Acyl-CoA:Lysophospholipid Acyltransferases
2008; Elsevier BV; Volume: 284; Issue: 1 Linguagem: Inglês
10.1074/jbc.r800046200
ISSN1083-351X
Autores Tópico(s)Sphingolipid Metabolism and Signaling
ResumoCell membranes contain several classes of glycerophospholipids, which have numerous structural and functional roles in the cells. Polyunsaturated fatty acids, including arachidonic acid and eicosapentaenoic acid, are located at the sn-2 (but not sn-1)-position of glycerophospholipids in an asymmetrical manner. Using acyl-CoAs as donors, glycerophospholipids are formed by a de novo pathway (Kennedy pathway) and modified by a remodeling pathway (Lands' cycle) to generate membrane asymmetry and diversity. Both pathways were reported in the 1950s. Whereas enzymes involved in the Kennedy pathway have been well characterized, including enzymes in the 1-acylglycerol-3-phosphate O-acyltransferase family, little is known about enzymes involved in the Lands' cycle. Recently, several laboratories, including ours, isolated enzymes working in the remodeling pathway. These enzymes were discovered not only in the 1-acylglycerol-3-phosphate O-acyltransferase family but also in the membrane-bound O-acyltransferase family. In this review, we summarize recent studies on cloning and characterization of lysophospholipid acyltransferases that contribute to membrane asymmetry and diversity. Cell membranes contain several classes of glycerophospholipids, which have numerous structural and functional roles in the cells. Polyunsaturated fatty acids, including arachidonic acid and eicosapentaenoic acid, are located at the sn-2 (but not sn-1)-position of glycerophospholipids in an asymmetrical manner. Using acyl-CoAs as donors, glycerophospholipids are formed by a de novo pathway (Kennedy pathway) and modified by a remodeling pathway (Lands' cycle) to generate membrane asymmetry and diversity. Both pathways were reported in the 1950s. Whereas enzymes involved in the Kennedy pathway have been well characterized, including enzymes in the 1-acylglycerol-3-phosphate O-acyltransferase family, little is known about enzymes involved in the Lands' cycle. Recently, several laboratories, including ours, isolated enzymes working in the remodeling pathway. These enzymes were discovered not only in the 1-acylglycerol-3-phosphate O-acyltransferase family but also in the membrane-bound O-acyltransferase family. In this review, we summarize recent studies on cloning and characterization of lysophospholipid acyltransferases that contribute to membrane asymmetry and diversity. All organisms are composed of cells that are enclosed by a cell membrane, which contains phospholipids, cholesterol, and proteins. Lipids fulfill four general functions. (i) They serve as an efficient source of energy; (ii) they form cell membranes that contain the bipolar lipids of glycerophospholipids and sphingophospholipids; (iii) they participate in the regulation of particular proteins through post-translational lipid modification; and (iv) they serve as messengers during cellular signal transduction (1.van Meer G. Voelker D.R. Feigenson G.W. Nat. Rev. Mol. Cell Biol. 2008; 9: 112-124Crossref PubMed Scopus (4234) Google Scholar). Thus, glycerophospholipids are important not only as structural and functional components of cell membranes but also as precursors of various lipid mediators, such as PAF 2The abbreviations used are: PAF, platelet-activating factor; PA, phosphatidic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; CL, cardiolipin; PI, phosphatidylinositol; PS, phosphatidylserine; PLA2, phospholipase A2; LPLAT, lysophospholipid acyltransferase; LPA, lysophosphatidic acid; GPAT, glycerol-3-phosphate acyltransferase; LPAAT, LPA acyltransferase; DAG, diacylglycerol; TAG, triacylglycerol; AGPAT, 1-acylglycerol-3-phosphate O-acyltransferase; LCLAT, lyso-CL acetyltransferase; ER, endoplasmic reticulum; LPGAT, lyso-PG acetyltransferase; LPCAT, lyso-PC acyltransferase; TLR, Toll-like receptor; LPEAT, lyso-PE acyltransferase; LPSAT, lyso-PS acyltransferase; MBOAT, membrane-bound O-acyltransferase; LPIAT, lyso-PI acyltransferase. and eicosanoids (2.Ishii S. Shimizu T. Prog. Lipid Res. 2000; 39: 41-82Crossref PubMed Scopus (326) Google Scholar, 3.Shimizu T. Ohto T. Kita Y. IUBMB Life. 2006; 58: 328-333Crossref PubMed Scopus (69) Google Scholar). Each tissue maintains a distinct content and composition of various phospholipids, such as PA, PC, PE, PG, CL, PI, and PS (1.van Meer G. Voelker D.R. Feigenson G.W. Nat. Rev. Mol. Cell Biol. 2008; 9: 112-124Crossref PubMed Scopus (4234) Google Scholar, 4.Yamashita A. Sugiura T. Waku K. J. Biochem. (Tokyo). 