- Email: [email protected]
Remodeling of phosphatidylglycerol in Synechocystis PCC6803
Biochimica et Biophysica Acta 1801 (2010) 163–170
Contents lists available at ScienceDirect
Biochimica et Biophysica Acta j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b b a l i p
Remodeling of phosphatidylglycerol in Synechocystis PCC6803 Hajnalka Laczko-Dobos a, Petr Fryčák b,1, Bettina Ughy a, Ildiko Domonkos a, Hajime Wada c, Laszlo Prokai b, Zoltan Gombos a,⁎ a b c
Institute of Plant Biology, Biological Research Center, Hungarian Academy of Sciences, H-6701 Szeged, P.O. Box 521, Hungary Department of Molecular Biology and Immunology, University of North Texas Health Science Center, 3500 Camp Bowie Boulevard, Fort Worth, TX 76107, USA Department of Life Sciences, Graduate School of Arts and Sciences, University of Tokyo, Komaba, Tokyo 153-8902, Japan
a r t i c l e
i n f o
Article history: Received 9 July 2009 Received in revised form 16 October 2009 Accepted 19 October 2009 Available online 24 October 2009 Keywords: Phosphatidylglycerol Remodeling Membrane lipid Photosynthetic organism Mass spectrometry Synechocystis PCC6803
a b s t r a c t The phosphatidylglycerol deficient ΔpgsA mutant of Synechocystis PCC6803 provided a unique experimental system for investigating in vivo retailoring of exogenously added dioleoylphosphatidylglycerol in phosphatidylglycerol-depleted cells. Gas chromatographic analysis of fatty acid composition suggested that diacyl-phosphatidylglycerols were synthesized from the artificial synthetic precursor. The formation of new, retailored lipid species was confirmed by negative-ion electrospray ionization–Fourier-transform ion cyclotron resonance and ion trap tandem mass spectrometry. Various isomeric diacyl-phosphatidylglycerols were identified indicating transesterification of the exogenously added dioleoylphosphatidyl-glycerol at the sn-1 or sn-2 positions. Polyunsaturated fatty acids were incorporated selectively into the sn-1 position. Our experiments with Synechocystis PCC6803/ΔpgsA mutant cells demonstrated lipid remodeling in a prokaryotic photosynthetic bacterium. Our data suggest that the remodeling of diacylphosphatidylglycerol likely involves reactions catalyzed by phospholipase A1 and A2 or acyl-hydrolase, lysophosphatidylglycerol acyltransferase and acyl-lipid desaturases. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Remodeling of lipids by manipulating fatty acid composition is one of potential factors involved in the adaptation to environmental stresses [1,2]. However, this metabolic process has not been well understood. Both prokaryotes and eukaryotes respond to various environmental stresses [3] by altering the fatty acid composition of
Abbreviations: PG, phoshatidylglycerol; PG(18:1/18:1), 1,2-di-(9E-octadecenoyl)sn-glycero-3-phospho-(1′-rac-glycerol); PG(14:0,14:0), 1,2-dimyristoyl-sn-glycero-3phospho-(1′-rac-glycerol); PG(18:2/18:0), 1-(9Z,12Z-octadecadienoyl)-2-octadecanoyl-sn-glycero-3-phospho-(1′-rac-glycerol); PG(18:2/18:1), 1-(9Z,12Z-octadecadienoyl)-2-(9E-octadecenoyl)-sn-glycero-3-phospho-(1′-rac-glycerol); PG(18:3/18:0), 1-(9Z,12Z,15Z-octadecatrienoyl)-2-octadecanoyl-sn-glycero-3-phospho-(1′-rac-glycerol); PG(18:1/16:0), 1-(9Z-octadecenoyl)-2-hexadecanoyl-sn-glycero-3-phospho(1′-sn-glycerol); PG(16:0/18:1), 1-hexadecanoyl-2-(9Z-octadecenoyl)-sn-glycero-3phospho-(1′-rac-glycerol); PG(16:1/18:0), 1-(9Z-hexadecenoyl)-2-octadecanoyl-snglycero-3-phospho-(1′-rac-glycerol); PG(18:2/16:0), 1-(9Z,12Z-octadecadienoyl)-2hexadecanoyl-sn-glycero-3-phospho-(1′-rac-glycerol); PG(16:1/18:1), 1-(9Z-hexadecenoyl)-2-(9Z-octadecenoyl)-sn-glycero-3-phospho-(1′-rac-glycerol); 16:0, hexadecanoic acid (palmitic acid); 16:1, hexadecenoic acid (palmitoleic acid); 18:0, octadecanoic acid (stearic acid); 18:1, 9-octadecenoic acid (oleic acid); 18:2, 9,12-octadecadienoic acid (linoleic acid); 18:3, 9,12,15-octadecatrienoic acid (α-linolenic acid); ΔpgsA+PG 1d and 21 d indicate mutant cells grown in the presence of PG for 1 and 21 days, respectively ⁎ Corresponding author. Tel.: +3662599704; fax: +3662433434. E-mail address: [email protected] (Z. Gombos). 1 Present address: Department of Analytical Chemistry, Faculty of Science, Palacky University in Olomouc, Tr. 17. listopadu 12, 77146 Olomouc, Czech Republic. 1388-1981/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.bbalip.2009.10.009
their membrane lipids. The extent of unsaturation of fatty acids in thylakoid membranes increases and/or the chain length of fatty acids decreases according to growth temperature. In photosynthetic organisms thylakoid membranes are composed of glycolipids, such as monogalactosyldiacylglycerol (MGDG), di-galactosyldiacylglycerol (DGDG), sulfo-quinovosyldiacylglycerol (SQDG), and one phospholipid, phosphatidylglycerol (PG) [4]. The remodeling of PG has crucial role in both cyanobacteria [5,6] and higher plants in the adaptation to low temperatures [7-10]. The biosynthetic pathway of PG helps to understand the remodeling process. The pathway of PG biosynthesis is well known in prokaryotes [11]. The biosynthesis of PG in cyanobacteria is similar to that in Escherichia coli [5,12]. The five major enzymes of PG biosynthesis are as follows: glycerol-3phosphate acyltransferase (G3P acyltransferase), lysophospatidicacid acyltransferase (LPA acyltransferase), CDP-diacylglycerol synthase (CDP-DG synthase), phosphatidylglycerol phosphate synthase (PGP synthase), and PGP phosphatase. As a consequence of the substrate preferences of G3P acyltransferase and LPA acyltransferase, in the newly synthesized PG 18:0 or 16:0 are esterified at the sn-1, and 16:0 at the sn-2 position of the glycerol backbone [13-15]. At the beginning of PG synthesis the newly synthesized PG molecules contain only saturated fatty acids at both the sn-1 and sn-2 positions [16,17]. At a later step of PG synthesis, they are desaturated by acyllipid desaturases, which introduce a double bond at a specific position of the fatty acids, predominantly at sn-1 [18,19]. Earlier PG remodeling studies used genetic manipulation of acyltransferases and desaturases to reveal the protective role of
164
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
