Molecular cloning of a phospholipid-cholesterol acyltransferase from Aeromonas hydrophila. Sequence homologies with lecithin-cholesterol acyltransferase and other lipases

Biochimica et Biophysics Acra 959 (1988) 153-159 Elsevier

153

BBA 52761

Molecular cloning of a phospholipid-cholesterol from Aeromonus hydrophifa

acyltransferase

Sequence homologies with lecithin-cholesterol

acyltransferase and other lipases

Julian Thornton a, S. Peter Howard b and J. Thomas Buckley a aDepartment of Biochemistry and Microbiology, University of Victoria, Victoria (Canada) and b Centre de Biochimie et de Biologie Moleculaire, CNRS, Marseille (France) (Received

13 August

1987)

Key words: Acyltransferase; Glycerophosphohpid-cholesterol acyltransferase; Phospholipid-cholesterol acyltransferase; Phosphatidylcholine-sterol acyltransferase; Nucleotide sequence; Gene clone; Sequence homology; (A. hydrophila)

We have determined the nucleotide sequence of a gene encoding Aeromonas hydrophircl phospholipid-cholesterol acyltransferase, an enzyme which shares many properties with mammalian lecithin:cholesterol acyltransfera~. The derived amino acid sequence of the protein contains two regions which are homologous to the proposed active sites and binding sites of the plasma acyltransferase and to similar sequences in other interfacially acting lipolytic enzymes. The amino terminus is preceded by a typical 18 amino acid signal sequence. The protein, which is released into the culture supematant by Aeromonas hydrophila, is confined to the periplasm of Escherichia coli.

Introduction The glycerophospholipid-cholesterol acyltransferase (GCAT) released by members of the K’ibrio family shares a number of properties with the mammalian enzyme lecithin-cholesterol acyltransferase (LCAT, [l-4]). Like the plasma enzyme, GCAT will catalyze fatty acid transfer between phosphatidylcholine and cholesterol, and acyl transfer is 2-position specific. Both enzymes will act as phospholipases in the absence of acyl acceptors, and neither requires calcium. The microbial enzyme has much less stringent requirements for the acyl donor than does LCAT, as all the comAbbreviations: GCAT, glycerophosphohpid-cholesterol transferase; LCAT, lecithin-cholesterol acyltransferase 2.3.1.43).

acyl(EC

Correspondence: and Microbiology, V8W 2Y2.

J.T. Buckley, Department of Biochemistry University of Victoria, Victoria BC, Canada

0005-2760/88/$03.50

0 1988 Elsevier Science Publishers

mon glycerophospholipids will function as substrates, and under some conditions GCAT will hydrolyse other hydrophobic substrates such as cholesteryl ester. As with LCAT, the selectivity for the acyl acceptor is considerable and appears to depend upon specific interactions between the phospholipid and cholesterol. The primary structure of human LCAT has recently been deduced from the nucleotide sequence of the cDNA [5] and confirmed by protein sequencing [6]. Two regions of the molecule have been identified which are homologous with amino acid sequences in porcine pancreatic lipase and bovine milk lipoprotein lipase [7]. Each contains a serine. Based on the experimental evidence obtained with porcine lipase, one site, found in a variety of lipases [8], is believed to be involved in substrate binding. The other is thought to be the active site. In spite of these observations, nothing is known of the role of individual amino acids in the reactions catalysed in the lipases, nor is it

B.V. (Biomedical

Division)

154

known what portions of the molecule participate in the unusual reaction mechansm of LCAT. None of the enzymes so far compared to LCAT preferentially catalyses acyl transfer. For these reasons we felt it especially important to determine the sequence of the microbial acyltransferase. In this communication we show that sequences similar to the binding and active sites of the mammalian lipases can be identified in GCAT. Experimental

procedures

Materials

Restriction enzymes and DNA modification enzymes were purchased from Pharmacia or Boehringer. Replicative form DNA of Ml3 mp18 and mp19 and Ml3 17-base primer were obtained from Pharmacia and New England Biolabs. Deoxy and dideoxynucleotide triphosphates were from Pharmacia. [a- 32P]dATP was purchased from Amersham. Other chemicals were the purest commercially available. Bacterial strains, plasmids

and media

The bacterial strains and plasmids used in this study and their sources are listed in Table I. The organisms were grown in YT medium, supplemented where specified with antibiotics at the following concentrations: ampicillin, 200 pg/ml; kanamycin, 50 pg/ml. Blood agar was prepared by adding 20 ml of human blood to 500 ml of tryptic soy agar tempered to 50 o C.