1997; 122: 1-16Crossref PubMed Scopus (233) Google Scholar, 5.Schlame M. Rua D. Greenberg M.L. Prog. Lipid Res. 2000; 39: 257-288Crossref PubMed Scopus (651) Google Scholar). For the biosynthesis of glycerophospholipids, fatty acids first need to be activated to acyl-CoAs as described by Kornberg and Pricer (6.Kornberg A. Pricer Jr., W.E. J. Biol. Chem. 1953; 204: 329-343Abstract Full Text PDF PubMed Google Scholar). Using acyl-CoAs as donors, phospholipids are formed from glycerol 3-phosphate by the de novo pathway, originally described by Kennedy and Weiss in 1956 (7.Kennedy E.P. Weiss S.B. J. Biol. Chem. 1956; 222: 193-214Abstract Full Text PDF PubMed Google Scholar). However, the acyl groups of glycerophospholipids are highly diverse and are distributed in an asymmetric manner (4.Yamashita A. Sugiura T. Waku K. J. Biochem. (Tokyo). 1997; 122: 1-16Crossref PubMed Scopus (233) Google Scholar, 8.Lands W.E. Biochim. Biophys. Acta. 2000; 1483: 1-14Crossref PubMed Scopus (120) Google Scholar). Saturated and monounsaturated fatty acids are usually esterified at the sn-1-position, whereas polyunsaturated acyl groups are esterified at the sn-2-position. This diversity and asymmetry is not fully explained by the Kennedy pathway. Rapid turnover of the sn-2-acyl moiety of glycerophospholipids was originally described by Lands as the remodeling pathway (Lands’ cycle) (9.Lands W.E. J. Biol. Chem. 1958; 231: 883-888Abstract Full Text PDF PubMed Google Scholar) and is attributed to the concerted and coordinated actions of PLA2s and LPLATs (3.Shimizu T. Ohto T. Kita Y. IUBMB Life. 2006; 58: 328-333Crossref PubMed Scopus (69) Google Scholar, 8.Lands W.E. Biochim. Biophys. Acta. 2000; 1483: 1-14Crossref PubMed Scopus (120) Google Scholar, 10.Waku K. Nakazawa Y. J. Biochem. (Tokyo). 1972; 72: 495-497Crossref PubMed Scopus (18) Google Scholar). Although these metabolic processes occur in a variety of tissues, information on the enzymes involved in phospholipid remodeling has been lacking for the past 50 years. Dr. Lands asked in his review, “Which enzymes distinguish between saturated and unsaturated acyl chains?” (8.Lands W.E. Biochim. Biophys. Acta. 2000; 1483: 1-14Crossref PubMed Scopus (120) Google Scholar). Now, we may be able to answer that question because several LPLATs have been recently cloned and characterized. In this review, after a brief description of the enzymes of the Kennedy pathway, we will summarize recent findings on the cloning and characterization of remodeling enzymes in the Lands’ cycle. In the de novo pathway of glycerophospholipid biosynthesis, LPA is first formed from glycerol 3-phosphate by GPAT (11.Van den Bosch H. Vance D.E. Biochim. Biophys. Acta. 1997; 1348: 1-2Crossref Google Scholar, 12.Coleman R.A. Lee D.P. Prog. Lipid Res. 2004; 43: 134-176Crossref PubMed Scopus (691) Google Scholar). Next, LPA is converted to PA by LPAATs, and PA is metabolized into two types of glycerol derivatives (11.Van den Bosch H. Vance D.E. Biochim. Biophys. Acta. 1997; 1348: 1-2Crossref Google Scholar, 12.Coleman R.A. Lee D.P. Prog. Lipid Res. 2004; 43: 134-176Crossref PubMed Scopus (691) Google Scholar). One is DAG, which is then converted to TAG, PC, and PE. Subsequently, PS is synthesized from PC or PE. The other glycerol derivative is cytidine diphospho-DAG, which is transformed into PI, PS, PG, and CL (Fig. 1). Several key enzymes in the de novo pathways have been characterized, and additional information is available in other review articles (1.van Meer G. Voelker D.R. Feigenson G.W. Nat. Rev. Mol. Cell Biol. 2008; 9: 112-124Crossref PubMed Scopus (4234) Google Scholar, 11.Van den Bosch H. Vance D.E. Biochim. Biophys. Acta. 1997; 1348: 1-2Crossref Google Scholar, 12.Coleman R.A. Lee D.P. Prog. Lipid Res. 2004; 43: 134-176Crossref PubMed Scopus (691) Google Scholar). Several acyltransferases that form LPA or PA have been identified, and all of them are members of the AGPAT family, which possesses LPLAT motifs (13.Lewin T.M. Wang P. Coleman R.A. Biochemistry. 1999; 38: 5764-5771Crossref PubMed Scopus (224) Google Scholar, 14.Yamashita A. Nakanishi H. Suzuki H. Kamata R. Tanaka K. Waku K. Sugiura T. Biochim. Biophys. Acta. 2007; 1771: 1202-1215Crossref PubMed Scopus (67) Google Scholar). Because several groups independently cloned LPLATs, they have multiple names. For example, AGPAT1 is also called LPAATα, and AGPAT8 is also known as AGPAT9, LPAATθ, or GPAT3. Confusingly, LCLAT1 was also given the name AGPAT8. To eliminate the confusion about the nomenclature, we propose that the enzymes be renamed based on their substrate specificities and by the order of their cloning publications (Table 1).TABLE 1Summary of LPLATs: a proposal for the standardization of LPLAT nomenclature Open table in a new tab GPATs—Four mammalian GPATs have been cloned (15.Yet S.F. Lee S. Hahm Y.T. Sul H.S. Biochemistry. 