Báo cáo khoa học: Expression in yeast of a novel phospholipase A1 cDNA from Arabidopsis thaliana docx

Furthermore, a complete lipid analysis of the transformed wild-type yeast showed that its phospholipid content was lower than that of the control void plasmid-transformed yeast whereas l

Expression in yeast of a novel phospholipase A1 cDNA

Alexandre Noiriel1, Pierre Benveniste1, Antoni Banas2, Sten Stymne2and Pierrette Bouvier-Nave´1

1

Institut de Biologie Mole´culaire des Plantes du CNRS, De´partement Isopre´noı¨des, Institut de Botanique, Strasbourg, France;

2

Department of Crop Science, Swedish University of Agricultural Sciences, Alnarp, Sweden

During a search for cDNAs encoding plant sterol

acyl-transferases, we isolated four full-length cDNAs from

Ara-bidopsis thaliana that encode proteins with substantial

identity with animal lecithin : cholesterol acyltransferases

(LCATs) The expression of one of these cDNAs, AtLCAT3

(At3g03310), in various yeast strains resulted in the doubling

of the triacylglycerol content Furthermore, a complete lipid

analysis of the transformed wild-type yeast showed that its

phospholipid content was lower than that of the control

(void plasmid-transformed) yeast whereas

lysophospho-lipids and free fatty acids increased When microsomes from

the AtLCAT3-transformed yeast were incubated with

di-[1-14C]oleyl phosphatidylcholine, both the

lysophospho-lipid and free fatty acid fractions were highly and similarly

labelled, whereas the same incubation with microsomes from

the control yeast produced a negligible labelling of these

fractions Moreover when microsomes from

AtLCAT3-transformed yeast were incubated with either sn-1- or

sn-2-[1-14C]acyl phosphatidylcholine, the distribution of the

labelling between the free fatty acid and the lysophospha-tidylcholine fractions strongly suggested a phospholipase A1 activity for AtLCAT3 The sn-1 specificity of this phos-pholipase was confirmed by gas chromatography analysis of the hydrolysis of 1-myristoyl, 2-oleyl phosphatidylcholine Phosphatidylethanolamine and phosphatidic acid were shown to be also hydrolysed by AtLCAT3, although less efficiently than phosphatidylcholine Lysophospatidylcho-line was a weak substrate whereas tripalmitoylglycerol and cholesteryl oleate were not hydrolysed at all This novel

A thaliana phospholipase A1 shows optimal activity at

pH 6–6.5 and 60–65
C and appears to be unaffected by

Ca2+ Its sequence is unrelated to all other known phos-pholipases Further studies are in progress to elucidate its physiological role

Keywords: Arabidopsis thaliana; expression in yeast; phos-pholipase A1; triacylglycerol increase

Phospholipases A1 (PLA1) and A2 (PLA2) hydrolyse,

respectively, the sn-1 and sn-2 acylester bond of

phospho-lipids, generating free fatty acids (FAs) and

lysophospho-lipids Phospholipases B sequencially remove two FA from phospholipids and thus have both phospholipase A and lysophospholipase activities [1] These three types of phos-pholipase activities (A1, A2 and B) have been described in microsomal preparations from triacylglycerol (TAG)-accu-mulating tissues of various plants [2] A PLA1 activity has been identified in the tonoplast from Acer pseudoplatanus cells [3] and an Arabidopsis thaliana cDNA encoding a PLA1 was shown to be expressed in the chloroplast [4] But most of the plant PLA papers describe soluble PLA(2) activities [1] Participation of PLAs in plant signal transduction is mentioned for auxin stimulation of growth [5–7] and in response to bacterial and fungal elicitors [8–10], wounding [11] or viral infection [10,12] This involvement of plant PLAs in signal transduction has just been reviewed [13] PLAs are also directly implicated in phospholipid retailor-ing or degradation durretailor-ing TAG synthesis [2,14] or senes-cence [15] The participation of PLAs in these various aspects of plant development and response to stress is likely

to occur in coordination with phospholipases C and D [16] Several plant cDNAs encoding PLAs have been cloned and characterized They can be classified into three distinct groups according to their sequence The first group includes the small (12–14 kDa) secretory PLA2s [7,17] which contain

12 conserved Cys residues and conserved regions that are likely to represent the active site and Ca2+-binding loop found in animal secretory PLA2s [18]

Correspondence to P Bouvier-Nave´, Institut de Biologie Mole´culaire

des Plantes CNRS, De´partement Isopre´noı¨des, Institut de Botanique,

28 rue Goethe, 67083 Strasbourg Cedex, France.

Fax: +33 3 90 24 19 21, Tel.: +33 3 90 24 18 46,

E-mail: Pierrette.Nave@bota-ulp.u-strasbg.fr,

URL: http://ibmp.u-strasbg.fr/

Abbreviations: DGAT, diacylglycerol:acylCoA acyltransferase; FA,

fatty acid; FS, free sterol; FAME, fatty acid methyl ester; G3PAT,

glycerol-3-phosphate acyltransferase; LCAT, lecithin : cholesterol

acyltransferase; LPAAT, lysophosphatidic acid acyltransferase; LPC,

lysophosphatidylcholine; LPE, lysophosphatidylethanolamine; PC,

phosphatidylcholine; PDAT, phospholipid:diacylglycerol

acyltrans-ferase; PE, phosphatidylethanolamine; PLA1, phospholipase A1;

PS, phosphatidylserine; SE, steryl ester; TAG, triacylglycerol.

Enzymes: DGAT, diacylglycerol:acylCoA acyltransferase (EC

2.3.1.20); LCAT, lecithin:cholesterol acyltransferase (EC 2.3.1.43);

PDAT, phospholipid:diacylglycerol acyltransferase (EC 2.3.1.158);

PLA1, phospholipase A1 or phosphatidylcholine 1-acylhydrolase

(EC 3.1.1.32).

Note: Part of this study was presented at the 16th International Plant

Lipid Symposium, Budapest, Hungary, 1–4 June 2004 (abstract

book pp 33 and 105; http://www.mete.mtesz.hu/pls/).