TABLE

I

BACTERIA,

PLASMIDS

AND

BACTERIOPHAGES

Strain

Genotype

E. coli

recA13, hsdS20, aral4, proA lacY1, gaIK2, IeuB6, rpsL20 ~~115, mtll, supE44

E.E. Ishiguro

E. co/i JM105

pro-lac, thi, rpsL, hsdR4, endA, sbcB, F’traD36, proA+ B+, lacIq, 1acZ Ml5

Pharmacia

KmR, TcR ApR, TcR See text See text Ml3 sequencing

E.W. Nester E.E. Ishiguro S.P. Howard This study Pharmacia

pVK102 pBR322 pHEc1 PHEc2.2 mp18/mp19

or description

vectors

Source

DNA preparation A. hydrophila

Ah65 chromosomal DNA was isolated by the method of Puhler and Timmis [9]. The DNA was partially digested with Sal1 (0.3 U SalI/pg chromosomal DNA) for 1 h at 37 o C and fragments of approx. 25 kb were isolated by agarose electrophoresis and further purified by centrifugation through a lo-40% sucrose density gradient as described by Maniatis et al. [lo]. Plasmid DNA, when used for cloning or restriction mapping, was isolated via the cleared lysate method of Godson and Vapnek [ll] and purified by centrifugation in a cesium chloride-ethidium bromide gradient as described by Maniatis et al. [lo]. Minipreparations of plasmids, used for screening of insert size in cloning experiments were isolated according to the method of Birnboim and Doly [12].

Construction of the pVKlO2 gene bank and detection of clones containing GCA T-expressing plasmids

The wide host range cosmid pVK102 [13] was digested with Sal1 and dephosphorylated with calf intestinal alkaline phosphatase. The digested cosmid was mixed with the size-selected Sal1 partial digest of A. hydrophifa chromosomal DNA and ligated overnight at 12°C. Aliquots containing 1.2 /.tg of the mixture were packaged in vitro into lambda phage particles using a commercial packaging kit (Amersham) and used to transduce E. coli HBlOl. Detection of clones containing GCAT-expressing plasmids was based on our observation that cells which produce the enzyme (including Aeromonas salmonicida and aerolysin negative mutants of A. hydrophila), exhibit alpha hemolytic phenotypes after 48 h when plated on human blood agar. Positive clones detected in this way were tested for acyltransferase activity as described by MacIntyre and Buckley [l].

Subcloning

metho&

Vector DNA was pBR322, or the RF DNA of Ml3 mp18 or mp19. Target DNA was the recombinant plasmid pHEC1 which contains a 18 kb insert of A. hydrophila with the entire GCAT gene. Subcloning procedures were carried out essentially as described by Maniatis et al. [lo].

155

DNA sequence analysis The plasmid pHEC2.2, which contains a 1.2 kb insert in pBR322, was digested with the restriction enzymes shown in Fig. 1, either alone or in combination, and the resulting fragments were inserted into the polylinker cloning regions of Ml3 mp18 and mp19 [14]. The DNA inserts were sequenced using the dideoxy chain termination method of Sanger et al. [15] according to the strategy depicted in Fig. 1. The universal 17-base single stranded primer was used to begin sequencing from the 3’ end of the inserts in each of the recombinant mp18 and mp19 bacteriophages. Where indicated in Fig. 1, sequencing was continued using 18-base oligonucleotide primers which were synthesized using a Sam One DNA synthesizer (Biosearch Inc.). Cell fractionation procedures Shock fluids were obtained by the sucroseEDTA method of Willis et al. [16]. The osmotically shocked cells were disrupted by passing them through a French pressure cell (1100 kg/cm*). RNAase, a periplasmic marker, was assayed by measuring the release of acid-soluble adenylate from [ ‘Hlpolyadenylate [17]. The intracellular marker glutamate dehydrogenase was assayed as described by Halpem and Lupo [18]. Enzyme activities for all fractions are given in units/ml of original culture. Results and Discussion