1993; 32: 9486-9491Crossref PubMed Scopus (76) Google Scholar, 16.Harada N. Hara S. Yoshida M. Zenitani T. Mawatari K. Nakano M. Takahashi A. Hosaka T. Yoshimoto K. Nakaya Y. Mol. Cell. Biochem. 2007; 297: 41-51Crossref PubMed Scopus (29) Google Scholar, 17.Cao J. Li J.L. Li D. Tobin J.F. Gimeno R.E. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 19695-19700Crossref PubMed Scopus (159) Google Scholar, 18.Chen Y.Q. Kuo M.S. Li S. Bui H.H. Peake D.A. Sanders P.E. Thibodeaux S.J. Chu S. Qian Y.W. Zhao Y. Bredt D.S. Moller D.E. Konrad R.J. Beigneux A.P. Young S.G. Cao G. J. Biol. Chem. 2008; 283: 10048-10057Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar, 19.Nagle C.A. Vergnes L. Dejong H. Wang S. Lewin T.M. Reue K. Coleman R.A. J. Lipid Res. 2008; 49: 823-831Abstract Full Text Full Text PDF PubMed Scopus (94) Google Scholar). GPAT1 and GPAT2 (also called xGPAT (16.Harada N. Hara S. Yoshida M. Zenitani T. Mawatari K. Nakano M. Takahashi A. Hosaka T. Yoshimoto K. Nakaya Y. Mol. Cell. Biochem. 2007; 297: 41-51Crossref PubMed Scopus (29) Google Scholar)) are located in the outer mitochondrial membrane, whereas GPAT3 (called AGPAT8, AGPAT9, or LPAATθ) and GPAT4 (called AGPAT6 or LPAATζ) are localized to the ER. GPAT1 is resistant to sulfhydryl agents like N-ethylmaleimide and prefers 16:0-CoA as a substrate (12.Coleman R.A. Lee D.P. Prog. Lipid Res. 2004; 43: 134-176Crossref PubMed Scopus (691) Google Scholar). In the liver of GPAT1 knock-out mice, the palmitate (16:0) content was lower at the sn-1-position of TAG, PC, and PE, indicating the important role of the mitochondrial form of GPAT in TAG and glycerophospholipid formation (12.Coleman R.A. Lee D.P. Prog. Lipid Res. 2004; 43: 134-176Crossref PubMed Scopus (691) Google Scholar). GPAT2 is N-ethylmaleimide-sensitive, has no preference for 16:0-CoA, and is expressed mainly in mouse testis (20.Wang S. Lee D.P. Gong N. Schwerbrock N.M. Mashek D.G. Gonzalez-Baro M.R. Stapleton C. Li L.O. Lewin T.M. Coleman R.A. Arch. Biochem. Biophys. 2007; 465: 347-358Crossref PubMed Scopus (67) Google Scholar). The microsomal form of GPAT constitutes ∼90% of the total GPAT activity in most tissues but only 50–80% of the activity in the liver (12.Coleman R.A. Lee D.P. Prog. Lipid Res. 2004; 43: 134-176Crossref PubMed Scopus (691) Google Scholar). Moreover, in differentiating 3T3-L1 adipocytes, the specific activity of microsomal GPAT is 70-fold higher, whereas mitochondrial GPAT activity is only 10-fold higher (12.Coleman R.A. Lee D.P. Prog. Lipid Res. 2004; 43: 134-176Crossref PubMed Scopus (691) Google Scholar). The mRNA level of GPAT3 is consistently 60-fold higher in 3T3-L1 adipocytes than in preadipocytes (17.Cao J. Li J.L. Li D. Tobin J.F. Gimeno R.E. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 19695-19700Crossref PubMed Scopus (159) Google Scholar). On the other hand, GPAT4 is expressed in many tissues. Both GPAT3 and GPAT4 recognize a broad range of substrates from 12:0-CoA to 18:1- or 18:2-CoA as donors. The microsomal form of GPAT is thought to play vital roles in TAG synthesis. The mitochondrial form of GPAT is regulated nutritionally and hormonally (12.Coleman R.A. Lee D.P. Prog. Lipid Res. 2004; 43: 134-176Crossref PubMed Scopus (691) Google Scholar). LPAATs—GPATs catalyze the formation of LPA from glycerol 3-phosphate, and LPAATs subsequently catalyze the formation of PA from LPA in the de novo pathway. To date, two LPAATs (LPAAT1 and LPAAT2) have been cloned and characterized (21.West J. Tompkins C.K. Balantac N. Nudelman E. Meengs B. White T. Bursten S. Coleman J. Kumar A. Singer J.W. Leung D.W. DNA Cell Biol. 1997; 16: 691-701Crossref PubMed Scopus (111) Google Scholar, 22.Stamps A.C. Elmore M.A. Hill M.E. Kelly K. Makda A.A. Finnen M.J. Biochem. J. 1997; 326: 455-461Crossref PubMed Scopus (42) Google Scholar, 23.Kume K. Shimizu T. Biochem. Biophys. Res. Commun. 1997; 237: 663-666Crossref PubMed Scopus (49) Google Scholar, 24.Eberhardt C. Gray P.W. Tjoelker L.W. J. Biol. Chem. 1997; 272: 20299-20305Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar), and three additional LPAAT candidates (AGPAT3–5) have been reported but have not been analyzed in any detail (25.Lu B. Jiang Y.J. Zhou Y. Xu F.Y. Hatch G.M. Choy P.C. Biochem. J. 2005; 385: 469-477Crossref PubMed Scopus (91) Google Scholar). They are all members of the AGPAT family and have four LPLAT motifs. Human LPAAT1 (called AGPAT1 or LPAATα) and human LPAAT2 (called AGPAT2 or LPAATβ) were cloned based on their homologies to yeast, Escherichia coli, and coconut AGPATs (21.West J. Tompkins C.K. Balantac N. Nudelman E. Meengs B. White T. Bursten S. Coleman J. Kumar A. Singer J.W. Leung D.W. DNA Cell Biol. 1997; 16: 691-701Crossref PubMed Scopus (111) Google Scholar, 22.Stamps A.C. Elmore M.A. Hill M.E. Kelly K. Makda A.A. Finnen M.J. Biochem. J. 1997; 326: 455-461Crossref PubMed Scopus (42) Google