(Received 16 April 2004, revised 27 July 2004, accepted 2 August 2004)

A second group of plant PLAs is formed by soluble,

patatin-like (phospho)lipases A(2): an allergen from the latex

of Hevea brasiliensis [19], three tobacco leaf soluble proteins

induced by virus infection [12], a cowpea galactolipid

acylhydrolase stimulated by drought stress [20] and four

A thalianaPLAs [6] Patatin is the major storage protein of

the potato tuber When cloned and expressed via a

baculo-virus vector in Sf9 insect cells, it was shown to be an aspecific

lipid acyl hydrolase that hydrolyses monoacylglycerols,

phosphatidylcholine (PC), monogalactosyldiacylglycerols,

di- and triacylglycerols with decreasing efficiency [21] When

assayed with PC, purified patatin exhibited a PLA2 activity

[22] Recent studies revealed that patatin has a Ser–Asp

catalytic dyad and a folding topology related to that of the

catalytic domain of animal cytosolic PLA2s Mutagenesis

confirmed the critical role of Ser77 and Asp215 in enzymatic

activity and of His109 in enzyme stability [23,24] Moreover

Rydel et al [24] described the crystal structure of patatin

Called either PLA(2)s or lipid acylhydrolases, patatin and

related proteins are 40–50-kDa proteins sharing 40–60%

identity (mostly in the N-terminal half) and possessing the

conserved Gly-X-Ser-X-Gly motif, around the catalytic Ser,

found in all the Ser hydrolases [25,26]

A third group of plant PLAs has been recently described

as lipase-like PLA1 Starting from an A thaliana mutant,

defective in anther dehiscence 1(dad1), the wild-type DAD1

gene was isolated, shown to complement the mutant and to

encode a chloroplastic protein of 45 kDa with PLA1

activity [4] Together with 11 homologous genes from the

Arabidopsis gene sequence databases, DAD1 presents

apparent similarities with some fungal lipases and

partic-ularly the characteristic
catalytic triad composed of a Ser,

an Asp and a His residue and the Gly-X-Ser-X-Gly

consensus motif around the catalytic Ser, both features that

are widely conserved in fungal and animal lipases and more

generally in Ser hydrolases [25,26] Apart from the

Gly-X-Ser-X-Gly motif common to groups 2 and 3, these three

groups of plant PLAs are unrelated

Here we wish to introduce a fourth group of plant PLAs,

the lecithin : cholesterol acyltransferase (LCAT)-like PLA1,

that we discovered in the course of a search for sterol

acyltransferases LCAT is the animal serum enzyme that

catalyses esterification of lipoprotein-associated cholesterol

by the sn-2 acyl group of PC In vitro studies showed that

LCAT, when incubated with PC in the absence of

cholesterol, also possess PLA2 activity [27] The cloning

and characterization of human LCAT [28] and its thorough

study by site-directed mutagenesis and molecular modelling

[29] showed that LCAT shares the Ser/Asp(Glu)/His

catalytic triad and the Gly-X-Ser-X-Gly motif with Ser

hydrolases We isolated four A thaliana cDNAs, the

deduced amino acid sequences of which share 25–35%

identity with human LCAT, including the catalytic triad

After expression in yeast, one of these cDNAs was clearly

shown to encode a PLA1

Experimental procedures

Chemicals

All of the lipids and Triton X-100 were 98–99% pure

products from Sigma [4-14C]Cholesterol (49 mCiÆmmol)1),

[1a,2a(n)-3H] cholesteryl oleate (29 CiÆmmol)1), [1-14C]oleic acid (50 mCiÆmmol)1), [1-14C]oleylCoA (55 mCiÆmmol)1), 1-palmitoyl-2-[1-14C]oleyl phosphatidylcholine (56 mCiÆ mmol)1), 1,2-[1-14C]oleyl phosphatidylcholine (107 mCiÆ mmol)1) and tri-[1-14C]palmitoylglycerol (55 mCiÆmmol)1) were from NEN or Amersham

1-[1-14C]Palmitoyl-2-oleyl phosphatidylcholine (2.2 mCiÆ mmol)1), 1-[1-14C]oleoyl-2-oleyl phosphatidylcholine (2.2 mCiÆmmol)1) and 1-oleoyl-2-[1-14C]oleyl phosphatidyl-choline (4.5 mCiÆmmol)1) were synthetized as described previously [2]

Strains, media and culture conditions Escherichia coli strain used, XL1 blue recA–[recA1, lac–, endA1, gyrA96, thi, hsdR17, SupE44, relA1 (F’proAB, lac1q, lacZ,DM15, Tn10)]

For Saccharomyces cerevisiae, two strains of com-mon genetic background (can1-100, his3-11,15, leu2-3,112, trp1-1, ura3-1): are1are2 (SCY059, MATa, ade2-1, met14D14HpaI-SalI, are1DNA::HIS3, are2D::LEU2) and the corresponding wild-type (SCY062, MATa) were a kind gift of S L Sturley (Columbia University College of Physicians and Surgeons, New York) Two strains were from Euroscarf, Frankfurt: dga1 (BY4742, MATa, his3D1, leu2D0, lys2D0, ura3D0, YOR245c::kanMX4) and lro1 (FY, Mat a, ura3-52, HIS3, leu2D1, LYS2, trp1D63, YNR008w(8, 1768)::kanMX2)

Yeast strains transformed with plasmid pYeDP60, harbouring either no insert or the plant cDNA under study, were simultaneously grown for 3 days at 30
C in minimum medium containing suitable supplements, then transferred into complete medium and grown overnight

at 30
C as previously described [30] The cells were then centrifuged and either freeze-dried for neutral lipid analysis or disrupted for complete lipid analysis or subfractionation

Plasmid for yeast transformation The plasmid pYeDP60 [31] was used to transform yeast strains as in [30]

Cloning of LCAT-like cDNAs For A thaliana LCAT1 (AtLCAT1), as an EST clone AV4422635 (Kazusa DNA Research Insitute, Chiba, Japan) became available, this latter was sequenced and shown to correspond to a cDNA encompassing the ORF

of a gene (At1g27480) called AtLCAT1 Its sequence was shown to encode a polypeptide of 432 amino acids This sequence has been assigned the GenBank accession number AY443040

For Medicago truncatula LCAT1 (MtLCAT1), the cDNA EST clone BE322181 (Samuel Roberts Noble Foundation Medicago truncatula insect herbivory library, USA) presented strong homology with AtLCAT1 After complete sequencing, it was shown to correspond to a cDNA of 1604 bp encoding a polypeptide of 450 amino acids with 57% identity with AtLCAT1