Isolation of GCA T-containing clones Acyltransferase activity was detected in disrupted cells after overnight growth of one of the alpha hemolytic clones. This confirmed the presence of the gene for GCAT rather than expression of a modified aerolysin gene or a clone of another phospholipase which could presumably also have been selected by the screening procedure. A large amount of the cosmid, containing an 18 kb insert, was purified from a cleared lysate by cesium chloride ccntrifugation. This cosmid (pHEC1) was restricted with a variety of endonucleases and the resulting fragments were ligated to the appropriately restricted subcloning vector pBR322. One of the subclones, identified on the basis of alpha hemolytic phenotype as well as enzyme pro-

Fig. 1. Strategy for sequencing the GCAT gene. Arrows represent sequences obtained using fragments produced by restriction enzyme digestion at the sites indicated or sequences obtained using synthetic oligonucleotides as primers.

duction, is depicted in Fig. 1. The recombinant plasmid pHEC2.2 contains a 1.2 Kb Pstl fragment inserted into the PstI site of pBR322. Location of GCA T in E. coli clones Most of the acyltransferase activity was recovered in the shock fluid of cells containing pHEC1 and pHEC2.2 (Table II). No enzyme was detected in the culture supernatant, indicating that GCAT is not released by E. coli. Similar observations have been made with a variety of extracellular proteins cloned into this species, leading to the conclusion that E. coli lacks a mechanism for the export of proteins across its outer membrane [19]. Nucleotide sequence of the GCAT gene Using the sequencing strategy shown in Fig. 1, the nucleotide sequence containing the GCAT gene was determined. The results are presented in Fig. 2. The nucleotide sequence corresponding to the beginning of the protein was identified by comparing the translated amino acid sequence with the previously determined amino terminal sequence of GCAT purified from A. salmonicida. Only one of the first 18 amino acids is different in the enzymes from the two species. GCAT from A. salmonicida contains a threonine at position 3 rather than a serine (data not shown here). The open reading frame shown in the figure encodes a protein of 281 amino acids with a molecular weight of 31303. This is considerably larger than the molecular weight of the protein from A. salmonicida as determined by sodium dodecyl sulphate electrophoresis [l]. This may reflect major differences between the two proteins. (In spite of the similarity of their amino termini, plyclonal antibody pre-

156

CCGACAcrccorax;ca:~rn~cx:crcAa:ncnccATCAATCAGCCATPCC~TCA(SPA

1266

I.278

1290

ECWl'CATpGATXQITCIGa3;cATGGl!GGcCAGocXCICCECAG 1332 1344 1356

1302 P&I

1314

1368

Fig. 2. DNA sequence of the GCAT gene. The nucleotide residues are numbered in the 5’ to 3’ direction. The translated above the DNA sequence. Numbers below the sequence refer to the nucleotide positions, numbers above positions. The sequence from - 18 to - 1 is the probable signal sequence. The sequence corresponding to the ammo the enzyme from A. salmonicida is underlined (see text). Several restriction sites are identified for reference

coding region is to the amino acid acid sequence of to Fig. 1.