various PG molecular species in resistance to low and high temperature stresses in both higher plants [20] and cyanobacteria [2,21,22]. In plants selective expression of acyltransferases allowed manipulation of the unsaturation of PGs. Overexpression of squash (Cucurbita moschata) and Arabidopsis G3P acyltransferases in tobacco (Nicotiana tabacum) altered the levels of saturated molecular species of PG [20,23]. Heterologous expression of the Arabidopsis gene that increased unsaturated fatty acid content of PG at the sn-1 position enhanced chilling resistance, whereas that from squash, which has strong preference to palmitoyl-acyl carrier protein, increased chilling susceptibility in transgenic tobacco [20]. The level of unsaturation of membrane glycerolipids was manipulated by modulating the genes of fatty acid desaturases both in cyanobacteria and plants [2,22,24]. The presence of diunsaturated glycerolipids in the membranes of photosynthetic organisms enhanced chilling tolerance and played a critical role in the recovery from low temperature photoinhibition [25,26]. Temperature stress-induced remodeling of lipids was demonstrated in the salt-tolerant green alga Dunaliella salina [27-29]. However, the mechanism of remodeling, which resulted in the synthesis of various molecular species with redistribution of acyl chains within individual lipid classes, has not been elucidated [30,31]. In Synechocystis PCC6803, the genes for LPA acyltransferase (sll1848) [15], CDP-DG synthase (cdsA, slr1369) [32], and PGP synthase (pgsA, sll1522) [33] involved in PG biosynthesis have recently been identified. Unlike higher plants, cyanobacteria are capable of uptaking exogenously added PG [33] therefore they can serve as ideal models for studying the role and metabolism of PG in photosynthetic organisms. The identification of the genes of PG biosynthesis opened the way for generating the ΔpgsA [33] and ΔcdsA [32] mutants of Synechocystis PCC6803 that are incapable of synthesizing PG, although they can utilize exogenously added artificial PG through uptake from the culturing medium. Therefore, the physiological roles of PG and the biochemical pathways of their biosynthesis can be studied in vivo by these mutants. For example, the effect of PG molecules with various fatty acids has been investigated on photosynthetic and other cellular functions using these biosynthetic mutants [34,35]. The intracellular content of PG can be manipulated in PG-deficient mutant cells by withdrawing exogenously supplied PG. 21-day PG depletion of Synechocystis PCC6803/ΔpgsA results in severe PG deficiency. This system provides an excellent experimental tool for monitoring the in vivo remodeling process of artificial PG molecules. In this study, we investigated the retailoring of PG in ΔpgsA mutant cells. Retailoring is a special case of remodeling, in which an exogenously supplied, non-natural PG species is converted to physiologically optimal forms. Taking advantage of a mutant that is incapable of synthesizing PG we could study the formation of essential PG molecular species derived from dioleoyl-PG [PG(18:1/18:1)] as a precursor. PG(18:1/18:1) contains a single double bond that is the minimum requirement for cell survival [36]. This PG species was added to PG-depleted ΔpgsA cells and retailoring of its fatty acyl moieties was monitored by mass spectrometry. Based on the identification of various PG species as products of the retailoring process we propose a remodeling mechanism for PG in a photosynthetic prokaryote grown under normal environmental conditions, without manipulation of the enzymes involved in this remodeling process. 2. Materials and methods 2.1. Organism and growth condition Cells of wild-type and ΔpgsA mutant strain of Synechocystis PCC6803 were grown photoautotrophically in BG11 medium [37] supplemented with 5 mM HEPES–NaOH (pH 7.5). The mutant cells were cultivated in BG11 medium supplemented with 20 μg/mL kanamycin and in the presence of 20 μM PG(18:1/18:1) (Sigma, St. Louis, MO, USA). Cells were grown at 30 °C, under continuous
illumination at the intensity of 30 μmol photons m−2 s−1. Cultures were aerated on a gyratory shaker operating at 100 rpm. PG depletion was achieved by washing the cells twice with PG-free medium, followed by cultivation in PG-free medium. The cells can survive in PG-free medium only for a limited time, since the decrease of PG content abolishes photosynthetic activity and other cellular functions [38]. Following 21 days of PG-free cultivation, we re-added PG to the medium. Recovering mutant cells, grown in the presence of PG for 1 and 21 days, are indicated as ΔpgsA+PG 1d and ΔpgsA+PG 21d. 2.2. Lipid extraction and analysis For gas chromatographic analysis, lipids were extracted from the collected cells (around OD750 = 50) by the method of Bligh and Dyer [39]. The analysis of lipids was carried out according to the method of Sato and Murata [40]. Lipids were fractionated on precoated thin layer chromatography (TLC) plates (Merck 5721) developed with CHCl3/CH3OH/28% NH4OH (65:35:5, v/v). Fifty micrograms pentadecanoic acid (15:0), an internal standard, was added to the PG band of the TLC plate. After detection of the lipids by primuline fluorescence, the appropriate area of the silica gel (PG band) was scraped and was subjected to methanolysis with 5% HCl in methanol at 85 °C for 2 h. The resulting fatty acid methyl esters of PG were analyzed with a Hewlett Packard (Palo Alto, CA, USA) HP6890 gas chromatograph equipped with a hydrogen flameionization detector. Fatty acid methyl esters were separated on a 30 m × 0.25 mm i.d. SP-2330 capillary column (Supelco, Bellefonte, PA, USA). Temperatures of the column and the flame-ionization detector were 180 and 260 °C, respectively. The relative amounts of fatty acid methyl esters were determined by comparing the areas under the peaks on the chromatogram to the area of the internal fatty acid standard (15:0). For electrospray ionization (ESI) mass spectrometry, total lipids were isolated from intact cells of cyanobacteria (around OD750 = 50) with the method of Bligh and Dyer [39]. For semi-quantitative analyses, 20 μg PG(14:0/14:0) (Avanti Polar Lipids, Alabaster, USA) was added to each sample as internal standard before lipid extraction. 2.3. Mass spectrometry The extracts of the biological material or calibration solutions, spiked with the PG(14:0/14:0) internal standard (20 μg/mL), were diluted 100 times with water/acetonitrile mixture (50:50) before analysis. The samples (50 μL) were directly infused at a flow rate of 5 μL/min into a hybrid linear ion trap–Fourier transform ion cyclotron resonance (FTICR) mass spectrometer with a 7-Tesla superconducting magnet (LTQ-FT, Thermo Fisher, San Jose, CA) and equipped with an ESI source operated in the negative ionization mode. The spray voltage was set to 5 kV, sheath gas flow rate to 10 units, capillary temperature to 275 °C, and the ion optics elements voltages were set by the built-in automatic tuning procedure using a solution of the PG (18:1/18:1) standard. Singly negatively charged phospholipid [M-H]- molecular ions were detected by full-scan FTICR acquisition in the 150–2000 Th mass-to-charge ratio (m/z) range. Mass resolution (m/Δm) was set to 50000 at m/z of 400 Th. Approximately 80 mass spectra were averaged per sample. For further structural characterization, MS/MS experiments were performed. The [M-H]- ions were isolated and subjected to collision-induced dissociation (CID) in the linear ion trap that was also used to detect the product ions (MS/MS spectra). The isolation width was set to 1.0 Th and the collision energy to 15%, respectively, with helium as target gas. To estimate the concentration of PG species in the biological extracts, the following simple equation was used: cA = cIS = kA ðIA = IIS Þ
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