For Lycopersicum esculentum LCAT1 (LeLCAT1), the EST clone BG127829 (Clemson University Genomics

Institute, USA) was sequenced and the resulting polypeptide

was shown to have 55% identity with AtLCAT1

For Arabidopsis thaliana LCAT3 (AtLCAT3), the EST

cDNA clone BE525177 (ABRC, Ohio State University,

USA) was shown by sequencing and comparison with the

sequence of the gene At3g03310 to encode a truncated

AtLCAT3 polypeptide lacking 43 amino acids A complete

ORF was reconstituted by PCR using a direct primer 353

(152 nucleotides) bringing the lacking moiety of the cDNA

and a reverse primer 354 complementary to the 3¢ end of the

ORF (Table 1) and BE525177 as template Final

concen-trations of primers were 400 nM, template (20 ng), High

Fidelity PCR Master DNA polymerase (Boehringer;

25 lL), total volume 50 lL The PCR was performed using

29 cycles (30 s 94
, 30 s 50
, 2 min 72
) This resulted in the

amplification of a 1344 bp fragment which after digestion

with BamHI and KpnI was subcloned into pBlueScript SK

yielding the plasmid AtLCAT3-pSK After checking for the

absence of mutations, the insert was subcloned into the

yeast shuttle vector pYeDP60, yielding the plasmid

AtLCAT3-pYeDP60 AtLCAT3 was deposited in

GenBank and assigned the accession number AF421148

The FLAG-tagged A thaliana LCAT3

(FLAG-AtLCAT3) was made by PCR so that the C-terminal

FLAG epitope (*KDDDDKYD) was fused to the LCAT3

protein To this purpose the reverse primer 359 containing

the FLAG sequence was opposed to the direct primer 358

(Table 1) in the presence of AtLCAT3 (20 ng) as template

The PCR product was checked for the absence of mutations

and subcloned into pYeDP60 as shown above

To clone Nicotiana tabacum LCAT3 (NtLCAT3), an

orthologue of AtLCAT3 in tobacco, we took advantage of

the presence in databases of an mRNA sequence of tobacco

(clone q8487, accession number L31415) described as a

plant activating sequence encoding a positively charged

peptide functioning putatively as a transcriptional activation domain [32] This peptide showed more than 90% identity with a domain of AtLCAT3, therefore this peptide could belong to an orthologue of AtLCAT3 In order to isolate q8487, we designed two primers 362 and 363 (Table 1) corresponding to the 5¢- and 3¢-ends of q8487 and we opposed them in PCR assays containing a tobacco cDNA library (1 lL¼ 103 pfu), primers (400 nM) and High Fidelity PCR Master DNA polymerase (Boehringer;

25 lL) in a total volume of 50 lL The PCR was performed using 29 cycles (30 s 94
, 30 s 50
, 2 min 72
) This resulted

in the amplification of a 300-bp fragment The PCR product was subcloned into pGEMT plasmid to allow the sequence

to be checked

A cDNA library (250 000 pfu) from a 3-week-old

N tabacum cv Xanthi line LAB 1-4 calli derived from leaf protoplasts [33] was screened with the 300-bp q8487 sequence Thirteen positive spots were found and the corresponding kZAP phages were recovered, checked to give an amplification product in PCR experiments with primers 362 and 363, and screened a second time with q8487 allowing isolation of positive clones Two of the clones (211 and 314) were selected and sequenced, 314 was shown to be

a full-length cDNA of 1691 bp encoding a polypeptide of

451 amino acids sharing 68% identity with AtLCAT3 Therefore it was considered as an orthologue of AtLCAT3 and named NtLCAT3 (GenBank accession number AF468223)

For Mesembryanthemum crystallinum LCAT3 (McLCAT3)a search in TIGR databases allowed us to find several orthologues of AtLCAT3 in the ice plant (Mesembryanthemum crystallinum) These clones originated from an ice plant k Uni-Zap XR expression library prepared

48 h after NaCl treatment (J C Cushman, Department of Biochemistry, University of Nevada, Reno, NV) One of

Table 1 Synthetic oligonucleotide primer sequences (5¢ fi fi 3¢) used for gene cloning and site-directed mutagenesis Bold characters correspond to restriction sites Codons for the changed amino acids are underlined Nucleotides represented in bold characters indicate the point mutations produced For each mutation two oligonucleotides were synthesized: the one shown below and that with the complementary sequence.

Number Gene cloning

ATATATGGATCCATGTCTCTATTACTGG AAGAGATC 337

ATATATGGATCCATGGGCTGGATTCCGTGTCCGTGCTGGGGAACC 353 AACGACGATGAAAACGCCGGCGAGGTGGCGGATCGTGATCCGGTG

CTTCTAGTATCTGGAATTGGAGGCTCTATTCTGCATTCTAAGAAGA

AGAATTCAAAGTCTGAAATTCGGGTTTG

TATATAGGTACCTTACTTGTCATCGTCGTCCTTGTAGTCACCAGA 359 ATCAACTACTTTGTGAG

Site-directed mutagenesis

GCGTAGGAGTTTCGGGTAGCCTCCGCGGGCTTCTCCGTGATGAAAG H409L GGAGTGTCCTTCTATAACATATTTGGAGTGTCACTTAATACACC Y346F GTCACTATCATCTCCCATGCAATGGGAGGACTTATGGTTTC S177A