Scholar, 24.Eberhardt C. Gray P.W. Tjoelker L.W. J. Biol. Chem. 1997; 272: 20299-20305Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). Human LPAAT1 is expressed ubiquitously (21.West J. Tompkins C.K. Balantac N. Nudelman E. Meengs B. White T. Bursten S. Coleman J. Kumar A. Singer J.W. Leung D.W. DNA Cell Biol. 1997; 16: 691-701Crossref PubMed Scopus (111) Google Scholar, 22.Stamps A.C. Elmore M.A. Hill M.E. Kelly K. Makda A.A. Finnen M.J. Biochem. J. 1997; 326: 455-461Crossref PubMed Scopus (42) Google Scholar). LPAAT1 showed higher activity with 14:0-, 16:0-, and 18:2-CoAs and showed intermediate activities with 18:1- and 20:4-CoAs (26.Hollenback D. Bonham L. Law L. Rossnagle E. Romero L. Carew H. Tompkins C.K. Leung D.W. Singer J.W. White T. J. Lipid Res. 2006; 47: 593-604Abstract Full Text Full Text PDF PubMed Scopus (48) Google Scholar). The LPLAT motifs of human LPAAT1 contain the sequences NHX4D (motif I, residues 103–109), GVIFIDR (motif II, residues 143–149), EGTR (motif III, residues 178–181), and IVPIVM (motif IV, residues 205–210) (14.Yamashita A. Nakanishi H. Suzuki H. Kamata R. Tanaka K. Waku K. Sugiura T. Biochim. Biophys. Acta. 2007; 1771: 1202-1215Crossref PubMed Scopus (67) Google Scholar). Site-directed mutagenesis of LPAAT1 demonstrated that these motifs are essential for LPAAT activity (14.Yamashita A. Nakanishi H. Suzuki H. Kamata R. Tanaka K. Waku K. Sugiura T. Biochim. Biophys. Acta. 2007; 1771: 1202-1215Crossref PubMed Scopus (67) Google Scholar). Human LPAAT2 mRNA is found in most tissues, with the highest expression seen in the heart, liver, and adipocytes (21.West J. Tompkins C.K. Balantac N. Nudelman E. Meengs B. White T. Bursten S. Coleman J. Kumar A. Singer J.W. Leung D.W. DNA Cell Biol. 1997; 16: 691-701Crossref PubMed Scopus (111) Google Scholar, 24.Eberhardt C. Gray P.W. Tjoelker L.W. J. Biol. Chem. 1997; 272: 20299-20305Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar, 27.Agarwal A.K. Arioglu E. De Almeida S. Akkoc N. Taylor S.I. Bowcock A.M. Barnes R.I. Garg A. Nat. Genet. 2002; 31: 21-23Crossref PubMed Scopus (409) Google Scholar). LPAAT2 prefers 20:4-CoA rather than 16:0- or 18:0-CoA (24.Eberhardt C. Gray P.W. Tjoelker L.W. J. Biol. Chem. 1997; 272: 20299-20305Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar, 26.Hollenback D. Bonham L. Law L. Rossnagle E. Romero L. Carew H. Tompkins C.K. Leung D.W. Singer J.W. White T. J. Lipid Res. 2006; 47: 593-604Abstract Full Text Full Text PDF PubMed Scopus (48) Google Scholar). Mutations in LPAAT2 have been linked to congenital generalized lipodystrophy (also known as Berardinelli-Seip syndrome) (27.Agarwal A.K. Arioglu E. De Almeida S. Akkoc N. Taylor S.I. Bowcock A.M. Barnes R.I. Garg A. Nat. Genet. 2002; 31: 21-23Crossref PubMed Scopus (409) Google Scholar), indicating that LPAAT2 is involved in TAG synthesis and storage in adipocytes. Putative LPLATs in the AGPAT Family—Three other putative LPAATs (AGPAT3–5; also called LPAATγ, LPAATδ, and LPAATε, respectively) have been reported, but their LPAAT activity with 18:1-CoA was very low (25.Lu B. Jiang Y.J. Zhou Y. Xu F.Y. Hatch G.M. Choy P.C. Biochem. J. 2005; 385: 469-477Crossref PubMed Scopus (91) Google Scholar). Additionally, a putative mouse acyltransferase gene, AT-like 1B, is registered in the NCBI Database; however, the biochemical characteristics were not provided, and a human homolog has not been found. Phospholipids are first synthesized in the de novo pathway, and their fatty acyl composition at the sn-2-position is altered in the remodeling pathway (Lands' cycle) through the concerted actions of PLA2s and LPLATs (9.Lands W.E. J. Biol. Chem. 1958; 231: 883-888Abstract Full Text PDF PubMed Google Scholar). Although several PLA2s have been identified and well characterized (3.Shimizu T. Ohto T. Kita Y. IUBMB Life. 2006; 58: 328-333Crossref PubMed Scopus (69) Google Scholar), the cloning and characterization of acyltransferases are only occurring now. LPLATs were recently shown to be remodeling enzymes, and their biochemical analyses are in progress. LCLAT1 and LPGAT1—CL is the only known dimeric glycerophospholipid and consists of four fatty acyl chains, a linoleoyl group (C18:2). CL is required for the reconstituted activity of several key mitochondrial enzymes involved in energy metabolism (5.Schlame M. Rua D. Greenberg M.L. Prog. Lipid Res. 2000; 39: 257-288Crossref PubMed Scopus (651) Google Scholar). CL is synthesized from dilyso-CL and monolyso-CL by LCLAT. Mouse LCLAT1 (called ALCAT1 by Cao et al.) was identified and found to possess LPLAT motifs (28.Cao J. Liu Y. Lockwood J. Burn P. Shi Y. J. Biol. Chem. 2004; 279: 31727-31734Abstract Full Text Full Text PDF PubMed Scopus (167) Google Scholar). Overexpression of LCLAT1 in Sf9 cells or COS-7 cells led to a significant increase in di-LCLAT and mono-LCLAT activities. The enzyme recognized both dilyso-CL and monolyso-CL as acceptors with a preference for 18:1- and 18:2-CoAs. Mouse LCLAT1 is widely distributed, with the highest expression seen in the heart and liver. When LCLAT1 was overexpressed in COS-7 cells, the enzyme localized to the ER. The subcellular localization was predicted by the existence of a KKXX motif, an ER retention signal at the C-terminal end (29.Shikano S. Li M. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 5783-5788Crossref PubMed Scopus (88) Google Scholar) of LCLAT1. Although another AGPAT8 (not GPAT3) was reported as being an LCLAT, its biochemical characteristics were not analyzed in detail (30.Agarwal A.K. Barnes R.I. Garg A. Arch. Biochem. Biophys. 2006; 449: 64-76Crossref PubMed Scopus (51) Google Scholar). CL remodeling is believed to play an important role in the maintenance of normal heart functions. Defective CL is associated with Barth syndrome, a family disease caused by mutations of putative acyltransferase genes that manifests itself as cardiomyopathy and skeletal myopathy (31.Schlame M. Kelley R.I. Feigenbaum A. Towbin J.A. Heerdt P.M. Schieble T. Wanders R.J. DiMauro S. Blanck T.J. J. Am. Coll. Cardiol. 2003; 42: 1994-1999Crossref PubMed Scopus (155) Google Scholar). Thus, studies of LCLAT are needed to clarify the regulatory role of CL remodeling in cardiac function. PG is a precursor for the synthesis of CL and a potential activator of the protein kinase C family, including nuclear protein kinase CβII (32.Murray N.R. Fields A.P. J. Biol. Chem. 1998; 273: 11514-11520Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). The same group who cloned LCLAT1 found that LPGAT1 catalyzed the synthesis of PG from lyso-PG and acyl-CoA (33.Yang Y. Cao J. Shi Y. J. Biol. Chem. 2004; 279: 55866-55874Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). The enzyme also possesses the LPLAT motifs. Human LPGAT1 overexpressed in COS-7 cells showed a clear preference for 16:0-, 18:0-, and 18:1-CoAs as donors, which is consistent with the composition of endogenous PG in several tissues. Human LPGAT1 is widely distributed in tissues. The enzyme has the KKXX motif and is localized to the ER (33.Yang Y. Cao J. Shi Y. J. Biol. Chem. 2004; 279: 55866-55874Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). These two studies are a pioneering achievement in initiating studies on acyltransferases at the molecular level. LPCAT1 and LPCAT2—We (34.Nakanishi H. Shindou H. Hishikawa D. Harayama T. Ogasawara R. Suwabe A. Taguchi R. Shimizu T. J. Biol. Chem. 2006; 281: 20140-20147Abstract Full Text Full Text PDF PubMed Scopus (183) Google Scholar) and Chen et al. (35.Chen X. Hyatt B.A. Mucenski M.L. Mason R.J. Shannon J.M. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 11724-11729Crossref PubMed Scopus (149) Google Scholar) independently discovered LPCAT1 (also called AGPAT9 or AT-like 2), which has LPLAT motifs and catalyzes dipalmitoyl-PC synthesis. The enzyme is expressed mainly in the lung, especially in alveolar type II cells, and its mRNA is increased during the perinatal period. Dipalmitoyl-PC is a major component of pulmonary surfactant, which decreases surface tension, thereby preventing alveolar collapse, small airway closure, and alveolar flooding. Pulmonary surfactant deficiency is recognized to be an important contributing factor in the pathogenesis of infant respiratory distress syndrome, acute respiratory distress syndrome, asthma, and bronchiolitis (36.Stevens T.P. Sinkin R.A. Chest. 2007; 131: 1577-1582Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar). Thus, LPCAT1 may synthesize the PC of pulmonary surfactant and play a critical role in respiratory physiology. It is of note that LPCAT1 utilizes 18:2- or 18:3-acyl-CoA and exhibits LPGAT activity, reflecting the exact lipid composition of pulmonary surfactant. Further studies are needed to elucidate the exact roles of LPCAT1 in vivo and to determine whether there is a direct relationship between LPCAT1 and surfactant lipid synthesis. LPCAT2 (called LysoPAFAT/LPCAT2 and AT-like 1) was identified by our group as the long-sought lyso-PAF acetyltransferase involved in PAF biosynthesis in the remodeling pathway (37.Shindou H. Hishikawa D. Nakanishi H. Harayama T. Ishii S. Taguchi R. Shimizu T. J. Biol. Chem. 2007; 282: 6532-6539Abstract Full Text Full Text PDF PubMed Scopus (181) Google Scholar). LPCAT2 also has the LPLAT, EF hand-like, and KKXX motifs, which is consistent with its subcellular localization to the ER. The highest level of LPCAT2 expression was observed in resident macrophages and casein-induced neutrophils, followed by skin, colon, spleen, and thioglycolate-induced macrophages. Moreover, in mouse thioglycolate-induced macrophages, the