Synthesis of azidophospholipids and labeling of lysophosphatidylcholine acyltransferase from developing soybean cotyledons

Biochimica et Biophysica Acta 1439 (1999) 47^56 www.elsevier.com/locate/bba

Synthesis of azidophospholipids and labeling of lysophosphatidylcholine acyltransferase from developing soybean cotyledons Ajay W. Tumaney, Ram Rajasekharan * Department of Biochemistry, Indian Institute of Science, Bangalore 560 012, India Received 12 March 1999; accepted 3 May 1999

Abstract A photoreactive substrate analog of lysophosphatidylcholine (LPC), 1-[(4-azidosalicyl)-12-amino]dodecanoyl-sn-glycerol3-phosphocholine (azido-LPC) was synthesized. Fast atom bombardment mass spectrometry was employed to confirm the structures of azido-LPC and its intermediates. Azido-LPC was used to label putative acyl-CoA :LPC acyltransferase from microsomal membranes of developing soybean cotyledons. The synthesized substrate analog acts as a substrate for the target acyltransferases and phospholipases in the dark. When the microsomal membranes were incubated with the acyl acceptor analog and immediately photolyzed, LPC acyltransferase was irreversibly inhibited. Photoinactivation of the enzyme by the photoprobe decreased in the presence of LPC. Microsomal membranes were photolyzed with 125 I-labeled azido-LPC and analyzed by SDS-PAGE followed by autoradiography. These revealed that the analog preferentially labeled 54- and 114-kDa polypeptides. Substrate protected the labeling of both the polypeptides. In our earlier report, the same polypeptides were also labeled with photoreactive acyl-CoA analogs, suggesting that these polypeptides could be putative LPC acyltransferase(s). These results demonstrated that the photoreactive phospholipid analog could be a powerful tool to label acyltransferases involved in lipid biosynthesis. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Fatty acid biosynthesis; Phospholipid analog; Glycine max; Acyltransferase; Acyl-CoA analog; Photoa¤nity labeling

1. Introduction Fatty acids are incorporated into phospholipids Abbreviations: amino-PC, 1,2-diaminododecanoyl-sn-glycerol3-phosphocholine; ASD, [(4-azidosalicyl)-12-amino]dodecanoic acid; azido-LPC, [(4-azidosalicyl)-12-amino]dodecanoyl-sn-glycerol-3-phosphocholine ; azido-PC, 1,2-[(diazidosalicyl)-12-amino]dodecanoyl-sn-glycerol-3-phosphocholine ; FAB/MS, fast atom bombardment/mass spectrometry; ESI/MS, electroscopy ionization mass spectrometry; GPC, glycerophosphocholine; LPC, lysophosphatidylcholine ; PC, phosphatidylcholine; PE, phosphatidylethanolamine ; t-boc, 2-[(tert-butoxycarbonyl)-oxy]imino]2phenylpropionate * Corresponding author. Fax: +91-80-334-1683, 3341814; E-mail: [email protected]

either by the de novo biosynthetic pathway [1] or by the remodeling of phospholipids through the acylation^deacylation pathway [2]. The remodeling pathway for PC and PE [3] is catalyzed by an integral membrane protein acyl-CoA:1-acyl-sn-glycerol-3phosphocholine acyltransferase (EC 2.3.1.23). In plant systems, this enzyme has been shown to provide substrate for desaturases [4,5] and hydroxylase [6] by introducing oleic acid into the sn-2 position of LPC. The oleate in PC is then desaturated or hydroxylated to form polyunsaturated fatty acids or hydroxy oleic acid, respectively. The level of polyunsaturated fatty acids in oilseeds [7] and arachidonic acid in animal systems [8] may largely be controlled

1388-1981 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 8 - 1 9 8 1 ( 9 9 ) 0 0 0 7 3 - 6

