T. Dean B. Fenderson B.A. Ogawa T. Hakomori S. J. Biol. Chem. 1990; 265: 5385-5389Abstract Full Text PDF PubMed Google Scholar) and many other protein kinases that modulate chaperone effect, e.g.Sph-dependent kinase 1 for 14-3-3 protein (35Megidish T. Cooper J. Zhang L. Fu H. Hakomori S. J. Biol. Chem. 1998; 273: 21834-21845Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 36Megidish T. Takio K. Titani K. Iwabuchi K. Hamaguchi A. Igarashi Y. Hakomori S. Biochemistry. 1999; 38: 3369-3378Crossref PubMed Scopus (34) Google Scholar). Another example is activation by PLPS of Trk A tyrosine kinase in PC12 cells, leading to enhancement of mitogen-activated protein kinase. This mimics the effect of nerve growth factor, because withdrawal of PLPS induces apoptosis (7Sakakura C. Igarashi Y. Anand J.K. Sadozai K.K. Hakomori S. J. Biol. Chem. 1996; 271: 946-952Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). PLPS weakly inhibits PKC, whereas psychosine strongly inhibits PKC (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar). Systematic chemical analysis of cationic lipids from lower phase of Folch's partition of bovine brain revealed the following interesting points: (i) Although it was reported to be absent from human brain (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar), psychosine was clearly detectable in bovine brain white matter (Fig. 1; Table I); (ii) PLPS A and B, originally found in human brain white matter, were also detected in bovine brain, where their levels were, respectively, 12 and 18 times lower than in human (Table I); (iii) a novel GSL, Px, was isolated and characterized as a major plasmal conjugate in bovine brain. Whether Px occurs in human brain remains to be determined. The large difference in PLPS level between human and bovine brain, and its association with neuritogenesis, raises the interesting question of whether this compound is involved in neuronal communication, which may be qualitatively much higher in human brain. Although the structure of Px is distinct from that of PLPS in that Px contains glycerol, they have in common the structure of psychosine with plasmal conjugate. An important topic for future study is how Px affects signal transduction in neuronal cells. Psychosine is neurotoxic and strongly inhibits brain PKC activity. Its PKC-inhibitory effect is much stronger than that of PLPS (3Nudelman E.D. Levery S.B. Igarashi Y. Hakomori S. J. Biol. Chem. 1992; 267: 11007-11016Abstract Full Text PDF PubMed Google Scholar). Although it was reported to be absent in normal human brain, psychosine is detectable in brains of patients with sphingolipidosis (e.g. Krabbe's disease) (37Hannun Y.A. Bell R.M. Science. 1989; 243: 500-507Crossref PubMed Scopus (1104) Google Scholar, 38Igarashi Y. Nojiri H. Hanai N. Hakomori S. Methods Enzymol. 1989; 179: 521-541Crossref PubMed Scopus (69) Google Scholar). An interesting question is whether conversion of psychosine to PLPS or to Px may result in entirely different neuronal signaling, because PLPS is neurotrophic whereas psychosine is neurotoxic. We thank Dr. Martin Sadilek (Department of Chemistry, University of Washington, Seattle, WA) for GC-MS analysis, Dr. Masayuki Kubota (ThermoQuest K.K., Tokyo, Japan) for ESIMS analysis, and Dr. Stephen Anderson for scientific editing and preparation of the manuscript.
Referência(s)