157

pared against A. salmonicida GCAT did not cross-react with A. hydrophila GCAT.) Alternatively, GCAT may be processed post-translationally. We have shown that another protein produced by this species, the hole-forming toxin aerolysin, is released from the bacteria as a protoxin which is activated by removal of approximately 20 amino acids from the carboxy terminus

a variety of lipases [7,8,23]. The results in Fig. 3A show that GCAT contains a sequence which is highly conserved in lipases with different origins and different biochemical properties. Based on data obtained with porcine pancreatic lipase [24], this region is believed to form the interfacial lipid-binding site of these enzymes. Komaromy and Schotz [8] have noted that hepatic lipase contains two such sequences and they suggest that this enzyme may be able to bind lipids more efficiently than the other enzymes, thereby eliminating the need for protein cofactors. GCAT contains only a single copy of the binding sequence, however, and it has no cofactor requirements. GCAT also contains a sequence similar to regions in several lipases which Maraganore and Heinrickson [7] suggest contain the active sites of these enzymes. The data in Fig. 3B indicate that, although the apparent homology is not convicing, this region in GCAT is if any thing more similar to pancreatic lipase than is the sequence in LCAT. Interestingly, like pancreatic lipase, GCAT is insensitive to diisopropylphosphofluoroidate at concentrations which completely inhibit LCAT [25]. Like the other lipases, GCAT is a hydrophilic

WI. The codon usage frequency in the translated reading frame is very similar to the frequency we have observed for the aerolysin gene of this species (not shown here). In contrast to the aerolysin gene [21], there are no obvious promoter and terminator regions surrounding the gene for GCAT, which would suggest that it is translated from a polycistronic message. Comparison with the amino acid sequences of other lipases

A search of the National Biomedical Research Foundation protein sequence library [22] failed to reveal any sequences with extended regions homologous to GCAT. Recently, however, several authors have reported sequence similarities among

TABLE

II

DISTRIBUTION Shocked

OF ACYLTRANSFERASE

cells were French

HBlOl supematant

pressed

AND

after shocking

MARKER

ENZYMES

AFTER

OSMOTIC

SHOCK

to release their contents.

Acyltransferase (nmol cholesteryl ester formed/min per ml)

Glutamate ( p mol NADH /min per ml)

U/ml

U/ml

(W)

oxidized

RNAase (nmol adenylate /mitt per ml)

(W)

U/ml

0

0

0

wash shock fluid shocked cells

0 0 0

0 0 0.48

(100)

0.18 0.94 0.34

HBlOl (pHEC1) supematant

0

0

0

0 0 0.36

(100)

0.12 1.22 0.27

0

0

0 0.06 0.32

(15) (85)

0.12 0.80 0.42

wash shock fluid shocked cells

0 5.5 1.8

HBlOl (pHEC2.2) supematant

0.1

wash shock fluid shocked cells

0 7.0 0.8

(0) (0) (75) (25)

(1.4) (89) (9.7)

released

(%)

(12) (65) (23)

(7) (75) (8)

(8) (61) (31)

158 A: Lipid Binding

Domain

MicrcbialGCAT Rat heptic

gAq.Ile.Val.Met.Fbe.Gly.Asp.Ser.Le".Ser

lips2

26ker.Val.His.Leu.Phe.Ile.Asp.Ser.leu.Gln

~lippmteinlipse

244Ser.Ile.His.Leu.me.Ile.Asp.Ser.IEu.Le"

Rat lingual lipase

163Lys.Ile.His.~.Val.Gly.His

HLmlanLcAT

173F'ro.Val.Rze.I_eu.Ile.Gly.His.Ser.Leu.Gly

Fmxine pancreatic 1ipas.e

145Asn.Val.His.Val.Ile.Gly.His.Ser.Leu.Gly

B:Active

Ser.Gln.Gly

site

GCAT

232Trp.Lys.F%-o.Ehe.Ala.Ser.Aq.Sor.Ala.Ser.~.Asp.

Pancreatic lipas

l%rp.Lys.

.Gly.Gly.Ser.Ary.l%r.Gly.Tyr.Thr.Glu.Aa.Ser.Gln

LCAT

%l-rp.

.Gly.Gly.Ser.Ile.Lys.F?m.Met.Leu.Val.Leu.Ala.Ser.