BBAMCB 55460 25-6-99

48

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56

by this enzyme. However, this enzyme has not been puri¢ed from any eukaryotic source. To purify this enzyme, we have used an approach that involves covalent labeling of membrane proteins with active site directed radiolabeled substrates, bearing a photoreactive group. Our laboratory has been studying the synthesis and use of photoa¤nity reagents in the identi¢cation of membrane-bound acyltransferases that are involved in plant lipid biosynthesis [9,10]. Photoreactive phospholipid analogs have been used to speci¢cally label the aminophospholipid transporter of erythrocyte membranes [11], ATP synthetase in reconstituted lipids [12], ATPase from sarcoplasmic reticulum [13] and the interfacial recognition site of phospholipase A2 [14]. These analogs have also proven to be useful in determining intermolecular cross-linking between phospholipids [15,16] and the topological arrangement of membrane proteins [17,18]. In this paper, we report the synthesis of a photoa¤nity reagent 1-[(3-iodo-4-azidosalicyl)-12-amino]dodecanoyl-glycerophosphocholine in an attempt to identify LPC acyltransferase in developing soybean cotyledons. Azido-LPC acts as a substrate in the dark and as an irreversible inhibitor upon UV irradiation. Upon photolysis, the photoprobe was covalently incorporated into 54- and 114-kDa polypeptide from microsomal membranes in a speci¢c manner. 2. Materials and methods 2.1. Materials [1-14 C]Oleoyl-CoA (58.3 Ci/mol) and [125 I]NaI (17 Ci/mg/0.1 ml) were obtained from New England Nuclear. Ammonium persulfate and reagents for electrophoresis were from Bio-Rad. NHS-ASA and bicinchoninic acid protein assay reagents were from Pierce. Lipase (Rhizopus arrhizus), phospholipase A2 (Crotalus adamanteus), phospholipase D (Streptomyces chromofuscus) and catalase were from Sigma. Soybean (Glycine max, var. Wye) plants were grown in the ¢eld and pods were harvested from 18 to 20 days after £owering. Cotyledons were collected and stored at 385³C until use.

2.2. Microsomal membrane preparation Developing soybean cotyledons (10 g) were ground in a prechilled mortar and pestle for 5 min with two parts (w/v) of 50 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1000 U/ml catalase and 0.25 M sucrose. The crude homogenate was passed through two layers of miracloth and the ¢ltrate was centrifuged at 18 000Ug for 15 min. The resulting supernatant was further centrifuged at 100 000Ug for 60 min to obtain microsomal membrane fraction. The pellet was washed with the grinding medium and resuspended in a small volume of homogenizing medium without sucrose with the aid of a glass homogenizer. All operations were carried out at 4³C. Microsomal membranes were either used immediately or frozen as small aliquots in liquid nitrogen and stored at 385³C. The activity of LPC acyltransferase was stable at this temperature for a minimum of 3 months. Protein concentrations were determined by the bicinchoninic acid method [19] using bovine serum albumin as the standard. 2.3. LPC acyltransferase assay The assay mixture consisted of 50 mM Tris-HCl, pH 8.0, 20 WM [14 C]oleoyl-CoA (55 000 cpm), 20 WM LPC and 20^30 Wg of microsomal membrane proteins in a total volume of 100 Wl. The reaction was carried out at room temperature for 10 min and stopped by the addition of 0.7 ml chloroform/methanol (1:2, v/v) followed by 0.2 ml 1% acetic acid and the lipid was extracted [20]. The lower chloroformsoluble material was concentrated in a speed-vac and separated on a precoated 250-Wm silica gel G plate using chloroform/methanol/acetic acid/water (170:25:25:6, v/v) as the solvent system. PC was scraped and radioactivity was determined in a liquid scintillation counter. Boiled membranes were used as a control. Control values were subtracted from the values measured in the presence of active microsomal membranes. 2.4. Synthesis of 1,2-diaminododecanoylglycerophosphocholine (amino-PC) 12-Aminododecanoic acid was converted to tertbutoxycarbonyl aminododecanoic acid [21]. Solid 2-