these clones (BE131533) was sequenced and shown to be a

cDNA of 1724 bp with a deletion of 60 bp inside the ORF

A complete cDNA of 1793 bp could be reconstituted owing

to two other EST cDNAs BE034988 and BE131478 whose

sequences encompassed and complemented BE131533 This

cDNA encoded a protein of 460 amino acids having 63%

identity with AtLCAT3

For Glycine max LCAT3 (GmLCAT3), Glycine max

orthologues of AtLCAT3 were found in TIGR databases

Superposition of these clones (G max gene index TC

reports: TC200937 and TC197463) allowed reconstitution

of a putative cDNA of 1551 bp encoding a polypeptide

having 63% identity with AtLCAT3

To get Arabidopsis thaliana LCAT4 (AtLCAT4), the

EST clone AV549462 (Kazusa DNA Research Insitute,

Chiba, Japan) was the starting point After sequencing it

was shown that AV549462 was a cDNA of 1802 bp

encoding a polypeptide of 536 amino acids In order to

determine the function of this cDNA, the ORF was

amplified by PCR using primers 337 and 338 (Table 1) at

a concentration of 400 nM, AV549462 (20 ng) as template,

and High Fidelity PCR Master DNA polymerase

(Boeh-ringer; 25 lL) The PCR was performed using 29 cycles

(30 s 94
, 30 s 50
, 2 min 72
) This resulted in the

amplification of a product of 1608 bp which was cloned

into pYeDP60 previously opened by BamHI and KpnI The

ORF was called AtLCAT4 and was registered in GenBank

under accession number AF421149 AtLCAT4 was derived

from At4g19860

For Lycopersicum esculentum LCAT4 (LeLCAT4), after

sequencing the EST clone BG125533 (Clemson University

Genomics Institute, USA), a cDNA of 1853 bp (GenBank

accession number AF465780) presenting strong homology

with AtLCAT4 was identified This cDNA encoded a

polypeptide of 535 amino acids having 66% identity with

AtLCAT4

For Medicago truncatula LCAT4 (MtLCAT4),

twenty-eight EST cDNA clones showing strong homology with

AtLCAT4 and coming from the same gene have been

reported in TIGR databases A nucleotide sequence

(TC86247) originating from the superimposition of these

clones has also been given After conceptual translation, a

polypeptide sequence of 539 amino acids showing 64%

identity with AtLCAT4 has been deduced

For Glycine max LCAT4 (GmLCAT4),

twenty-seven EST cDNA clones showing strong homology with

AtLCAT4 and coming from the same gene have been

reported in TIGR databases A nucleotide sequence

(TC192038) originating from the superimposition of these

clones has also been given A polypeptide sequence of

536 amino acids having 65% identity with AtLCAT4 and

79% identity with MtLCAT4 has been deduced after

conceptual translation

Site-directed mutagenesis onFLAG-tagged AtLCAT3

The mutated alleles of FLAG-AtLCAT3 were obtained by

introducing point mutations in the DNA sequence as

follows: two separate PCR reactions were performed with

20 ng of the pBluescript vector containing the ORF of

AtLCAT3 The first reaction was carried out with the direct

primer 358 (Table 1) and a reverse primer introducing the

chosen mutation The second PCR was performed with a sense primer, complementary to the antisense primer introducing the mutation, and primer 359 (Table 1) After phenol/chloroform extraction, precipitation of the amplified fragments with 3M NaCl and purification from agarose (Nucleospin Extract purification kit), the two fragments were hybridized due to the overlapping regions from the direct primer and the one introducing the mutation The hybridization was carried out in a final volume of 20 lL in the presence of PCR buffer (2 lL) (2 min at 100
C, 20 min

at 42
C, and 10 min at room temperature) Finally, a PCR

on 1 lL of the hybridization mix using primers 358 and 359 allowed the synthesis of the mutated ORF of FLAG-AtLCAT3 Amplifications of DNA fragments were per-formed using High Fidelity PCR Master DNA polymerase (Boehringer) in a final volume of 50 lL Amplification was

5 min at 92
C, followed by 29 cycles of 30 s at 95
C, 30 s

at 52
C, 2 min at 72
C, and then a 10 min elongation

at 72
C

Nucleotide sequence determination was performed as described previously [30]

Transformation of yeast Transformation was performed according to [30] with some modifications After the heat shock at 42
C the cells were centrifuged, resuspended in 2% (w/v) glucose (100 lL) and plated on minimum medium containing suitable supple-ments (50 lgÆmL)1each)

Lipid analysis Steryl esters (SEs), free sterols (FSs) and TAGs (for colorimetric quantification) were extracted from freeze-dried yeast cells and analysed as described previously [30] The complete lipid analysis of control and transformed yeast was performed as described previously [34] except that the chloroform extracts of the fresh cell pellets were shared for separate TLCs of neutral lipids in hexane/diethylether/acetic acid (70 : 30 : 1, v/v/v) and polar lipids in chloroform/ methanol/acetic acid/water (85 : 15 : 10 : 3.5, v/v/v/v) Subcellular fractionation

Yeasts were grown for 3 days in 100 mL glucose minimum medium followed by 16 h in 200 mL galactose complete medium (see above) The harvested cells were then disrupted

in 0.1MTris/HCl pH 7.5 containing 0.6Msorbitol, 1 mM EDTA and 0.5% (w/v) BSA, in the presence of glass beads (0.45–0.50 mm diameter) by vigorous hand shaking accord-ing to Pompon et al [35] The homogenate was centrifuged for 10 min at 12 000 g and the supernatant for 60 min at

100 000 g The microsomal pellet was resuspended (2–5 mg proteinÆmL)1) in 0.1M Tris/HCl pH 7 Microsomal sus-pensions and supernatant samples were kept at)80
C for several months without significant loss of activity

Proteins were quantified as before [30] Western blots of microsomes (50 lg protein) and supernatant (5 lL) from FLAG-AtLCAT3-transformed yeast were achieved after SDS/PAGE, electrotransfer to a nylon membrane (Immo-bilon-P; Millipore) and immunoblotting with the anti-FLAG M2 mAb from Sigma

Enzymatic assays

Sterol acyltransferase Sterol acyltransferase assays were

carried out with [4-14C]cholesterol according to [36]

Lecithin cholesterol acyltransferase LCAT assays

inclu-ded either [4-14C]cholesterol or di-[1-14C]oleylPC under

various conditions [36–38]

Phospholipid diacylglycerol acyltransferase (PDAT)

Assays were performed according to [34]