mRNA level of LPCAT2 was increased by treatment with TLR agonists, lipopolysaccharide (a TLR4 agonist) and ODN1826 (a TLR9 agonist), both of which are bacterial cell components. The induction was suppressed by dexamethasone treatment. Because the virus component poly(I:C) (a TLR3 agonist) did not induce LPCAT2 expression, LPCAT2 is induced by bacterial infection but not by viral infection. Surprisingly, LPCAT2 also possesses LPCAT activity using 20:4-CoA as the best donor. The reaction product is 1-O-alkyl-2-arachidonoyl-PC, which is a major membrane constituent of inflammatory cells as well as a precursor of lyso-PAF. Lyso-PAF is produced from 1-O-alkyl-2-arachidonoyl-PC by PLA2s and used to form PAF (Fig. 1) (3.Shimizu T. Ohto T. Kita Y. IUBMB Life. 2006; 58: 328-333Crossref PubMed Scopus (69) Google Scholar, 38.Shindou H. Ishii S. Uozumi N. Shimizu T. Biochem. Biophys. Res. Commun. 2000; 271: 812-817Crossref PubMed Scopus (58) Google Scholar). Thus, LPCAT2 is a critically important enzyme not only in the biosynthesis of PAF (lyso-PAF acetyltransferase activity) but also in membrane homeostasis of inflammatory cells (LPCAT activity). Upon acute inflammatory stimulation with lipopolysaccharide, the activated enzyme utilizes acetyl-CoA more efficiently and produces PAF. It is speculated that LPCAT2 is phosphorylated and activated by p38 mitogen-activated protein kinase (MAPK) dependently because an inhibitor of this kinase abolished the activation of LPCAT2 and endogenous lyso-PAF acetyltransferase (37.Shindou H. Hishikawa D. Nakanishi H. Harayama T. Ishii S. Taguchi R. Shimizu T. J. Biol. Chem. 2007; 282: 6532-6539Abstract Full Text Full Text PDF PubMed Scopus (181) Google Scholar, 39.Shindou H. Ishii S. Yamamoto M. Takeda K. Akira S. Shimizu T. J. Immunol. 2005; 175: 1177-1183Crossref PubMed Scopus (39) Google Scholar). Therefore, specific inhibitors of LPCAT2 may be better anti-inflammatory drugs than PAF receptor antagonists because they also inhibit proliferation of inflammatory cells by disturbing membrane biogenesis. It will be important to characterize both acetyltransferase and acyltransferase activities of LPCAT2, including identification of binding sites for each substrate (acetyl-CoA and arachidonoyl-CoA). Additionally, it will be necessary to study the differential regulation of each enzyme activity. A recent study demonstrated that LPCAT1 catalyzes not only dipalmitoyl-PC synthesis (LPCAT) but also PAF synthesis (lyso-PAF acetyltransferase) (Fig. 1) (40.Harayama T. Shindou H. Ogasawara R. Suwabe A. Shimizu T. J. Biol. Chem. 2008; 283: 11097-11106Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar). In mouse macrophages, LPCAT1 was neither activated nor up-regulated by lipopolysaccharide stimulation, which was in contrast to LPCAT2. Moreover, both LPCAT1 and LPCAT2 have EF hand-like motifs; however, only LPCAT2 showed Ca2+-dependent activity. Site-directed mutagenesis of LPCAT1 demonstrated that the LPLAT motif 2 is an acyl-CoA-binding site in LPCAT1. Thus, two distinct lyso-PAF acetyltransferases are present: LPCAT1, a constitutively expressed enzyme, and LPCAT2, an inducible enzyme. This relationship is similar to cyclooxygenases 1 and 2, which are constitutively expressed and inducible enzymes, respectively (41.Smith W.L. Langenbach R. J. Clin. Investig. 2001; 107: 1491-1495Crossref PubMed Scopus (526) Google Scholar). This finding indicates that there are two kinds of PAF remodeling pathways: the inflammatory/inducible (LPCAT2) and non-inflammatory/constitutive (LPCAT1) remodeling pathways (40.Harayama T. Shindou H. Ogasawara R. Suwabe A. Shimizu T. J. Biol. Chem. 2008; 283: 11097-11106Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar). Recently, AGPAT7, an enzyme that is similar to LPCAT1 and LPCAT2, was characterized and renamed as LPEAT2 (also called AGPAT7, LPAATη, and AT-like 3) (42.Cao J. Shan D. Revett T. Li D. Wu L. Liu W. Tobin J.F. Gimeno R.E. J. Biol. Chem. 2008; 283: 19049-19057Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). LPEAT2 possesses LPEAT, LPGAT, LPSAT, and LPCAT activities with 18:1- or 20:4-CoA; however, its small interfering RNA decreased only LPEAT activity in HEK293T cells. Because the enzyme is expressed mainly in the brain, it was suggested that LPEAT2 is an important enzyme for the biogenesis of brain PE. In addition, it has been reported that AGPAT7 (LPEAT2) has LPCAT activity in red blood cells, although its activity is relatively weak (43.Soupene E. Fyrst H. Kuypers F.A. Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 88-93Crossref PubMed Scopus (78) Google Scholar). MBOAT Family—MBOAT family members are putative acyltransferases (44.Hofmann K. Trends Biochem. Sci. 2000; 25: 111-112Abstract Full Text Full Text PDF PubMed Scopus (380) Google Scholar, 45.Shimizu T. Annu. Rev. Pharmacol. Toxicol. 