BBAMCB 55460 25-6-99

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56

[(tert-butoxycarbonyl)oxy]amino]2-phenylpropionate (t-boc), 10 mmol, was added to a solution of aminododecanoic acid (1 mmol) in tetrahydrofuran/methanol/water (3:1:3), v/v) which was adjusted to pH 10 with triethylamine. This mixture was stirred at room temperature for 18 h and the organic solvent was evaporated under reduced pressure followed by lipid extraction under acidic conditions. The lower chloroform phase was evaporated under a stream of nitrogen and the residue dissolved in dry methylene chloride. The synthesized compound showed a single spot (Rf 0.68) by TLC using chloroform/methanol/water (65:25:4, v/v) as the solvent system and comigrated with commercial [(t-boc)amino]hexanoic acid. The product was stored at 320³C until use. The blocked fatty acid (0.25 mmol) was stirred with dicyclohexylcarbodiimide (0.15 mmol) in methylene chloride for 2 h at room temperature to give [(t-boc)amino]dodecanoyl anhydride [22]. Commercially available GPC^cadmium chloride complex (100 mg, 0.23 mmol) was suspended in 6 ml of dry chloroform. The synthesized fatty acid anhydride, in 1 ml of chloroform, was added to the above with constant stirring followed by the addition of 61 mg (0.5 mmol) if N,N-dimethyl-4-aminopyridine [15]. The reaction £ask was £ushed with nitrogen and sealed. The mixture was stirred at room temperature for 48 h and the solvent was removed under a stream of nitrogen. The residue was redissolved in 2 ml of chloroform and loaded onto a silicic acid (20 g) column that has been equilibrated with chloroform. The column was washed with chloroform and then eluted with mixtures of chloroform and then eluted with mixtures of chloroform/ methanol (1:1, v/v). The purity was established by TLC using same solvent system as described above. The yield was approximately 57%. The protected phosphatidylcholine analog was deblocked with 6% anhydrous HCl in glacial acetic acid at room temperature for 2 h [23]. The deblocked lipid bound tightly to the silicic acid column and did not elute even with methanol. Hence after deblocking, lipid was extracted and the chloroform soluble material was used for further synthesis without puri¢cation.

49

2.5. Synthesis of 1,2-diazidosalicylaminododecanoylglycerophosphocholine (azido-PC) The synthesis of azido-PC was achieved by acylating the primary amino group of 1,2-diaminododecanoyl-glycerophosphocholine with 4-azidosalicylic acid as described earlier [9]. Amino-PC (0.12 mmol) was dissolved in 4 ml of chloroform and 50 Wl of triethylamine was added to neutralize the solution. N-Hydroxysuccinimidyl 4-azidosalicylic acid (83 mg, 0.3 mmol) in 1 ml chloroform was then added with constant stirring. After 24 h in the dark at room temperature, the solvent was evaporated under a stream of nitrogen. The residue was dissolved in 1 ml chloroform and loaded onto a silicic acid (20 g) column that had been equilibrated with chloroform. The column was washed with 100 ml chloroform and then eluted with a 45-ml mixture of chloroform/methanol (1:1, v/v). The solvent was evaporated under reduced pressure and stored in chloroform at 385³C. 2.6. Synthesis of azido-LPC Azido-PC (0.5 mg) was hydrolyzed with phospholipase A2 to obtain azido-LPC [24]. The enzyme was added thrice in 5-U portions at 1-h intervals. At the end of 3 h, the substrate was completely hydrolyzed. The product was loaded onto a 1-ml Prep-Sep C18 reverse-phase extraction column (Fisher Scienti¢c) which had been washed with 3 ml of methanol and equilibrated with 50 mM Tris-HCl, pH 8.0. The column was washed with 1 ml of equilibration bu¡er followed by 1 ml of 50% methanol and azido-LPC was eluted with 8 ml of methanol. The solvent was evaporated and the residue redissolved in methanol and stored at 385³C. The puri¢ed azido-LPC was iodinated using Na125 I and chloramine-T [25]. The iodinated product was also puri¢ed using reversephase column chromatography. The e¤ciency of iodination was 58^65%. All operations involving azide were carried out under dim safe light. 2.7. Fast atom bombardment/mass spectrometry (FAB/MS) Fast atom bombardment mass spectra were acquired on a VG ZAB-HF double focusing mass spec-