where cA is the concentration of analyte, cIS is the concentration of internal standard [PG(14:0/14:0)] and IA and IIS are the ion intensities of analyte and the internal standard, respectively. Calibration solutions containing 1, 2, 5, 10, 20 and 50 μg/mL of PG(18:1/18:1) or PG(16:0/18:1) (Avanti Polar Lipids, Alabaster, USA) and 20 μg/mL of the internal standard were analyzed by full-scan negative-ion ESI– FTICR mass spectrometry, and the cA/cIS ratios were plotted against the IA/IIS ratios. The kPG(18:1/18:1) and kPG(16:0/18:1) constants were determined as the slopes of linear regressions with zero intercept. The constants were then used to calculate PG(18:1/18:1) (m/z 773) and PG(18:1/16:0) (m/z 747) concentrations from the observed IA/IIS ratios in the full-scan negative-ion ESI–FTICR mass spectra of biological samples. To estimate the concentration of PG (18:2/18:1) at m/z 771 and PG(18:2/16:0) at m/z 745, we used the average of kPG(18:1/18:1) and kPG(16:0/18:1) determined. 3. Results 3.1. Gas chromatographic analyses revealed remodeling of PG fatty acids PG of Synechocystis PCC6803/ΔpgsA mutant was separated and the fatty acid composition was measured by gas chromatography according to Hagio et al. [33]. The results are summarized in Table 1. The PG-deficient mutant was grown in the presence of exogenously added PG(18:1/18:1). After 1 day of culturing of the mutant cells in the presence of PG(18:1/18:1), the most abundant fatty acid species was oleic acid. In the ΔpgsA+PG 21d sample, oleic acid content decreased from 60% to 11%. At the expense of oleic acid, palmitic acid content increased from 19% to 57%, and polyunsaturated fatty acids were present in much higher amounts, as compared to ΔpgsA+PG 1d sample. The estimation of fatty acid composition of the PG in wild-type cells was done by Murata et al. [13]. This estimation was based on the fatty acid composition determined with Rhizopus phospholipase A2 (PLA2). The most abundant PG species were those which contained mainly unsaturated fatty acids at the sn-1 position (18:2/16:0; 18:1/ 16:0), whereas the ones with saturated fatty acids at the same position (18:0/16:0) were less abundant, in agreement with the results published by Okazaki and coworkers [14]. 3.2. Mass spectrometry confirms the presence of newly formed PGs The biochemical modification of artificial PG was monitored by mass spectrometry. Fig. 1 shows the negative-ion ESI–FTICR mass spectra of dioleoyl-PG (Fig. 1a) and those of the lipid extracts of mutant cells following 1 day (Fig. 1b), 12 days (Fig. 1c), 18 days (Fig. 1d) and 21 days (Fig. 1e) of culturing after the re-addition of dioleoylPG. Since after 21 days of PG depletion the mutant cells did not contain accurately detectable levels of PG, the first time point of the analysis was 1 day following re-addition. The monoisotopic [M-H]ion of dioleoyl-PG is detected at m/z 773.5339 (Fig. 1a). Apparently, several new deprotonated [M-H]- PG species (at m/z 771.5159,
Table 1 Fatty acid composition (mol%) of PG from Synechocystis PCC6803/ΔpgsA mutant cells. Strain
ΔpgsA+PG 1d ΔpgsA+PG 21d
Acyl chains in PG molecules (mol%) 16:0
16:1
18:0
18:1
18:2
18:3
19 57
2 4
10 20
60 11
7 8
2 n.d.
Synechocystis PCC6803/ΔpgsA cells were grown in PG-supplemented medium for 1 or 21 days (ΔpgsA+PG 1d and ΔpgsA+PG 21d, respectively). 16:0, hexadecanoic acid (palmitic acid); 16:1, hexadecenoic acid (palmitoleic acid); 18:0, octadecanoic acid (stearic acid); 18:1, 9-octadecenoic acid (oleic acid); 18:2, 9,12-octadecadienoic acid (linoleic acid); 18:3, 9,12,15-octadecatrienoic acid (α-linolenic acid); n.d., not detected.
165
747.5164 and 745.5004) appeared after 12 days of culturing in the presence of the added synthetic PG (Fig. 1c). Accurate m/z values (Fig. 1e) of the molecular ions observed in the ESI-FTICR mass spectra (Table 2) permitted the verification of the composition of these new PG forms and, consequently, guided the identification of their accurate chemical structure. In all cases, the mass accuracy was within 5 ppm. Following 12-day culturing (Fig. 1d and e) new PG molecular species did not emerge, however, the amount of the newly formed PG species continuously increased during the 21-day culturing period with concomitant decrease in the dioleoyl-PG content (Fig. 1e). Structural identification of artificial and reformed PG molecules was performed by low-energy CID-MS/MS of [M-H]- ions observed in the full-scan mass spectrum displayed in Fig. 1e (ΔpgsA+PG 21d sample). Also relying upon earlier results implicating the preference of acyl-lipid desaturases to introduce double bonds into the sn-1 fatty acid of PG in cyanobacteria [18,19], we could infer fatty acyls bound to the sn-1 and sn-2 positions by detailed MS/MS analyses using the results of earlier studies by Hsu and Turk [41] on CID of [M-H]- ions of various synthetic PG species. The features they identified permitted the reliable assignments of the fatty acid substituents and their position in the glycerol backbone. In particular, we deduced structural information from major fragment ions that were characteristic for the individual PG molecules and arose from a neutral loss of free fatty acid substituents [M-HRxCO2H]- or [M-H-R'xCH = C = O]- ketenes, by consecutive losses of the fatty acid and the glycerol head group [M-HRxCOOH-C3H6O2]-, and by formation of fatty carboxylate anions [RxCO-2], where x = 1, 2 and Rx = R'xCH2. We also utilized the feature that the abundances of the ions arising from neutral loss of the sn-2 substituent as a free fatty acid [M-H-R2CO2H]- or as a ketene [M-H-R'2CH = C = O]- were generally greater than those of the product ions from the analogous losses at sn-1 [41]. This was due to the nucleophilic attack of the anionic phosphate site on the C-1 or the C-2 of the glycerol, to which the carboxylates attached, to expel sn-1 (R1CO-2) or the sn-2 (R2CO-2) carboxylate anion with preference to R2COO- over R1COO[41]. Identification of the major isomers by this rule of Hsu and Turk [41] did not, however, exclude the presence of the other isobaric isomers in small amounts. Fig. 2a shows the MS/MS spectrum of the added synthetic PG(18:1/18:1) as reference. We indeed observed fragment ions at m/z 491, 509 and 417, reflecting neutral loss of the 18:1 fatty acyl substituent as an acid, neutral loss of a ketene, and consecutive losses of the fatty acid and glycerol head group, respectively. A single carboxylate anion (m/z 281) appeared, because the fatty acyl constituent was the same in both the sn-1 and sn-2 positions of PG(18:1/18:1). 3.3. Identification of PG(18:2/18:0) The product–ion spectrum of the m/z 773 [M-H]- ions of the lipid extract obtained 21 days after PG(18:1/18:1) re-addition is shown in Fig. 2b. In addition to the fragment ions observed for PG (18:1/18:1), the MS/MS spectrum also contained additional ions at m/z 493 and 489 reflecting neutral losses of 18:2 and 18:0 fatty acids, respectively. Ions at m/z 419 and 415 were derived from the consecutive losses of fatty acid and the glycerol head group. The ions at m/z 511 and 507 corresponded to neutral losses as ketenes. The carboxylate anions were m/z 279 and 283, respectively. The spectrum showed that ions m/z 489, 415, 507 and 283 were more abundant than ions at m/z 493, 419, 511 and 279. These fragment ions confirmed the presence of isomeric PG species, and the intensities suggested that the 18:2 fatty acid residues were located predominantly in the sn-1 position, while the 18:0 residues were predominantly in the sn-2 position. Thus the MS/MS spectrum also indicated presence of PG(18:2/18:0), in addition to the added artificial PG(18:1/18:1).