PLA1 assay PLA1 activity was first observed when

performing LCAT or PDAT assays The conditions were

then adjusted so that the phospholipase activity was

proportional to protein concentration and time and

opti-mized with respect to the substrate and detergent

concen-trations, while the reaction yield was kept below 20%: the

microsomal preparation (0.125 mg proteinÆmL)1) was

incu-bated with [1-14C]acyl-labelled PC (250 lM and usually

15 nCi) and Triton X-100 (0.15%) in 0.1MTris/HCl pH 7

(final volume, 100 lL) usually for 30 min at 30
C The

reaction was stopped by adding a mixture of methylene

chloride (400 lL) and methanol (100 lL) containing oleic

or palmitic acid, soybean PC and egg lysoPC (50 lg each)

as carriers After further addition of 0.03 N HCl (100 lL)

and methylene chloride (600 lL), the organic phase was

withdrawn and the water phase was extracted twice more

with methylene chloride The lipidic extract was

separ-ated by TLC in chloroform/methanol/water/acetic acid

(65 : 25 : 4 : 1; v/v/v/v) and the radioactive bands of free

FA (Rf¼ 0.9), PC (Rf¼ 0.4) and lysoPC (Rf¼ 0.2) were

detected with an automatic TLC-linear analyser (Berthold),

and scraped from the plate for liquid scintillation counting

A large-scale PLA1 assay was set up for GLC and

GLC/MS The incubation mixture (1 mL) had the same

composition as described above except that unlabelled

1-myristoyl, 2-oleylPC or various dioleylphospholipids

was used After purification, the free FA and lysoPC

fractions were converted to their fatty acid methyl ester

(FAME) derivatives Fatty acids were methylated by

10% boron trifluoride/methanol (Fluka), according to

Morrison and Smith [39] whereas LPC was

transmethy-lated by 0.5M sodium methoxide/methanol (Supelco)

Special care was taken to avoid loss of myristoyl

methylester as advised by Christie [40] The resulting

FAMEs were extracted in hexane then identified and

quantified by GLC on a DB-1 capillary column,

according to their relative retention time and peak area

to the internal standard heptadecanoic acid methylester

The GLC temperature program was from 60
C to

120
C, 20
CÆmin)1, from 120
C to 200
C, 2
CÆmin)1

and from 200
C to 280
C, 20
CÆmin)1 Identification

of FAMEs was confirmed by their mass spectra

AcylCoA synthase The AcylCoA synthase assay was

carried out according to [41]

Glycerol-3-phosphate acyltransferase The

glycerol-3-phosphate acyltransferase (G3PAT) assay was carried out

as described previously [42] but 0.1MTris/HCl, pH 7, was

used instead of 0.25MHEPES, pH 8

Lysophosphatidic acid acyltransferase The lysophospha-tidic acid acyltransferase (LPAAT) assay was derived from Bourgis et al [43] with the following modifications: the [1-14C]oleylCoA concentration was 100 lM instead of

10 lM, the lysophosphatidic acid concentration was 100 lM and not 55 lM, the Tris/HCl was pH 7 instead of 8 and phosphatidic acid (20 lM) was added

For all of assays mentioned above, the extraction and TLC procedures were as described for PLA1

HMGCoA reductase assay This activity was tested and analysed as described previously [44]

Results

Cloning of plant homologues of theHomo sapiens lecithin cholesterol acyltransferase

During a search for plant genes encoding enzymes involved

in sterol esterification by FA, we noticed A thaliana EST clones presenting homology with HsLCAT, the human gene encoding LCAT This enzyme catalyses the transfer of a FA from PC to cholesterol in the blood Our search led to the identification of four genes which were named AtLCAT1, AtLCAT2, AtLCAT3 and AtLCAT4 (At1g27480, At1g04010, At3g03310, At4g19860, respectively) In addi-tion two genes presenting homology with the recently discovered gene of phosphatidylcholine diacylglycerol acyl-transferase from Saccharomyces cerevisiae (ScPDAT) [34,45] were found and named AtPDAT1 and AtPDAT2 (At5g13640, At3g44830), the first of which has been recently cloned and shown to encode indeed a PDAT [46,47] These EST cDNA clones allowed the isolation and sequencing of cDNAs corresponding to AtLCAT1, 2, 3 and 4 All these cDNAs encompassed the coding region

A more extensive search of orthologues of the

A thaliana LCATgenes in various plants has been possible through use of TIGR (The Institute for Genomic Research, http://www.tigr.org) databases (Lactuca sativa LCAT1, Glycine max LCAT3 and LCAT4, Medicago truncatula LCAT2 and LCAT4) and thanks to cloning work performed in our laboratory (Medicago truncatula LCAT1, Nicotiana tabacum and Mesembryanthemum crystallinum LCAT3, Lycopersicum esculentum LCAT4)

A phylogenetic tree was constructed for several plant, animal, fungal and bacterial LCAT-like proteins which was rooted with the bacterial Bacillus licheniformis esterase as outgroup (Fig 1)

According to this tree the so-named plant LCATs are clearly divided into five subfamilies The LCAT1 subfamily

is the closest to mammalian and avian authentic LCATs It

is worthy of interest that very close to the mammalian LCATs, one can find proteins such as Bos taurus phospholipid ceramide acyltransferase (PLCAT) or Homo sapiensLCAT-like lysophospholipase (LLPL) which have both been shown recently to possess phospholipase A2 activity and to catalyse in vitro the transfer of a FA group from position 2 of a phospholipid to ceramide [48] The LCAT3 and LCAT4 form two clearly distinct subfamilies which are more distant from the mammalian LCATs than the LCAT1 subfamily, but which are closer to the

B licheniformis esterase The plant LCAT2 subfamily

appears also more distant from the mammalian and avian LCATs than the plant LCAT1 but close to the plant PDAT subfamily which groups with the yeast PDATs

The deduced protein sequences of all of these LCAT-like genes were aligned and several highly conserved regions were shown (Fig 2) They include the regions around the three essential amino acids of HsLCAT which form the catalytic triad common with all the Ser hydrolases

In order to determine the functions of the products of the AtLCATgenes their expression in yeast was undertaken The cDNA coding regions of AtLCAT1, 3 and 4 were inserted into the yeast expression vector pYeDP60 So far, the expression studies of AtLCAT1 and 4 in yeast has not allowed their function to be determined The functional expression of AtLCAT2 was performed independently in

A thaliana The encoded protein was shown to be a phospholipid sterol acyltransferase [48a,48b] We report hereafter the characterization of the AtLCAT3-encoded protein

Expression ofAtLCAT3 in various yeast strains Because AtLCAT3 was potentially encoding a sterol acyltransferase, it was first expressed in the yeast mutant are1are2 defective in sterol acyltransferase genes In this mutant, sterol ester biosynthesis is absent [49] and micro-somes from this strain lack sterol acyltransferase activity [50]