BBAMCB 55460 25-6-99

50

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56

trometer (Manchester, UK) and processed with an OPUS data system. The samples were deposited on a thin layer of glycerol and then introduced into the mass spectrometer by direct probe. Ionization was achieved with an Ion Tech Saddle Field fast atom gun (Middlesex, UK) producing 7-kV Xenon atoms at 1-mA emission current. The mass spectrometer was scanned from 800 to 80 amu at a rate of 10 s/ decade and the positive ions were recorded. 2.8. High resolution/fast atom bombardment/mass spectrometry (HR/FAB/MS) HR/FAB/MS analysis performed with a VGZABHF double-focusing mass spectrometer equipped with an Ion Tech Saddle ¢eld FAB gun. The samples were deposited by a needle syringe on a thin layer of glycerol on a FAB target with or without a reference standard for high resolution analysis and the introduced into the mass spectrometer. Ionization was achieved with the FAB gun typically producing 7.5-kV Xenon atom at 1-mA emission current. The mass spectrometer was adjusted to 8000 resolution. Peak matching using the decade unit on the spectrometer was applied for high resolution analysis. An accurate mass analysis of azido-LPC was done at m/z 616 that used as the reference standard for the analysis of amino-PC at m/z 652. The 652 ion was used as the internal reference for the analysis of fragment ion of (127 I)azido-LPC at m/z 716, which was used later as the internal reference for the protonated molecular ion at m/z 742. Positive ions were analyzed. 2.9. Liquid chromatograph/electroscopy ionization/ mass spectrometry (LC/ESI/MS) Mass spectra were recorded with a JEOL SX-102A double sector mass spectrometer (Tokyo, Japan) and processed with a Hewlett Packard Apollo series 400 computer using JEOL complement software. The sample was introduced into the Analytica Ultraspray interface via a loop injection. The solvent, 50% acetonitrile/1% formic acid was delivered at a rate of 50 Wl/min with an ABI140A solvent delivery system. Ionization was achieved with high-voltage application (3500^4000 V) on the Ultraspray unit. Nitrogen gas heated at 200³C to decluster the solvent (aceto-

nitrile) from the analytes. The spectrometer was scanned from 200 to 700 amu in 3 s and positive ions were recorded. 2.10. Photolabeling The photolabeling reaction was carried out in the cap of a microcentrifuge tube in a ¢nal volume of 50 Wl containing 60 Wg of microsomal membrane proteins in 50 mM Tris-HCl, pH 8.0, 0.5 mM 2-mercaptoethanol and 0.5 WCi (0.5 WM) of the labeled photoprobe [26]. Reaction was preincubated on ice in the dark for 1 min and irradiated for 3 min with a hand-held UV-lamp with the ¢lter removed (5000 WW/cm2 , model UVG-54, UV Products) at a distance of 7 cm. As a control, the photoprobe was ¢rst exposed to UV-light and then added to the microsomal membranes. No labeling was observed in the absence of UV-light. Proteins were precipitated with 10% trichloroacetic acid on ice for 15 min and centrifuged in a microfuge for 15 min. The pellets were washed with cold acetone, solubilized with 50 Wl of loading bu¡er and analyzed by 12% SDS-PAGE [27]. The gel was stained with Coomassie blue R-250, dried, and the labeled proteins were identi¢ed by autoradiography. The stained protein bands were excised and radioactivity was determined by Q-counting. These experiments were repeated three times, each from di¡erent microsomal membrane preparations. 3. Results 3.1. Synthesis of azido-LPC The chemical structure of 125 I-labeled 1-[(4-azidosalicyl)-12-amino]dodecanoyl-sn-glycerol-3-phosphocholine is presented in Fig. 1. The synthesis of the

Fig. 1. Structure of 1-[(3-125 Iodo-4-azidosalicyl)-12-amino]dodecanoyl-sn-glycerol-3-phosphocholine.

BBAMCB 55460 25-6-99

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56

photoreactive LPC analog is described in details in Section 2. The synthesis involved the acylation of protected aminododecanoic acid with GPC to form t-boc-amino-PC. We coupled azidosalicylic acid with deblocked amino-PC to form azido-PC with a ¢nal yield of 44%, based on the amount of aminododecanoic acid used. This compound was completely hydrolyzed with phospholipase A2 and the resultant product, azido-LPC, was radioiodinated by oxidative chloramine-T. A single radioactive and iodine-positive spot was observed with the mobility expected for azido-LPC (Rf 0.11). No free fatty acid, iodine or azido-PC was detected. The synthesized radiolabeled azido-LPC was hydrolyzed with phospholipases C, D

51

and lipase forming 1-ASD-sn-glycerol, 1-ASD-glycerol-3-phosphate and free ASD, respectively. These results indicated that the phospholipases accept the synthesized azido-LPC as the substrate. The puri¢ed compounds and intermediates were con¢rmed by mass spectral analysis (Fig. 2). In the cases of azido-LPC and (127 I)azido-LPC, fragmented ions corresponding to their amine analogs (26 atomic mass units less than the target azide molecules) were observed in Fig. 2B,D. The amine formation clearly occurred in the FAB/MS source since the positive ion electroscopy/mass spectrometry (ESI/MS) was recognized as a more gentle ionization technique [28] than FAB/MS. Azido-LPC showed the only protonated

Fig. 2. Mass spectral data of photoreactive phospholipid and related compounds. (A) FAB/mass spectrum of amino-PC. (B) FAB/ mass spectrum of azido-LPC. (C) ESI/mass spectrum of azido-LPC. (D) FAB/mass spectrum of (127 I)azido-LPC.