166
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
Fig. 1. The “phosphatidylglycerol range” (showing the [M-H]- ions of this class of lipids) in the negative-ion ESI-FTICR mass spectra of (a) artificial synthetic PG(18:1/18:1) (b), ΔpgsA+PG 1d (c), ΔpgsA+PG 12d (d), ΔpgsA+PG 18d and (e), ΔpgsA+PG 21d lipid extracts.
3.4. Identification of PG(18:2/18:1) and its isoform, PG(18:3/18:0) The product–ion spectrum of the [M-H]- ions of m/z 771 (Fig. S1) showed ions at m/z 491 and 489 reflecting neutral loss of 18:2 and 18:1 fatty acids, respectively. The m/z 417, 415 ions arose from the consecutive losses of fatty acid and the glycerol head group. The spectrum also contained ions at m/z 509 and 507, reflecting neutral loss of the fatty-acyl substituents as ketenes. The abundance of the m/z 489 ion was greater than that of the m/z 491 ion, and the abundance of the m/z 415 ion was higher than that of 417. The abundance of the m/z 507 ion was also greater than that of the m/z 509 ion. The abundance of the carboxylate anion at m/z 281 was more abundant than the
corresponding ion arising from the similar loss at m/z 279. From the fragment ion intensities we inferred that the 18:2 and 18:1 fatty acids were esterified to the sn-1 and the sn-2 positions of the glycerol backbone, respectively. In this MS/MS spectrum, obtained from m/z 771 as precursor, we also identified other ions of low intensity. However, the abundances at m/z 487, 413, 505 and 283 were clearly greater than those of m/z 493, 419, 511 and 277. Therefore, we attributed these fragment ions to losses of 18:3 and 18:0 fatty acyl groups from the sn-1 and sn-2 positions, respectively. Altogether, the MS/MS spectrum of m/z 771 (Fig. S1) indicated the presence of both PG(18:2/18:1) and PG(18:3/18:0). 3.5. Identification of PG(18:1/16:0) and its isoform PG(16:1/18:0)
Table 2 Molecular ions (m/z) of putative PG species. Theoretical (m/z)
Detected (m/z)
Error (ppm)
Number of carbon (C) atoms
Number of desaturation
Putative PG species (sn-1/sn-2)
773.5338
773.5325
−1.7
C42
2
771.5201
771.5164
−4.8
C42
3
747.5182
747.5167
−1.9
C40
1
745.5025
745.5007
−2.4
C40
2
18:1/18:1 18:2/18:0 18:2/18:1 18:3/18:0 18:1/16:0 16:0/18:1 16:1/18:0 18:2/16:0 16:1/18:1
Italics indicate the most abundant PG molecular species for the given m/z.
Fig. S2 shows the MS/MS spectrum of the [M-H]- ions of m/z 747. Ions at m/z 465 and 491, reflecting neutral losses of 18:1 and 16:1 fatty acids, were produced. The m/z 391 and 417 ions arose from the consecutive losses of fatty acid and the glycerol head group. Other abundant ions at m/z 483 and 509 reflect neutral loss of the fatty acyl substituent as a ketene. We observed that the abundance of the m/z 491 ion was greater than that of the m/z 465 ion, the abundance of the m/z 417 ion was higher than that of the m/z 391 ion, and the abundance of the m/z 509 ion is also greater than that of the m/z 483 ion. The carboxylate anion at m/z 255 was less abundant than the corresponding ion arising from the similar loss at m/z 281. The relative abundance of these fragment ions indicated that the 18:1 and 16:0 fatty acids were derived from the sn-1 and sn-2 positions, respectively. In this m/z 747 ion spectrum we could also identify
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
167
Fig. 2. The CID product–ion spectra of (a), [M-H]- of artificial PG(18:1/18:1) standard (precursor: m/z 773), (b), [M-H]- of PG(18:1/18:1) and its isomer PG(18:2/18:0) (precursor: m/z 773) extracted from ΔpgsA+PG 21 d.