.Ser.Gln

Fig. 3. Amino acid homology between GCAT and other hpases. A: homologies with the lipid-binding domain of porcine pancratic lipase and other hpases. The numbers refer to amino acid positions. B: comparison with the proposed active site of porcine lipase and a similar region of LCAT. See text for details.

protein. Analysis of the predicted amino acid sequence [26,27] indicated an average (H) 19 hyand no long hydrophobic dropathy of -0.56 stretches. Site-directed mutagenesis should establish whether or not the regions identified in Fig. 3 are actually involved in the GCAT reaction and detailed analysis of the sequences of the mammalian and microbial acyltransferases may lead to an understanding of why their reaction mechanisms appear so similar. Acknowledgement We gratefully acknowledge the assistance of Margaret Green. This research was supported by the National Science and Engineering Research Council of Canada and the British Columbia Heart Foundation. References 1 Buckky, J.T., Halasa,

L.N. Biol. Chem. 255, 3320-3325.

and

Maclntyre,

S. (1981)

J.

2 Buckley, J.T. (1982) Biochemistry 21, 6699-6703. 3 Buckley, J.T. (1983) Biochemistry 22, 5490-5493. 4 Buckley, J.T., McLeod, R. and Frohlich, J. (1984) J. Lipid Res. 25, 913-918. 5 McLean, J., Fielding, C., Drayna, D., Dieplinger, H., Baer, B., Kohr, W., Henzel, W. and Lawn, R. (1986) Proc. Natl. Acad. Sci. USA 83, 2335-2339. 6 Yang, C., Manoogian, D., Pao, Q., Lee, F., Knapp, R.D., Gotto, A.M. and PownaB, H.J. (1987) J. Biol. Chem. 262, 3086-3091. 7 Maraganore, J.M. and Heimikson, R.L. (1986) Trends Biothem. Sci. 11, 497-498. 8 Komaromy, M.C. and Schotz, M.C. (1987) Proc. Natl. Acad. Sci. USA 84, 1526-1530. 9 Puhler, A. and Timmis, K.N. (1984) Advanced Molecular Genetics, Springer-Verlag, Berlin. 10 Maniatis, T., Fritsch, E.F. and Sambrook, J. (eds.) (1982) Molecular Cloning - A Laboratory Handbook Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 11 Godson, G.N. and Vapnek, D. (1973) B&him. Biophys. Acta 299, 516-520. 12 Bimboi, H.C. and Daly, J. (1979) Nucleic Acids Res. 7, 1513-1523. 13 Knauf, V.C. and Nester, E.W. (1982) Plasmid 8, 45-54. 14 Messing, J. (1983) Methods Enzymol. 101, 20-77. 15 Sanger, F., Nicklen, S. and Cot&on, A.R. (1979) Proc. Natl. Acad. Sci. USA 74, 5463-5467.

159 16 Willis, R.C., Morris, R.G., Ciiakoght, C., Schellenberg, G.D., Gerber, N.H. and Furlong, C.E. (1974) Arch. Biothem. Biophys. 161, 64-75. 17 Lopes, J., Gottfried, S. and Rothfield, L. (1972) J. Bacterial. 109, 520-525. 18 Halpem, Y.S. and Lupo, M. (1965) J. Bacterial. 90, 1288-1295. 19 Howard, S.P. and Buckley, J.T. (1986) Mol. Gen. Genet. 204, 289-295. 20 Howard, S.P. and Buckley, J.T. (1985) J. Bacterial. 163, 336-340. 21 Howard, S.P., Garland, W.J., Green, M.J. and Buckley, J.T. (1987) J. Bacterial. 169, 2869-2871.

22 Lipman, D.J. and Pearson, W.R. (1985) Science 227, 1435-1441. 23 Wion, K.L., Kirchgessner, T.G., Lusis, A.J., Schotz, M.C. and Lawn, R.M. (1987) Science 235, 1638-1641. 24 Verger, R. (1984) in Lipases (Borgstrom, B. and Brockman, H.L., eds.), pp. 84-150, Elsevier, Amsterdam. 25 Juahiainen, M. and Dolphin, P.J. (1986) J. Biol. Chem. 261, 7032-7043. 26 Kyte, J. and Doolittle. R.F. (1982) J. Mol. Biol. 157, 105-132. 27 Eisenberg, D. (1984) J. Mol. Biol. 179, 125-142.