BBAMCB 55460 25-6-99

52

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56

Table 1 Accurate mass data of azidophospholipids and related compound Compounds

Ion

Accurate mass

Error (mmu)

Formula

Azido-LPC

616 590 652 716 742

616.3111 590.3206 652.4662 716.2173 742.2078

3.0 1.5 1.6 1.2 3.1

C27 H47 N5 O9 P1 C27 H49 N3 O9 P1 C32 H67 N3 P1 C27 H48 N3 O9 P1 C27 H46 N5 O9 P1

Amino-PC (127 I)Azido-LPC

molecular ion at m/z 616 (Fig. 2C). The accurate mass determinations obtained by positive high resolution FAB/MS are shown in Table 1. The photolability of azido-LPC was demonstrated by monitoring the change in the absorption spectra of the compound following UV irradiation at regular time intervals. A substantial photosensitivity of azido-LPC was evident from the decrease in absorption at 270 and 308 nm with increasing time of irradiation (Fig. 3). Upon irradiation, there was a shift in the absorption maximum from 270 to 274 nm. The absorption at 308 nm completely disappeared after 1 min of photolysis. Complete photodecomposition was achieved after 2 min of irradiation.

Fig. 3. Ultraviolet irradiation of azido-LPC. The photoprobe was in a standard photolabeling reaction mixture and irradiated for various lengths of time, 0 s (999), 15 s (-9W-9W-9), 30 s (-W-W-), 60 s (- - -), 120 s (WWWWWW) and the absorption spectra were taken.

Fig. 4. Analysis of acyltransferase reaction products using [125 I]azido-LPC. The enzyme was assayed and the products analyzed as described in Section 2. A represents an autoradiogram of the TLC. The synthesized [125 I]azido-LPC was analyzed on a TLC using chloroform/methanol/acetic acid/water (170:25:25:6, v/v) as the solvent system. B is an autoradiogram of the reaction products analyzed by two-dimensional TLC. The ¢rst dimension (ascending) was developed using chloroform/methanol/ ammonium hydroxide (65:25:4, v/v), and the second dimension was with chloroform/methanol/acetic acid/water (170:25:25:6, v/v) as the solvent system.

3.2. Interaction of azido-LPC with acyltransferase Microsomal membranes of developing soybean cotyledons were used as a model system to evaluate the e¡ectiveness and speci¢city of the synthesized photoreactive reagent. Microsomal membranes were incubated with 125 I-labeled azido-LPC and oleoyl-CoA in the dark under standard assay conditions. Reaction products were analyzed by thin-layer chromatography followed by autoradiography (Fig. 4B), which indicated a product (Rf 0.21) that migrated close to PC. These results suggested that the phospholipid analog acts as a substrate for LPC acyltransferase. In the presence of 20 WM oleoyl-CoA, the Km values of 22.3 and 20.7 WM were determined for the acyltransferase using azido-LPC and LPC, respectively. In addition, microsomal membranes were assayed for acyltransferase activity using [14 C]oleoyl-CoA after photolysis for 3 min with the analog. Table 2 shows the enzyme activity was una¡ected by UVlight in the absence of photoprobe. Microsomal membranes were photolyzed in the presence of azido-LPC, washed to remove the unbound photoprobe and assayed under standard conditions. The enzyme activity was lost up to 60% after photolysis as com-

BBAMCB 55460 25-6-99

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56 Table 2 Photoinactivation of acyltransferase activity by azido-LPC Addition (WM) LPC 25 0 0 125 250

Irradiation

Activity (%)

+ 3 + + +

100 þ 9.7 96 þ 8.5 38 þ 2.8 64 þ 5.7 89 þ 7.3

Azido-LPC 0 25 25 25 25

Microsomal membranes (120 Wg protein) were photolyzed in a total volume of 0.1 ml containing 25 WM azido-LPC followed by dilution with 50 mM Tris-HCl, pH 8.0. The mixture was pelleted at 100 000Ug for 1 h and assayed under standard conditions using 37.0 Wg protein. The presence or absence of UV irradiation is indicated by + and 3, respectively. One hundred percent activity represents 0.401 nmol of PC formed/min/mg protein. Values are mean percent þ S.D. of three determinations.

pared to unphotolyzed samples suggesting an irreversible inhibition. The acyltransferase activity was protected from inhibition when 10-fold excess of natural LPC was included during photolysis. These results suggested that azido-LPC and LPC were interacting with enzyme at the same site.