The neutral lipid content of the transformed yeast was compared to the control (void plasmid-transformed) yeast Surprisingly the expression of AtLCAT3 resulted in the doubling of the yeast TAG (Fig 3) and FS contents whereas the SE content remained unchanged (data not shown)

Incubation of di-[1-14C]oleyl PC with microsomes from AtLCAT3- or void plasmid-transformed are1are2, in the presence of cholesterol or dioleylglycerol, did not show any measurable acyltransferase activity but resulted in a high hydrolysis of PC into lysoPC (LPC) and FAs for micro-somes from the transformed yeast, whereas the control microsomes produced only a low hydrolysis of PC Considering the increase in TAG content of are1are2 when transformed with AtLCAT3, we wondered whether AtLCAT3 might be involved in plant TAG synthesis Because TAG synthesis in yeast is performed by several enzymes, mainly by diacylglycerol : acylCoA acyltrans-ferase (DGAT), partly by PDAT and a little by ARE1 and ARE2 [51–53], we transformed the corresponding mutants dga1 and lro1 and the wild-type strain with AtLCAT3.Whereas AtLCAT3-transformed wild-type and lro1strains as well as are1are2 had a doubled TAG content compared to that of the corresponding control strain, transformation of the dga1 mutant did not produce any change (Fig 3) These results clearly show that the yeast DGAT is involved in the observed TAG increase and consequently that AtLCAT3 is not directly implicated Next, microsomes from these three transformed (and control) strains were prepared and tested with di-[1-14C] oleyl PC: for each AtLCAT3-transformed strain including dga1, the PC acylhydrolase activity was much higher (10–100 times according to the assay conditions) than that

of the control microsomes

Fig 1 Phylogenetic tree showing LCAT-like proteins The phylogenic

tree was constructed for several plant, animal, fungal and bacterial

LCAT-like proteins BlESTER (B licheniformis esterase U35855);

AtLCAT4 (Arabidopsis thaliana AF421149, comes from At4g19860);

LeLCAT4 (Lycopersicum esculentum AF465780); GmLCAT4 (Glycine

max TC192038); MtLCAT4 (Medicago truncatula TC86247);

McLCAT3 (Mesembryanthemum crystallinum EST, BE131533);

GmLCAT3 (G max EST, TC200937); AtLCAT3 (A thaliana

phos-pholipase A1, AF421148, comes from At3g03310); NtLCAT3

(Nicotiana tabacum phospholipase A1, AF468223); CeLCAT

(Caenorhabditis elegans NP_492033); BtPLCAT (Bos taurus

phos-pholipid ceramide acyltransferase NP_776985); HsLLPL (Homo

sapiens LCAT-like lysophospholipase NP_036452); GgLCAT (Gallus

gallus lecithin cholesterol acyltransferase P53760); MmLCAT (Mus

musculus NP_032516); HsLCAT (H sapiens NP_000220); OcLCAT

(Oryctophagum communis P53761); AgLCAT (Anopheles gambiae

XP_319740); DmLCAT (Drosophila melanogaster AF145599);

AtLCAT1 (A thaliana AY443040, comes from At1g27480);

LsLCAT1 (Lactuca sativa EST BQ864610); MtLCAT1 (M truncatula

AF533771); AtLCAT2 (A thaliana NP_171897, comes from

At1g04010); MtLCAT2 (M truncatula AF493159); AtPDAT2

(A thaliana NP_190069, comes from At3g44830); AtPDAT1

(A thaliana AY160110, comes from At5g13640); MtPDAT1

(M truncatula AY210981); ScPDAT (Saccharomyces cerevisiae

phospholipid diacylglycerol acyltransferase P40345); SpPDAT

(Schizoaccharomyces pombe phospholipid diacylglycerol

acyltrans-ferase O94680) Accession numbers beginning by AF and AY

corres-pond to products which have been cloned and/or characterized in our

laboratory The phylogenetic tree has been rooted with BlESTER as the

outgroup Numbers at the nodes of the phylogenetic tree are bootstraps,

indicating the frequencies of occurrence of partitions found in the tree.

Expression ofAtLCAT3 in wild-type yeast: lipid analysis

For a complete study of the effects of AtLCAT3

transfor-mation on the yeast lipid content, the neutral and polar lipid

contents of AtLCAT3-transformed wild-type yeast were

compared to those of the control yeast by mean of GLC

analysis of the FAMEs generated from these fractions

(Fig 4) The PC, phosphatidylethanolamine (PE) and

phosphatidylserine (PS) contents of the

AtLCAT3-trans-formed yeast were found to be half those of the control yeast

while LPC, lysophosphatidylethanolamine (LPE) and free

FA were strongly increased The increase in TAG that we first measured by colorimetry (Fig 3) was clearly con-firmed, although to a lesser extent, by this GLC analysis Finally the total FA content was slightly (by 16%) but significantly increased in the AtLCAT3-transformed yeast and the amount of overproduced total FA (24 nmÆmg dry weight)1) corresponds roughly to the overproduced TAG (19.5 nmÆmg dry weight)1) (Fig 4) In the same analysis, steryl esters, diacylglycerol, phosphatidylinositol, cardio-lipin and lysophosphatidic acid were shown not to be changed significantly (data not shown)

Expression ofAtLCAT3 in wild-type yeast: enzyme characterization

As mentioned above, incubation of di-[1-14C]oleyl PC with microsomes from various AtLCAT3-transformed yeast strains yielded high labelling of the free FA and LPC fractions whereas the control microsomes produced a much weaker hydrolysis of PC After optimization of this PC acylhydrolase assay with microsomes from AtLCAT3-transformed and control wild-type yeast (see Experimental procedures), the yields of PC hydrolysis amounted to around 15% and 0.5%, respectively (Table 2), correspond-ing to PC acylhydrolase specific activities of around 600 and

20 nmolesÆmg protein)1Æh)1, respectively

When this di-[1-14C]oleyl PC was incubated with micro-somes from AtLCAT3-transformed yeast, the free FA and LPC fractions were labelled equally (Table 3) To study the positional specificity of AtLCAT3 towards the two acyl groups of PC, these microsomes were incubated with

sn-1-or sn-2-specifically labelled dioleylPC sn-1-or 1-palmitoyl, 2-oleyl

Fig 2 Alignment of the deduced aminoacid sequences of LCAT-like cDNAs Five highly conserved regions are shown The conserved amino acids are boxed The Ser177, Asp384 and His409 residues of AtLCAT3 corresponding to the catalytic triad of HsLCAT are indicated by a triangle, as well as two other residues (Tyr346 and Thr352) of AtLCAT3 which have been mutated.