other ions, albeit with very low abundance. Ion abundances from the loss of fatty-acyl constituents at m/z 463, 389, 481 and 283 were greater than those at m/z 493, 419, 511 and 253, which reflected the loss of 16:1 and 18:0 fatty-acyl constituents from the sn-1 and sn-2 positions of the glycerol backbone, respectively. Therefore, the MS/ MS spectrum of m/z 747 (Fig. S2) indicated the presence of both PG (18:1/16:0) and PG(16:1/18:0). 3.6. Identification of PG (18:2/16:0) and its isoform PG(16:1/18:1) When the parent [M-H]- ions at m/z 745 were fragmented, the product ions permitted the identification and position assignment of fatty-acyl groups attached to the glycerol backbone, and isomers such as PG(18:2/16:0) and PG(16:1/18:1) could be easily distinguished (Fig. S3). MS/MS product ions at m/z 465 and 489 reflected neutral loss of 18:2 and 16:0 fatty acids. The m/z 391, 415 product ions arose from the consecutive losses of fatty acid and the glycerol head group. Other abundant ions at m/z 483 and 507 reflected neutral losses of the fatty-acyl substituents as ketenes. The abundance of the m/z 489 ion was greater than that of the m/z 465 ion, the abundance of the m/z 415 ion was higher than that of m/z 391. The abundance of the m/z 507 ion was also greater than that of the m/z 483 ion. The carboxylate anion at m/z 255 was more abundant than the corresponding ion
arising from a similar fragmentation process yielding m/z 279 suggesting that 18:2 fatty acid was esterified to the sn-1, while 16:0 fatty acid to the sn-2 position, corresponding to PG(18:2/16:0). In this MS/MS spectrum from m/z 745 as the precursor ion (Fig. S3), we also noticed other ions corresponding to the loss of 16:1 and 18:1 fattyacyl substituents. The relative abundances of the ions at m/z 463, 389, 481, and 281 were greater than those derived from ions at m/z 491, 509, 417 and 253, respectively. Therefore, they indicated the presence of both PG(18:2/16:0) and PG(16:1/18:1). 3.7. Semi-quantitative mass spectrometric analysis of PG molecular species To follow the dynamics of PG retailoring, we made semiquantitative analyses of PG molecular species. We used an internal standard, PG(14:0/14:0), because it was not a natural phospholipid and, moreover, its molecular ion (m/z 665) did not overlap with those of the naturally occurring molecular species in the negative-ion ESI mass spectra. Relative ion intensities obtained from full-scan FTICR measurements were used for quantitation, therefore, we could not distinguish isobaric PG isomers. Based on comparing the relative ion intensities of the m/z 773, 771, 747 and 745 ions, we observed a gradual decrease of PG(18:1/18:1) (m/z 773) during the culturing
168
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
period (1, 14 and 21 days) (Fig. 3a). At the same time, the concentration of other PG species at m/z 771 (Fig. 3b), 747 (Fig. 3c) and 745 (Fig. 3d) increased at the expense of m/z 773, indicating remodeling of the PG species. The concentrations of PG species were also estimated using calibration coefficients. PG(18:1/18:1) decreased from 1.77 to 0.77 μM, together with simultaneous increase in the concentration of the other three most abundant PG molecular species (Table 3). The tendency was similar to that shown in Fig. 3. 4. Discussion GC analyses indicated retailoring of PG(18:1/18:1) in Synechocystis PCC6803/ΔpgsA, in agreement with our previous results obtained with the PAL/ΔcdsA mutant [42]. This exogenous PG was reformed by the cyanobacteria according to their functional needs. In cells, natural PG molecular species were synthesized which provided essential PG molecules for cellular functions. Our earlier studies highlighted the importance of PG(18:2/16:0), revealing that this molecular species is needed for resistance to low temperature photoinhibition [24,43]. This can explain why the remodeling proceeds mainly to the production of PG(18:2/16:0). Whereas our earlier GC results suggested remodeling, accurate and site-specific information has been provided by the mass spectrometric analyses performed. For the reliability of our results it was important verifying that the Synechocystis PCC6803/ΔpgsA cells remained fully viable during the 21-day PG depletion period. Earlier we demonstrated that, following PG re-addition, the photosynthetic activity of the cultures fully recovered from the PG starvation-related decrease [33,44]. Furthermore, the viability of the cells was confirmed in each stage of the experiments by determining the numbers of colony-forming units (data not shown). The results of mass spectrometric analyses, together with earlier evidence about the preference of acyl-lipid desaturases to introduce double bonds into the sn-1 fatty acid of PG in cyanobacteria [18,19], suggested the biosynthetic scheme shown in Fig. 4a. Starting from the
Table 3 Temporal changes in molar concentrations of PG species. Strain
Molecular species 18:1/16:0 16:0/18:1 16:1/18:0
18:2/16:0 16:1/18:1
Molecular ion (m/z) 773 771
747
745
Molar concentration (μM) 1.77 0.12 0.77 1.47
0.17 0.67
n.d. 0.88
18:1/18:1 18:2/18:0
ΔpgsA+PG 1d ΔpgsA+PG 21d
18:2/18:1 18:3/18:0
Samples are denoted as in Table 1. Italics indicate the most abundant PG molecular species; n.d., not detected.
added PG(18:1/18:1), the first possible enzymatic reaction is 1 (indicated with solid arrow), which involves an acyl-lipid desaturation that results in the production of linoleoyl from oleoyl residue. The acyl-lipid desaturase (Fig. 4b) that catalyses this reaction is a wellcharacterized enzyme [18,19]. In cyanobacteria lipid desaturation is carried out exclusively by acyl-lipid desaturases [45]. Routes 2, 3, 4 and 8 (Fig. 4a) (open arrows) represent synthetic routes that require at least two separate enzyme reactions. The first step is the cleavage of fatty acids, either at sn-2 or sn-1 positions, which need PLA2 or phospholipase A1 (PLA1), respectively. An alternative biosynthetic route could be mediated by acyl-hydrolases [46,47]. These processes result in the formation of PG species containing fatty acids other than 18:1, such as 16:0 and 18:0. For the enzyme processes in 3 and 4 the cells need PLA2 (Fig. 4b) that can eliminate an acyl group from the sn2 position. Reactions 2 and 8 requires PLA1 (Fig. 4b) that can cleave the ester bound at the sn-1 position. Acyl-hydrolases can remove fatty acids from both the sn-1 and sn-2 positions [48]. In Synechocystis PCC6803 phospholipases, acyl-hydrolases and their genes have not yet been identified. Sequence homology search could not provide positive results, suggesting that if there are phospholipases and acylhydrolases in Synechocystis PCC6803, these are substantially different
Fig. 3. Analyses of PG species, changes in ion intensities: (a), PG(18:1/18:1) (m/z 773); (b), PG(18:2/18:1), (m/z 771); (c), PG(18:1/16:0) (m/z 747); and (d), PG(18:2/16:0) (m/z 745), this molecular species was not detectable after one day re-addition of PG.
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
169
Fig. 4. Putative biosynthetic scheme for the remodeling of PG in Synechocystis PCC6803/ΔpgsA mutant cells. (a) Numbers indicate enzymatic reactions, bold characters denote reformed fatty acid moieties. The chemical reactions, replacing and desaturating fatty acids, are indicated by open and solid arrows, respectively. (b) “'” and “⁎” symbols indicate the replacement and desaturation of fatty acids at the sn-1 and sn-2 positions, respectively. Reformed fatty acids and known biosynthetic enzymes are highlighted by bold characters.