53

ates. However, no labeling was observed either in the dark or in the absence of protein. Under photolytic conditions, the 54- and 114-kDa polypeptides were labeled consistently (Fig. 5A, lane 2). There was no increase in the intensity of labeling at 41-kDa protein after photolysis. Addition of oleoyl-CoA in the photolysis mixture resulted in a 5.3-fold increase in labeling of 54-kDa polypeptide, but did not a¡ect the labeling of the 114-kDa polypeptide (Fig. 5A, lane 3). The total protein pattern observed that with Coomassie blue stained gel revealed that these labeled polypeptides are not abundant in microsomal membranes (Fig. 5B). The speci¢city of labeling was demonstrated by the protection from covalent incorporation of the photoprobe by natural LPC. Photolabeling of the 54and 114-kDa polypeptides in microsomal membranes was protected in a concentration-dependent manner (Fig. 6). The half-maximal protection of photoa¤nity labeling of 54- and 114-kDa polypeptides were

3.3. Photoa¤nity labeling of microsomal membranes with Azido-LPC The extent of the covalent incorporation of [ I]azido-LPC into microsomal membranes upon irradiation with UV-light was determined after complete denaturation of proteins followed by SDSPAGE. In these photolabeling experiments, the concentration of photoprobe used (0.5 WM) was less. This was necessary to minimize the non-speci¢c labeling of proteins in the membranes. Autoradiography of the gel revealed that the polypeptides of 41 þ 2, 54 þ 3 and 114 þ 5 kDa were labeled (Fig. 5A). However, a considerable number of polypeptides at about the 60^80-kDa region were also labeled to a much lesser extent and the intensity of labeling of these polypeptides varied considerably between experiments. A heavily labeled band was also observed near the top of the gel. Photolysis of labeled azido-LPC prior to the addition of microsomal membranes labeled a 41-kDa polypeptide (Fig. 5A, lane 1) suggesting that the possibility of existence of long-lived chemically reactive intermedi125

Fig. 5. Photoa¤nity labeling of microsomal membranes with 125 I-labeled azido-LPC. A is an autoradiogram of the SDSPAGE. Lanes 1^3 each contained 60 Wg microsomal membrane proteins and 0.5 WCi acyl acceptor analog. In lane 1, the photoprobe was irradiated with UV-light before the addition of microsomal membranes. In lane 2, microsomal membranes were photolyzed with azido-LPC in the absence of exogenous oleoylCoA. In lane 3, microsomal membranes were photolyzed with the photoprobe in the presence of oleoyl-CoA. B represents the Coomassie blue stained SDS-PAGE of soybean microsomal membrane proteins. The positions of molecular weight markers are indicated by arrow (kDa).

BBAMCB 55460 25-6-99

54

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56

Fig. 6. Protection against photolabeling of the 114- and 54-kDa polypeptides by LPC. Microsomal membranes were photolyzed with labeled azido-LPC in the presence of varying concentrations of LPC. Molecular weight markers are indicated by arrows (kDa). Intact N3 , labeled azido-LPC was added along with the microsomal membranes prior to irradiation with UV-light: 18:1-CoA, oleoyl-CoA.

achieved at approximately 5 WM of LPC. However, the labeling of 41-kDa protein could not be protected even at the highest concentration of LPC used (data not shown). These results suggested that the photoprobe interacted at the acyl acceptor-binding site of the acyltransferase. 4. Discussion 1-[(4-Azidosalicyl)-12-amino]dodecanoyl-sn-glycerol-3-phosphocholine was used as a tool to identify LPC acyltransferase from microsomal membranes of developing soybean cotyledons. Chemical synthesis of alkyl azide LPC analog (12-azido oleoyl-glycerophosphocholine) was described in our earlier study [10]. Labeling of the synthesized compound involved a multi-step process, which gives variable yield with very low speci¢c activity. An e¤cient method for the synthesis and characterization of the radioiodinatable and photoreactive lysophosphatidylcholine analog is described in this paper. AzidoLPC acts as a substrate for acyltransferase, indicating that it interacts at the active site of the enzyme. These experiments also suggest that the acyltransferase has broad substrate speci¢city towards the acyl acceptor. Similar results have also been obtained with alkenyl ether analogs [29] and 1-(12-azidooleoyl)glycero-3-phosphocholine analog of LPC in microsomal membranes of developing sun£ower cotyledons and developing soybean cotyledons, respectively. The fact that the photoprobe and the natural substrate have very close kinetic constants indicated that the analog could be used as a speci¢c