Fig 3 TAG content of various control and AtLCAT3-transformed

yeast strains TAGs were extracted from freeze-dried cells (at least two

clones per strain), purified by TLC and quantified threefold by the

colorimetric assay described in the experimental section Deviation

from the mean was less than 12.5% White bars,

void-plasmid-trans-formed strains; black bars, AtLCAT3-transvoid-plasmid-trans-formed strains.

PC (Table 3) The distribution of radioactivity between the

free FA and the LPC fractions indicates for AtLCAT3 a

selectivity for the sn-1 position of about 90% with dioleylPC

and 85% with 1-palmitoyl, 2-oleyl PC The sn-1 specificity

of AtLCAT3 was also studied by GLC analysis of the FA

and LPC released during the incubation of 1-myristoyl,

2-oleyl PC (Fig 5): myristic acid accumulated almost

exclusively in the free FA fraction and oleic acid in the

LPC fraction Therefore with 1-myristoyl, 2-oleyl PC, the

selectivity of AtLCAT3 for the sn-1 position is almost 100%

Dioleyl phosphatidylethanolamine and dioleyl

phos-phatidic acid as well as 1-oleyl LPC were compared to

dioleyl PC as substrates GLC determination of the released

oleic acid showed that phosphatidylethanolamine,

phos-phatidic acid and LPC hydrolysis amounted to 50, 40 and

8%, respectively, that of PC (Fig 6) On the other hand,

incubations of tri-[1-14C]palmitoyl glycerol or [2,3-3H]

cholesteryl oleate did not produce any labelled palmitic

acid or cholesterol, respectively, excluding for AtLCAT3 a

TAG lipase or steryl ester hydrolase activity (Table 2)

Finally, using the optimized assay conditions for the

phospholipid acylhydrolase activity of AtLCAT3, we

looked again for a potential acyltransferase activity for this

protein Sterols and dioleylglycerol were added to the incubation mixture but this did not disclose any LCAT or PDAT activity As AtLCAT3 (a) does not show any acyltransferase activity, (b) does not show any acylhydrolase activity towards TAG or cholesteryl ester, and (c) hydro-lyses PC specifically at the sn-1 position and hydrohydro-lyses other phospholipids, it is most probably a PLA1

Its pH optimum was carefully determined with either 1,2-di-[1-14C]oleyl PC or 1-palmitoyl, 2-[1-14C]oleyl PC: the activity curve has a maximum at pH 6–6.5 Its optimal incubation temperature was surprisingly shown to be 60–65
C and its activity was unaffected by 0.1, 1 or

10 mMCa2+, under our assay conditions

The distribution of AtLCAT3 between the microsomal and 100 000 g supernatant subfractions from the trans-formed yeast was determined by comparing their total activities They were in a ratio of 84 : 16 indicating that most of this protein is associated with the microsomal

Table 2 Hydrolase activity towards various lipids of microsomes from control and AtLCAT3-transformed wild-type yeast Microsomal prep-arations from AtLCAT3-transformed and control wild-type yeast (0.125 mgÆmL)1) were incubated with various lipids (250 l M ) in the presence of 0.15% (v/v)Triton X-100 for 30 min Lipids were extracted and separated by TLC The radioactivity of the products was meas-ured by liquid scintillation Values are the mean of duplicates and experiments were repeated at least twice.

Substrates

Percentage of substrate hydrolysis with microsomes from

Void plasmid-transformed

WT yeast

AtLCAT3-transformed

WT yeast Tri-[1-14C]palmitoylglycerol a 0.0 0.0 [1a,2a(n)3H]Cholesteryl oleate a 0.0 0.0 [1–14C]Acyl PCb 0.5 15.0

a Tripamitoylglycerol and cholesteryl oleate were assayed in the presence of either 0.03 or 0.15% Triton X-100 b The values for labelled PC are a mean of the incubations presented in Table 3 and the corresponding controls.

Table 3 Positional specificity of the phospholipid acylhydrolase activity

of AtLCAT3: hydrolysis of specifically labelled PCs Microsomal preparations from AtLCAT3-transformed wild-type yeast (0.125 mgÆmL)1) were incubated with various PCs (250 l M ) in the presence of 0.15% (v/v) Triton X-100 for 15 min For further details see Table 3 legend.

Substrates

Relative labelling (%) in Free fatty

acid fraction

LPC fraction 1,2-Di[1- 14 C]oleyl PC 52 48 1-[1- 14 C]Oleyl,2-oleyl PC 91 9 1-Oleyl,2-[1-14C]oleyl PC 11 89 1-[1- 14 C]Palmitoyl,2-oleyl PC 86 14 1-Palmitoyl,2-[1- 14 C]oleyl PC 16 84

Fig 4 Complete lipid analysis of control and AtLCAT3-transformed

wild-type yeast FAMEs from individual and total lipids were analysed

by GLC The cultures and analyses were carried out in triplicate.

Standard deviation was less than 2% in analyses of total fatty acids

(TFA) content and less than 15% in analyses of individual lipid classes.