from those known from higher plants. Acyltransferase enzymes (Fig. 4b) with specific acyl-carrier characteristics are required for new fatty acids esterified to PG. Lysoglycerophospholipid acyltransferase (LGP acyltransferase), a mainly sn-2-specific 16:0 acyltransferase, has been identified in cyanobacteria [15], which explains our finding that one of the main products of remodeling is PG(18:1/16:0) that could be desaturated further to PG(18:2/16:0) (Fig. 4a, route 3). The replacement of fatty acids is followed by desaturation (Fig. 4a, routes 5, 6, 7 and 9) (bold arrows). As a possible alternative mechanism (Fig. 4a, route 2) we could mention the exchange of acyl groups of PG with those of other lipids. In cyanobacteria there are acyltransferases, such as acyl-ACP: lyso-MGDG acyltransferase [49], therefore, 18:2 can be also transferred from MGDG to the sn-1 position of lyso-PG synthesized by acyl-hydrolase or PLA1 (Fig. 4b). One day after the re-addition of PG(18:1/18:1), the most abundant molecular species was still the synthetic lipid, representing more than 90% of the overall PG content. However, some minor PG derivatives containing 18:2 and 16:0 acyl chains could also be observed as low-abundance components, and were retained after PG depletion, when the cells needed specific residual PG for survival. The products of desaturation emerged 1 day after the re-addition of dioleoyl-PG. After 2 weeks of culturing in the presence of dioleoyl-PG, more than 50% of total PG contained 18:2, which was the most abundant polyunsaturated fatty acid. The 18:2 was esterified to the sn-1 position. Twenty-one days after re-addition of dioleoyl-PG, a rather significant accumulation of 16:0 occurred. This fatty acid esterified mainly the sn-2 position of PG. The above enzymatic reactions (Fig. 4b) are capable of remodeling lipid molecules, providing an important regulatory machinery for the adaptation to environmental changes [2,7]. The proposed biosynthetic scheme is based on the known processes of lipid synthesis in cyanobacteria and higher plants [35,50,51] although our results suggest the presence of yet unknown synthetic steps in cyanobacteria. In conclusion, in vivo retailoring of exogenously added dioleoyl-PG was found to produce several native PG species that are required for the survival of Synechocystis PCC6803/ΔpgsA cells. We propose that PG retailoring in Synechocystis PCC6803 is mediated by PLA1, PLA2 and acyl-hydrolases which are necessary to remove fatty acids from the glycerol backbone, lipid acyltransferases needed for constructing new fatty acid residues esterified to the glycerol backbone, and acyllipid desaturases that create the physiologically required lipid
molecules. Our data provide evidence that cyanobacteria possess the enzymatic machinery capable of PG remodeling. Acknowledgements This research was supported by the Hungarian Science Foundation (OTKA: grants K 60109 and K 68692 to Z.G.). Laszlo Prokai is the Robert A. Welch Professor of the University of North Texas Health Science Center (grant BK-0031). P.F. thanks the Ministry of Education of the Czech Republic (MSM6198959216). The authors are grateful to Anna Sallai for her technical assistance. We are thankful to Dr. Miklós Szekeres (Institute of Plant Biology, Biological Research Center of the Hungarian Academy of Sciences, Szeged) for critical reading of the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbalip.2009.10.009. References [1] J.L. Harwood, A.L. Jones, H.J. Perry, A.J. Rutter, K.L. Smith, M. Williams, Changes in plant lipids during temperature adaptation, in: A.R. Cossins (Ed.), Temperature Adaptation of Biological Membranes, Portland Press, London, 1994, pp. 107–118. [2] H. Wada, N. Murata, Temperature-induced changes in the fatty-acid composition of the cyanobacterium, Synechocystis PCC6803, Plant Physiol. 92 (1990) 1062–1069. [3] J.L. Harwood, Involvement of chloroplast lipids in the reaction of plants submitted to stress, in: P.A. Siegenthaler, N. Murata (Eds.), Advances in Photosynthesis, Lipids in Photosynthesis: Structure, Function and Genetics, vol. 6, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998, pp. 287–302. [4] P.A. Siegenthaler, Molecular organization of acyl lipids in photosynthetic membranes of higher plants, in: P.A. Siegenthaler, N. Murata (Eds.), Advances in Photosynthesis, Lipids in Photosynthesis: Structure, function and genetics, vol. 6, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998, pp. 119–144. [5] N. Murata, I. Nishida, Lipids in blue-green algae (cyanobacteria), in: P.K. Stumpf (Ed.), The Biochemistry of Plants, Academic Press, Orlando, 1987, pp. 315–347. [6] N. Sato, N. Murata, Y. Miura, N. Ueta, Effect of growth temperature on lipid and fatty-acid compositions in the blue-green-algae, Anabaena variabilis and Anacystis nidulans, Biochim. Biophys. Acta 572 (1979) 19–28. [7] N. Murata, N. Sato, N. Takahashi, Y. Hamazaki, Compositions and positional distributions of fatty-acids in phospholipids from leaves of chilling-sensitive and chilling-resistant plants, Plant Cell Physiol. 23 (1982) 1071–1079.
170
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
[8] N. Murata, Molecular-species composition of phosphatidylglycerols from chillingsensitive and chilling-resistant plants, Plant Cell Physiol. 24 (1983) 81–86. [9] P.G. Roughan, Phosphatidylglycerol and chilling sensitivity in plants, Plant Physiol. 77 (1985) 740–746. [10] Y. Tasaka, I. Nishida, S. Higashi, T. Beppu, N. Murata, Fatty-acid composition of phosphatidylglycerols in relation to chilling sensitivity of woody-plants, Plant Cell Physiol. 31 (1990) 545–550. [11] A. Ohta, K. Waggoner, A. Radominskapyrek, W. Dowhan, Cloning of genes involved in membrane lipid synthesis—effects of amplification of phosphatidylglycerophosphate synthase in Escherichia coli, J. Bacteriol. 147 (1981) 552–562. [12] H. Wada, N. Murata, Membrane lipids in cyanobacteria, in: P.A. Siegenthaler, N. Murata (Eds.), Advances in Photosynthesis, Lipids in Photosynthesis: Structure, Function and Genetics, vol. 6, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998, pp. 65–81. [13] N. Murata, H. Wada, Z. Gombos, Modes of fatty-acid desaturation in cyanobacteria, Plant Cell Physiol. 33 (1992) 933–941. [14] K. Okazaki, N. Sato, N. Tsuji, M. Tsuzuki, I. Nishida, The significance of C16 fatty acids in the sn-2 positions of glycerolipids in the photosynthetic growth of Synechocystis sp PCC6803, Plant Physiol. 141 (2006) 546–556. [15] D. Weier, C. Muller, C. Gaspers, M. Frentzen, Characterisation of acyltransferases from Synechocystis sp PCC6803, Biochem. Biophys. Res. Commun. 