photoa¤nity reagent. The speci¢city of the interaction was demonstrated by protecting the enzyme from photoinactivation upon addition of varying amounts of substrate (Table 2). Azido-LPC speci¢cally labeled the polypeptides of 54 and 114 kDa (Fig. 5) and the labeling depended upon UV irradiation. The intensity of photolabeling of each polypeptide may re£ect the proximity of the polypeptide to the reactive nitrene on the photoprobe for covalent bond formation. These cannot be used as a quantitative stoichiometric measurement of the amount of each polypeptide in the microsomal membranes. To obtain speci¢c labeling, we have done all of the photolabeling experiments in the presence of 0.5 mM 2-mercaptoethanol. At that monothiol concentration, the photoprobe was stable for a minimum of 60 min in the dark. Monothiols act as good nucleophiles and hydrogen atom donors to quench the reactive intermediates (nitrenes and anilinyl radicals) that are not bound to the proteins [30]. The speci¢city of labeling was also shown by the protection experiments (Fig. 6) where the natural acyl acceptor prevented photoinsertion of azido-LPC into proteins, clearly indicating that the photoprobe interacts with acyl acceptor binding site. From these results, it is unlikely that the labeled proteins (54 and 144 kDa) have resulted from non-speci¢c cross-linking of photoprobe. Photolysis of azido-LPC prior to the addition of membranes led to the labeling of a 41-kDa protein, indicating that the long-lived chemically reactive intermediates may be involved in labeling. It has been shown that the lifetime for phenyl nitrenes is in the order of 0.1^1 ms [31]. In our experiments, the light-

BBAMCB 55460 25-6-99

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56

initiated reaction continued in the dark for a minimum of 10 min. Similar observations have also been reported for other photoprobes [32^34]. The intensity of labeling of 41-kDa protein varied considerably between experiments. Labeling of this protein was not inhibited by the substrate and its identity remains to be investigated. In this study, we have identi¢ed a 54 þ 3-kDa polypeptide in microsomal membranes of immature soybean cotyledons. This protein is likely to be LPC acyltransferase. Positive identi¢cation as a putative acyltransferase is supported by the following ¢ndings: (1) photoa¤nity labeling is protected by natural LPC in a concentration-dependent manner; (2) the labeling was enhanced 5.3-fold upon addition of oleoyl-CoA; and (3) the same protein was also labeled with 12-azidooleoyl-CoA [26]. The interaction of azido-LPC with the 114 þ 4-kDa protein also suggests that this also could be LPC acyltransferase. However, we have not observed oleoyl-CoAdependent labeling in the 114-kDa protein. This could be due to the presence of endogenous acyl donor in microsomal membranes. It is also possible that the 114-kDa protein has higher a¤nity towards acyl-CoA than 54-kDa protein or possibly that it binds the acyl acceptor before it binds to acyl donor. Previously, we have shown that 12-azidooleoyl-CoA and ASD-CoA speci¢cally labeled a 110-kDa protein in microsomal membranes of developing soybean cotyledons [26]. It may be concluded from these experiments that the less abundant and strongly labeled 54and 114-kDa polypeptides are putative LPC acyltransferases. However, we do not have any direct evidence to support this conclusion. Photoreactive phospholipid analogs have been used for structural and topological studies of membrane proteins [17,35]. There are a very limited number of reports on the use of nitrene-generating lipid analogs for functional studies [36,37]. The present study represents an example of a successful application of a photoreactive phospholipid analog for labeling of putative microsomal membrane-bound LPC acyltransferase(s). Acknowledgements This research was supported by the Center for

55

Scienti¢c and Industrial Research and Department of Science and Technology, New Delhi.