2009; (in press)PubMed Google Scholar). A His residue positioned within a long hydrophobic region of these enzymes is invariant, likely making it one of the active-site residues. DGAT1 (DAG acyltransferase 1) and ACAT1 and ACAT2 (acyl-CoA:cholesterol acyltransferases 1 and 2) are members of the MBOAT family. Interestingly, the family also includes protein acyltransferases, such as Porcupine, Hedgehog acyltransferase, and GUP1 (46.Takada R. Satomi Y. Kurata T. Ueno N. Norioka S. Kondoh H. Takao T. Takada S. Dev. Cell. 2006; 11: 791-801Abstract Full Text Full Text PDF PubMed Scopus (553) Google Scholar, 47.Chen M.H. Li Y.J. Kawakami T. Xu S.M. Chuang P.T. Genes Dev. 2004; 18: 641-659Crossref PubMed Scopus (275) Google Scholar, 48.Bosson R. Jaquenoud M. Conzelmann A. Mol. Biol. Cell. 2006; 17: 2636-2645Crossref PubMed Scopus (113) Google Scholar). Porcupine is required for Ser209-dependent acylation (16:1) of Wnt-3a protein (already palmitoylated at Cys77) for secretion (46.Takada R. Satomi Y. Kurata T. Ueno N. Norioka S. Kondoh H. Takao T. Takada S. Dev. Cell. 2006; 11: 791-801Abstract Full Text Full Text PDF PubMed Scopus (553) Google Scholar). Hedgehog acyltransferase palmitates a Cys residue in Sonic Hedgehog, and this modification is essential for its activity as well as for the generation of a protein gradient in the developing embryo (47.Chen M.H. Li Y.J. Kawakami T. Xu S.M. Chuang P.T. Genes Dev. 2004; 18: 641-659Crossref PubMed Scopus (275) Google Scholar). In yeast, the GUP1 gene is proposed to catalyze the remodeling of the glycosylphosphatidylinositol anchor (48.Bosson R. Jaquenoud M. Conzelmann A. Mol. Biol. Cell. 2006; 17: 2636-2645Crossref PubMed Scopus (113) Google Scholar). Most members of the MBOAT family have not been characterized, such as MBOAT3–7 and LRC4 (leukocyte receptor cluster member 4). In the latter half of 2007, five groups independently reported an MBOAT homolog in yeast that showed several LPLAT activities (49.Riekhof W.R. Wu J. Jones J.L. Voelker D.R. J. Biol. Chem. 2007; 282: 28344-28352Abstract Full Text Full Text PDF PubMed Scopus (122) Google Scholar, 50.Benghezal M. Roubaty C. Veepuri V. Knudsen J. Conzelmann A. J. Biol. Chem. 2007; 282: 30845-30855Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar, 51.Jain S. Stanford N. Bhagwat N. Seiler B. Costanzo M. Boone C. Oelkers P. J. Biol. Chem. 2007; 282: 30562-30569Abstract Full Text Full Text PDF PubMed Scopus (84) Google Scholar, 52.Tamaki H. Shimada A. Ito Y. Ohya M. Takase J. Miyashita M. Miyagawa H. Nozaki H. Nakayama R. Kumagai H. J. Biol. Chem. 2007; 282: 34288-34298Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar, 53.Chen Q. Kazachkov M. Zheng Z. Zou J. FEBS Lett. 2007; 581: 5511-5516Crossref PubMed Scopus (52) Google Scholar). Soon thereafter, mammalian MBOAT members were reported in 2008 (45.Shimizu T. Annu. Rev. Pharmacol. Toxicol. 2009; (in press)PubMed Google Scholar, 54.Hishikawa D. Shindou H. Kobayashi S. Nakanishi H. Taguchi R. Shimizu T. Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 2830-2835Crossref PubMed Scopus (198) Google Scholar, 55.Zhao Y. Chen Y.Q. Bonacci T.M. Bredt D.S. Li S. Bensch W.R. Moller D.E. Kowala M. Konrad R.J. Cao G. J. Biol. Chem. 2008; 283: 8258-8265Abstract Full Text Full Text PDF PubMed Scopus (122) Google Scholar, 56.Lee H.C. Inoue T. Imae R. Kono N. Shirae S. Matsuda S. Gengyo-Ando K. Mitani S. Arai H. Mol. Biol. Cell. 2008; 19: 1174-1184Crossref PubMed Scopus (102) Google Scholar). We (54.Hishikawa D. Shindou H. Kobayashi S. Nakanishi H. Taguchi R. Shimizu T. Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 2830-2835Crossref PubMed Scopus (198) Google Scholar) and Zhao et al. (55.Zhao Y. Chen Y.Q. Bonacci T.M. Bredt D.S. Li S. Bensch W.R. Moller D.E. Kowala M. Konrad R.J. Cao G. J. Biol. Chem. 2008; 283: 8258-8265Abstract Full Text Full Text PDF PubMed Scopus (122) Google Scholar) independently discovered LPCAT3 (called MBOAT5), which has LPCAT, LPEAT, and LPSAT activities. Moreover, we showed that LPCAT4 (originally called MBOAT2) possesses LPCAT and LPEAT activities, whereas LPEAT1 (called MBOAT1) exhibits LPEAT and LPSAT activities (54.Hishikawa D. Shindou H. Kobayashi S. Nakanishi H. Taguchi R. Shimizu T. Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 2830-2835Crossref PubMed Scopus (198) Google Scholar). Thus, these results demonstrate that the MBOAT family is a novel LPLAT family. Mouse LPCAT3 mRNA was detected ubiquitously, with the highest expression seen in the testis. On the other hand, mouse LPCAT4 mRNA was highly expressed in the epididymis, brain, testis, and ovary, and mouse LPEAT1 mRNA was highly expressed in the stomach, epididymis, and colon. LPCAT3 showed higher activities with polyunsaturated fatty acyl-CoAs 20:4-CoA and 18:2-CoA than with saturated fatty acyl-CoA, whereas LPCAT4 and LPEAT1 had a clear preference for 18:1-CoA (Table 1). LPCAT3 and LPCAT4 had higher LPCAT activity with 1-acyllyso-PC than with 1-O-alkyllyso-PC or 1-O-alkenyllyso-PC as an acceptor. No clear differences were observed between 1-acyl-LPEAT and 1-O-alkenyl-LPEAT activities of each enzyme. Based on the apparent Km and Vmax values, LPCAT3 had higher LPCAT activity than LPEAT and LPSAT activities. Similarly, LPCAT4 had higher LPCAT activity than LPEAT activity. LPEAT1 showed similar activities for lyso-PE and lyso-PS. Using 20:4-CoA as a donor, endogenous LPCAT, LPEAT, and LPSAT activities were found to be decreased in B16 melanoma cells transfected with LPCAT3 small interfering RNA. Thus, LPCAT3 appears to be a key enzyme exhibiting LPCAT, LPEAT, and LPSAT activities in B16 cells (54.Hishikawa D. Shindou H. Kobayashi S. Nakanishi H. Taguchi R. Shimizu T. Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 2830-2835Crossref PubMed Scopus (198) Google Scholar). LPCAT3, LPCAT4, and LPEAT1 were localized primarily to the ER when enzymes were overexpressed in Chinese hamster ovary cells, which is consistent with having the ER-localized motif KKXX in the C-terminal region. Overexpression of each enzyme induced the formation of unidentified organelles (karmella-like structures). Transmission electron microscopy revealed that the atypical multilayer membrane components were detected in Chinese hamster ovary cells transfected with LPCAT3 cDNA (54.Hishikawa D. Shindou H. Kobayashi S. Nakanishi H. Taguchi R. Shimizu T. Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 2830-2835Crossref PubMed Scopus (198) Google Scholar). The nature and origin of these structures remain to be clarified. LPIAT1 (also called MBOA-7, MBOAT7, and LRC4) was identified as the first LPIAT, which catalyzes the incorporation of arachidonic acid and eicosapentaenoic acid into lyso-PI (56.Lee H.C. Inoue T. Imae R. Kono N. Shirae S. Matsuda S. Gengyo-Ando K. Mitani S. Arai H. Mol. Biol. Cell. 2008; 19: 1174-1184Crossref PubMed Scopus (102) Google Scholar). Site-directed mutagenesis of human LPIAT1 demonstrated that a predicted active-site residue, His350, within a long hydrophobic region is important for LPIAT activity. An LPIAT1 mutant of Caenorhabditis elegans showed a “bags of worms” phenotype whereby the embryos hatched within the mother, leaving a cuticle sack that contained multiple wriggling larvae. Recently, MBOAT4 was identified as a ghrelin O-acyltransferase, which esterifies Ser3 of ghrelin, an appetite-stimulating peptide hormone, with an octanoyl group (57.Yang J. Brown M.S. Liang G. Grishin N.V. Goldstein J.L. Cell. 2008; 132: 387-396Abstract Full Text Full Text PDF PubMed Scopus (931) Google Scholar). Acylation is important for the growth hormone-releasing activity of ghrelin. Thus, the MBOAT family is composed of enzymes that incorporate fatty acids into amino acid residues and lysophospholipids. It is unknown which amino acid residues or motifs are important in distinguishing between lysophospholipids and proteins as acceptors. In the last 4 years, many LPLATs have been identified, resulting in the most spectacular advance in the LPLAT field since the discovery of the Kennedy pathway and the Lands' cycle 50 years ago. In this review, we propose to rename LPLATs to clarify and standardize the nomenclature (Table 1). The possibility that additional LPLATs, with preferences for different acyl-CoAs, contribute to membrane composition and diversity will be addressed in future studies. The existence of multiple LPLATs is reminiscent of the 20 distinct aminoacyl-tRNA synthetases and acyltransferases that help incorporate amino acids into pre-existing polypeptides (58.Ibba M. Soll D. Annu. Rev. Biochem. 2000; 69: 617-650Crossref PubMed Scopus (1068) Google Scholar). The redundant and pleiotropic substrate preferences of LPLATs may explain the diversity in membrane glycerophospholipids, which vary among tissues and can change in response to external stimuli. So far, a great deal of effort has been expended in trying to understand the biochemical characteristics of individual enzymes. Further studies will be needed to elucidate the biological roles of these enzymes in vivo. It will also be important to analyze enzyme activities with mixed substrates of acyl-CoAs and to determine the functional coupling of acyl-CoA synthetases and acyltransferases. Nevertheless, recent findings with LPLATs go a long way toward answering the questions posed by Dr. Lands alluded to earlier (8.Lands W.E. Biochim. Biophys. Acta. 2000; 1483: 1-14Crossref PubMed Scopus (120) Google Scholar) and may will open the door to greater understanding of the biological significance of membrane diversity and asymmetry.
Referência(s)