334 (2005) 1127–1134. [16] N. Sato, N. Murata, Lipid biosynthesis in the blue-green-alga, Anabaena variabilis. 1. Lipid Classes, Biochim. Biophys. Acta 710 (1982) 271–278. [17] N. Sato, N. Murata, Lipid biosynthesis in the blue-green-alga, Anabaena variabilis. 2. Fatty-acids and lipid molecular-species, Biochim. Biophys. Acta 710 (1982) 279–289. [18] N. Murata, H. Wada, Acyl-lipid desaturases and their importance in the tolerance and acclimatization to cold of cyanobacteria, Biochem. J. 308 (1995) 1–8. [19] H. Wada, N. Murata, Synechocystis PCC6803 mutants defective in desaturation of fatty-acids, Plant Cell Physiol. 30 (1989) 971–978. [20] N. Murata, Q. Ishizakinishizawa, S. Higashi, H. Hayashi, Y. Tasaka, I. Nishida, Genetically engineered alteration in the chilling sensitivity of plants, Nature 356 (1992) 710–713. [21] T. Sakamoto, D.A. Los, S. Higashi, H. Wada, I. Nishida, M. Ohmori, N. Murata, Cloning of omega-3 desaturase from cyanobacteria and its use in altering the degree of membrane-lipid unsaturation, Plant Mol. Biol. 26 (1994) 249–263. [22] H. Wada, Z. Gombos, T. Sakamoto, N. Murata, Genetic manipulation of the extent of desaturation of fatty-acids in membrane-lipids in the cyanobacterium Synechocystis PCC6803, Plant Cell Physiol. 33 (1992) 535–540. [23] F.P. Wolter, R. Schmidt, E. Heinz, Chilling sensitivity of Arabidopsis-thaliana with genetically engineered membrane-lipids, EMBO J. 11 (1992) 4685–4692. [24] Y. Tasaka, Z. Gombos, Y. Nishiyama, P. Mohanty, T. Ohba, K. Ohki, N. Murata, Targeted mutagenesis of acyl-lipid desaturases in Synechocystis: evidence for the important roles of polyunsaturated membrane lipids in growth, respiration and photosynthesis, EMBO J. 15 (1996) 6416–6425. [25] Z. Gombos, E. Kanervo, N. Tsvetkova, T. Sakamoto, E.M. Aro, N. Murata, Genetic enhancement of the ability to tolerate photoinhibition by introduction of unsaturated bonds into membrane glycerolipids, Plant Physiol. 115 (1997) 551–559. [26] N. Murata, Low-temperature effects on cyanobacterial membranes, J. Bioenerg. Biomembr. 21 (1989) 61–75. [27] D.V. Lynch, G.A. Thompson, Retailored lipid molecular-species—a tactical mechanism for modulating membrane-properties, Trends Biochem. Sci. 9 (1984) 442–445. [28] D.V. Lynch, G.A. Thompson, Microsomal phospholipid molecular-species alterations during low-temperature acclimation in Dunaliella, Plant Physiol. 74 (1984) 193–197. [29] D.V. Lynch, G.A. Thompson, Chloroplast phospholipid molecular-species alterations during low-temperature acclimation in Dunaliella, Plant Physiol. 74 (1984) 198–203.
[30] I. Nishida, N. Murata, Chilling sensitivity in plants and cyanobacteria: the crucial contribution of membrane lipids, Annu. Rev. Plant Physiol. Plant Mol. Biol. 47 (1996) 541–568. [31] P.G. Roughan, C.R. Slack, Cellular organization of glycerolipid metabolism, Annu. Rev. Plant Physiol. Plant Mol. Biol. 33 (1982) 97–132. [32] N. Sato, M. Hagio, H. Wada, M. Tsuzuki, Requirement of phosphatidylglycerol for photosynthetic function in thylakoid membranes, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 10655–10660. [33] M. Hagio, Z. Gombos, Z. Varkonyi, K. Masamoto, N. Sato, M. Tsuzuki, H. Wada, Direct evidence for requirement of phosphatidylglycerol in photosystem II of photosynthesis, Plant Physiol. 124 (2000) 795–804. [34] I. Domonkos, H. Laczkó-Dobos, Z. Gombos, Lipid-assisted protein-protein interactions that support photosynthetic and other cellular activities, Prog. Lipid Res. 47 (2008) 422–435. [35] H. Wada, N. Murata, The essential role of phosphatidylglycerol in photosynthesis, Photosynth. Res. 92 (2007) 205–215. [36] I. Sakurai, M. Hagio, Z. Gombos, T. Tyystjarvi, V. Paakkarinen, E.M. Aro, H. Wada, Requirement of phosphatidylglycerol for maintenance of photosynthetic machinery, Plant Physiol. 133 (2003) 1376–1384. [37] M.M. Allen, Simple conditions for growth of unicellular blue-green algae on plates, J. Phycol. 4 (1968) 1–4. [38] I. Domonkos, P. Malec, A. Sallai, L. Kovacs, K. Itoh, G.Z. Shen, B. Ughy, B. Bogos, I. Sakurai, M. Kis, K. Strzalka, H. Wada, S. Itoh, T. Farkas, Z. Gombos, Phosphatidylglycerol is essential for oligomerization of photosystem I reaction center, Plant Physiol. 134 (2004) 1471–1478. [39] E.G. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. J. Biochem. Physiol. 37 (1959) 911–917. [40] N. Sato, N. Murata, Membrane lipids, Methods Enzymol. 167 (1988) 251–259. [41] F.F. Hsu, J. Turk, Studies on phosphatidylglycerol with triple quadrupole tandem mass spectrometry with electrospray ionization: fragmentation processes and structural characterization, J. Am. Soc. Mass Spectrom. 12 (2001) 1036–1043. [42] H. Laczko-Dobos, B. Ughy, S.Z. Toth, J. Kornenda, O. Zsiros, I. Domonkos, A. Parducz, B. Bogos, M. Komura, S. Itoh, Z. Gombos, Role of phosphatidylglycerol in the function and assembly of Photosystem II reaction center, studied in a cdsAinactivated PAL mutant strain of Synechocystis sp. PCC6803 that lacks phycobilisomes, Biochim. Biophys. Acta 1777 (2008) 1184–1194. [43] Z. Gombos, H. Wada, N. Murata, The recovery of photosynthesis from lowtemperature photoinhibition is accelerated by the unsaturation of membranelipids—a mechanism of chilling tolerance, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 8787–8791. [44] Z. Gombos, Z. Varkonyi, M. Hagio, M. Iwaki, L. Kovacs, K. Masamoto, S. Itoh, H. Wada, Phosphatidylglycerol requirement for the function of electron acceptor plastoquinone Q(B) in the photosystem II reaction center, Biochemistry 41 (2002) 3796–3802. [45] S. Higashi, N. Murata, An in vivo study of substrate specificities of acyl-lipid desaturases and acyltransferases in lipid synthesis in Synechocystis PCC6803, Plant Physiol. 102 (1993) 1275–1278. [46] A.R. Matos, A.T. Pham-Thi, Lipid deacylating enzymes in plants: old activities, new genes, Plan Physiol. Biochem. 47 (2009) 491–503. [47] G.A. Mignery, C.S. Pikaard, D.J. Hannapel, W.D. Park, Isolation and sequenceanalysis of cDNAs for the major potato-tuber protein, patatin, Nucleic Acids Res. 12 (1984) 7987–8000. [48] X.M. Wang, Lipid signaling, Curr. Opin. Plant Biol. 7 (2004) 329–336. [49] H.H. Chen, A. Wickrema, J.G. Jaworski, Acyl-acyl-carrier protein: lysomonogalactosyldiacylglycerol acyltransferase from the cyanobacterium Anabaena variabilis, Biochim. Biophys. Acta 963 (1988) 493–500. [50] X. Wang, Plant phospholipases, Annu. Rev. Plant Physiol. Plant Mol. Biol. 52 (2001) 211–231. [51] W. Yang, S.P. Devaiah, X. Pan, G. Isaac, R. Welti, X. Wang, AtPLAI is an acyl hydrolase involved in basal jasmonic acid production and Arabidopsis resistance to Botrytis cinerea, J. Biol. Chem. 282 (2007) 18116–18128.