References [1] E.P. Kennedy, Fed. Proc. 20 (1961) 934^940. [2] W.E.M. Lands, J. Biol. Chem. 235 (1960) 2233^2237. [3] R. Rajasekharan, V. Nachiappan, Arch. Biochem. Biophys. 311 (1994) 394^398. [4] M.I. Gurr, M.P. Robinson, A.T. James, Eur. J. Biochem. 9 (1969) 70^78. [5] C.R. Slack, P.G. Roughan, J. Browse, Biochem. J. 179 (1979) 649^653. [6] M. Bafor, M.A. Smith, L. Jonsson, K. Stobart, S. Sytmne, Biochem. J. 280 (1991) 507^514. [7] S. Stymne, A.K. Stobart, in: P.K. Stumpf (Ed.), The Biochemistry of Plants, Vol. 9, Academic Press, New York, 1987, pp. 177^214. [8] R. Irvine, Biochem. J. 204 (1982) 3^16. [9] R. Rajasekharan, R.C. Marian, J.M. Shockey, J.D. Kemp, Biochemistry 32 (1993) 12386^12391. [10] R. Rajasekharan, J.D. Kemp, J. Lipid Res. 35 (1994) 45^51. [11] A.J. Schriot, J. Madsen, A.E. Ruoho, Biochemistry 26 (1987) 1812^1819. [12] J. Hoppe, C. Montecucco, P. Friedl, J. Biol. Chem. 258 (1983) 2882^2887. [13] R. Bisson, C. Montecucco, Biochem. J. 193 (1981) 757^763. [14] K.-S. Huang, J.H. Law, Biochemistry 20 (1981) 181^187. [15] C.M. Gupta, R. Radhakrishnan, H.G. Khorana, Proc. Natl. Acad. Sci. USA 74 (1977) 4315^4319. [16] J. Brunner, F. Richards, J. Biol. Chem. 255 (1980) 3319^ 3329. [17] H. Baylay, J.V. Staros, in: E.F.V. Scriven (Ed.), Azides and Nitrenes: Reactivity and Utility, Academic Press, New York, 1984, pp. 433^490. [18] Y. Takagaki, R. Radhakrishnan, C.M. Gupta, H.G. Khorana, J. Biol. Chem. 258 (1983) 9128^9135. [19] P.K. Smith, R.I. Krohn, G.T. Hermanson, A.K. Mallia, F.H. Gartner, M.D. Provenzano, E.K. Fujimoto, N.M. Goeke, B.J. Olsen, D.C. Klenk, Anal. Biochem. 150 (1985) 76^85. [20] E.G. Bligh, W.J. Dyer, Can. J. Biochem. Physiol. 37 (1959) 911^917. [21] A.J. Schroit, J.W. Madsen, Biochemistry 22 (1983) 3617^ 3623. [22] Z. Selinger, Y. Lapidot, J. Lipid Res. 7 (1966) 174^175. [23] J.W. Nichols, R.E. Pagano, Biochemistry 20 (1981) 2783^ 2789. [24] M. Kates, in: Technique in Lipidology, Elsevier/North Holland, New York, 1972, pp. 568^570. [25] I. Ji, J. Shin, T.H. Ji, Anal. Biochem. 151 (1985) 348^349. [26] R. Rajasekharan, V. Nachiappan, H.R. Roychowdhury, Eur. J. Biochem. 220 (1994) 1013^1018. [27] U.K. Laemmli, Nature 227 (1970) 680^685.

BBAMCB 55460 25-6-99

56

A.W. Tumaney, R. Rajasekharan / Biochimica et Biophysica Acta 1439 (1999) 47^56

[28] C.M. Whitehouse, R.N. Dreyer, M. Yamashita, J.B. Fenn, Anal. Chem. 57 (1985) 675^679. [29] P. Sperling, E. Heinz, Eur. J. Biochem. 213 (1993) 965^971. [30] H. Baylay, J.R. Knowles, Methods Enzymol. 46 (1977) 69^ 114. [31] J.V. Staros, Trends Biochem. Sci. 5 (1980) 320^322. [32] J.A. Maassen, W. Moller, J. Biol. Chem. 253 (1978) 2777^ 2783. [33] M.T. Mas, J.K. Wang, P.A. Hargrave, Biochemistry 19 (1980) 684^692.

[34] A.W. Nicholson, C.C. Hall, W.A. Strycharz, B.S. Cooperman, Biochemistry 21 (1982) 3797^3808. [35] J. Brunner, Methods Enzymol. 172 (1989) 630^687. [36] J. Connor, A.J. Schroit, Biochim. Biophys. Acta 1066 (1991) 37^42. [37] M. Michalak, E.L. Wandler, K. Strynadka, G.L. Lopaschuk, W.M. Njue, H.-J. Liu, P.M. Olley, FEBS Lett. 265 (1990) 117^120.

BBAMCB 55460 25-6-99