FA, free fatty acids.

membranes, in agreement with a Western blot analysis using

microsomes and supernatant from the

FLAG-AtLCAT3-transformed yeast (data not shown)

Finally the involvement of Ser177, Asp384 and His409 in

the catalysis, by analogy with the conserved catalytic triad

of HsLCAT (Fig 2), was checked by directed mutagenesis

Indeed the point mutations S177A, D384A or H409L

resulted in the disappearance of PLA1 activity and of the

associated TAG production, although the mutated proteins

were produced in similar amounts as AtLCAT3, as

estimated by Western blots of these FLAG-tagged proteins

in the respective microsomes For comparison, Y346F and T352A mutations resulted in a decrease but not the suppression of PLA1 activity and TAG production (Fig 7) although these amino acids are also invariant (Fig 2)

Expression ofAtLCAT3 in wild-type yeast: consequences

on yeast lipid metabolism

In an attempt to relate phospholipid hydrolysis and TAG synthesis, [1-14C]oleic acid was incubated with 12 000 g supernatants from AtLCAT3-transformed and control wild-type yeast, under acylCoA synthase assay conditions (see Experimental procedures) In fact the labelling of TAGs

in the cell-free extract of the transformed yeast was half that

of the control yeast, whereas labelling of PC and other

Fig 6 Substrate specificity of AtLCAT3 towards various

phospho-lipids After incubation of dioleylPC (DOPC),

dioleylphosphatidyl-ethanolamine (DOPE), dioleylphosphatidic acid (DOPA) and

oleyllysoPC (OLPC) for 30 min with AtLCAT3-transformed wild-type

yeast microsomes, the free FA fraction was analysed by GLC of the

FAMEs The values found for the corresponding zero time and the

blank incubated without exogenous substrate were deduced Results

are from duplicate experiments.

Fig 7 Expression of several FLAG-tagged alleles of AtLCAT3 in wild-type yeast Control, void plasmid-transformed yeast; F-Asp384Ala, F-Tyr346Phe, F-His409Leu, F-Ser177Ala, F-Thr352Ala, yeast strains transformed with FLAG-tagged and mutated alleles of AtLCAT3 AtLCAT3 and F-AtLCAT3, non tagged and FLAG-tagged AtLCAT3-transformed yeast (A) Western analysis Microsomes (50 lg protein) were resolved by SDS/PAGE and proteins were immunoblotted with an anti-FLAG serum The mass of 46 kDa cor-responds to the expected mass for AtLCAT3 (B) Microsomal PLA1 activity and (C) TAG content (colorimetric determination) of these strains relative to those from AtLCAT3 Analyses were performed in duplicate on two clones per strain Deviation from the mean was less than 12.5%.

Fig 5 Phospholipid acylhydrolase activity of AtLCAT3 towards

1-myristoyl, 2-oleyl PC After incubation of this PC (250 l M ) with

AtLCAT3-transformed wild-type yeast microsomes (0.125 mgÆmL)1)

in 1 mL for the indicated times, the amounts of myristic and oleic acids

in the free FA (FFA) and lysoPC (LPC) fractions were determined by

GLC analysis of their FAMEs The values found for the

corres-ponding blanks incubated without exogenous substrate were deduced.

Results are from duplicate experiments.

phospholipids was increased by a factor of three-to-four.

The labelling of phospholipids was similarly increased

when microsomes from these yeasts were incubated

with [1-14C]oleylCoA in the presence of either glycerol

3-phosphate (G3PAT assay) or lysophosphatidic acid

(LPAAT assay) These results suggest that in

AtLCAT3-transformed yeast, the FAs and lysophospholipids released

by AtLCAT3 are recycled into the phospholipid pool

As the FS content was doubled in AtLCAT3-transformed

yeast, the HMGCoA reductase activities of microsomes

from transformed and control wild-type yeast were

com-pared This early and rate-limiting enzyme of the sterol

biosynthesis pathway was indeed stimulated by a factor of

2.8, suggesting a coregulation with the phospholipid

biosynthesis to fit the modified phospholipid content and

composition (Fig 4)

In any case the strong depletion in the phospholipid

content of AtLCAT3-transformed yeast as well as the

increased level of lysophospholipids and free FAs do not

seem to be deleterious to this yeast because it grew as well as

the control yeast in terms of cell mass However a

microscopic observation showed that the transformed cells

had a smaller average diameter, a thicker cell wall and

contained more oil bodies (Fig 8)

Discussion

This study describes the characterization of the A thaliana

gene AtLCAT3 (At3g03310) which encodes a protein of

46 kDa sharing 25% identity with human LCAT

(HsLCAT) and 23% identity with yeast phospholipid

diacylglycerol acyltransferase (ScPDAT) Its expression in

yeast, followed by various in vitro studies, clearly

demon-strated that AtLCAT3 encodes a PLA1 (Tables 2 and 3,

Figs 3–6) Subcellular fractionation indicated that

AtLCAT3 is mainly associated with the microsomal frac-tion This result is puzzling as the amino acid sequence of AtLCAT3 does not contain any membrane-spanning domain However no attempts were made to wash the microsomes: the protein might be only adsorbed or weakly bound to the microsomal membranes

The comparison of the lipid content of AtLCAT3-transformed yeast and that of the void plasmid-AtLCAT3-transformed yeast showed a decrease (by a factor 2) of the PC, PE and PS contents and a strong increase of the free FAs, LPC and LPE levels (Fig 4), in accordance with the phospholipid acyl-hydrolase activity shown for AtLCAT3 The decrease in PS suggests that this phospholipid would be another substrate for AtLCAT3, together with PC and PE, whereas phospha-tidylinositol, the level of which was not significantly altered, would not be substrate Interestingly, the FA composition of the PC and PE from the AtLCAT3-transformed yeast diverge significantly from those of the control yeast (Fig 4), suggesting for AtLCAT3 a preference for unsatured FAs If such a selectivity for unsatured FAs at the sn-1 position was confirmed in vitro for AtLCAT3, it could indicate a role in phospholipid remodelling for this novel PLA1

Moreover, the lipid analyses showed an increase of the total FA content in the AtLCAT3-transformed yeast together with an increase of the TAG content of the yeast (Fig 4) This last result confirms the role of yeast TAGs in

FA storage [54] It is noteworthy that the reverse situation was described in a yeast mutant where the three genes encoding phospholipases B had been deleted: the total cellular phospholipid content was increased by 40% whereas the TAG level was reduced by 50% [55]

We performed point mutations on the three conserved amino acids known to be essential for LCATs (catalytic triad common to all Ser hydrolases) and confirmed that AtLCAT3 also requires these three amino acids to be active

Fig 8 Light micrographs of yeast cells after staining with Red Sudan IV Void plasmid-transformed yeast (left) and AtLCAT3-transformed yeast (right) in stationary phase were collected, washed with water, immersed for 30 min in the ethanolic staining solution and rinsed with water The arrows indicate oil bodies The scale bar corresponds to 40 lm.