Contents lists available at ScienceDirect
Biochimica et Biophysica Acta j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b b a l i p
Remodeling of phosphatidylglycerol in Synechocystis PCC6803 Hajnalka Laczko-Dobos a, Petr Fryčák b,1, Bettina Ughy a, Ildiko Domonkos a, Hajime Wada c, Laszlo Prokai b, Zoltan Gombos a,⁎ a b c
Institute of Plant Biology, Biological Research Center, Hungarian Academy of Sciences, H-6701 Szeged, P.O. Box 521, Hungary Department of Molecular Biology and Immunology, University of North Texas Health Science Center, 3500 Camp Bowie Boulevard, Fort Worth, TX 76107, USA Department of Life Sciences, Graduate School of Arts and Sciences, University of Tokyo, Komaba, Tokyo 153-8902, Japan
a r t i c l e
i n f o
Article history: Received 9 July 2009 Received in revised form 16 October 2009 Accepted 19 October 2009 Available online 24 October 2009 Keywords: Phosphatidylglycerol Remodeling Membrane lipid Photosynthetic organism Mass spectrometry Synechocystis PCC6803
a b s t r a c t The phosphatidylglycerol deficient ΔpgsA mutant of Synechocystis PCC6803 provided a unique experimental system for investigating in vivo retailoring of exogenously added dioleoylphosphatidylglycerol in phosphatidylglycerol-depleted cells. Gas chromatographic analysis of fatty acid composition suggested that diacyl-phosphatidylglycerols were synthesized from the artificial synthetic precursor. The formation of new, retailored lipid species was confirmed by negative-ion electrospray ionization–Fourier-transform ion cyclotron resonance and ion trap tandem mass spectrometry. Various isomeric diacyl-phosphatidylglycerols were identified indicating transesterification of the exogenously added dioleoylphosphatidyl-glycerol at the sn-1 or sn-2 positions. Polyunsaturated fatty acids were incorporated selectively into the sn-1 position. Our experiments with Synechocystis PCC6803/ΔpgsA mutant cells demonstrated lipid remodeling in a prokaryotic photosynthetic bacterium. Our data suggest that the remodeling of diacylphosphatidylglycerol likely involves reactions catalyzed by phospholipase A1 and A2 or acyl-hydrolase, lysophosphatidylglycerol acyltransferase and acyl-lipid desaturases. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Remodeling of lipids by manipulating fatty acid composition is one of potential factors involved in the adaptation to environmental stresses [1,2]. However, this metabolic process has not been well understood. Both prokaryotes and eukaryotes respond to various environmental stresses [3] by altering the fatty acid composition of
Abbreviations: PG, phoshatidylglycerol; PG(18:1/18:1), 1,2-di-(9E-octadecenoyl)sn-glycero-3-phospho-(1′-rac-glycerol); PG(14:0,14:0), 1,2-dimyristoyl-sn-glycero-3phospho-(1′-rac-glycerol); PG(18:2/18:0), 1-(9Z,12Z-octadecadienoyl)-2-octadecanoyl-sn-glycero-3-phospho-(1′-rac-glycerol); PG(18:2/18:1), 1-(9Z,12Z-octadecadienoyl)-2-(9E-octadecenoyl)-sn-glycero-3-phospho-(1′-rac-glycerol); PG(18:3/18:0), 1-(9Z,12Z,15Z-octadecatrienoyl)-2-octadecanoyl-sn-glycero-3-phospho-(1′-rac-glycerol); PG(18:1/16:0), 1-(9Z-octadecenoyl)-2-hexadecanoyl-sn-glycero-3-phospho(1′-sn-glycerol); PG(16:0/18:1), 1-hexadecanoyl-2-(9Z-octadecenoyl)-sn-glycero-3phospho-(1′-rac-glycerol); PG(16:1/18:0), 1-(9Z-hexadecenoyl)-2-octadecanoyl-snglycero-3-phospho-(1′-rac-glycerol); PG(18:2/16:0), 1-(9Z,12Z-octadecadienoyl)-2hexadecanoyl-sn-glycero-3-phospho-(1′-rac-glycerol); PG(16:1/18:1), 1-(9Z-hexadecenoyl)-2-(9Z-octadecenoyl)-sn-glycero-3-phospho-(1′-rac-glycerol); 16:0, hexadecanoic acid (palmitic acid); 16:1, hexadecenoic acid (palmitoleic acid); 18:0, octadecanoic acid (stearic acid); 18:1, 9-octadecenoic acid (oleic acid); 18:2, 9,12-octadecadienoic acid (linoleic acid); 18:3, 9,12,15-octadecatrienoic acid (α-linolenic acid); ΔpgsA+PG 1d and 21 d indicate mutant cells grown in the presence of PG for 1 and 21 days, respectively ⁎ Corresponding author. Tel.: +3662599704; fax: +3662433434. E-mail address: [email protected] (Z. Gombos). 1 Present address: Department of Analytical Chemistry, Faculty of Science, Palacky University in Olomouc, Tr. 17. listopadu 12, 77146 Olomouc, Czech Republic. 1388-1981/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.bbalip.2009.10.009
their membrane lipids. The extent of unsaturation of fatty acids in thylakoid membranes increases and/or the chain length of fatty acids decreases according to growth temperature. In photosynthetic organisms thylakoid membranes are composed of glycolipids, such as monogalactosyldiacylglycerol (MGDG), di-galactosyldiacylglycerol (DGDG), sulfo-quinovosyldiacylglycerol (SQDG), and one phospholipid, phosphatidylglycerol (PG) [4]. The remodeling of PG has crucial role in both cyanobacteria [5,6] and higher plants in the adaptation to low temperatures [7-10]. The biosynthetic pathway of PG helps to understand the remodeling process. The pathway of PG biosynthesis is well known in prokaryotes [11]. The biosynthesis of PG in cyanobacteria is similar to that in Escherichia coli [5,12]. The five major enzymes of PG biosynthesis are as follows: glycerol-3phosphate acyltransferase (G3P acyltransferase), lysophospatidicacid acyltransferase (LPA acyltransferase), CDP-diacylglycerol synthase (CDP-DG synthase), phosphatidylglycerol phosphate synthase (PGP synthase), and PGP phosphatase. As a consequence of the substrate preferences of G3P acyltransferase and LPA acyltransferase, in the newly synthesized PG 18:0 or 16:0 are esterified at the sn-1, and 16:0 at the sn-2 position of the glycerol backbone [13-15]. At the beginning of PG synthesis the newly synthesized PG molecules contain only saturated fatty acids at both the sn-1 and sn-2 positions [16,17]. At a later step of PG synthesis, they are desaturated by acyllipid desaturases, which introduce a double bond at a specific position of the fatty acids, predominantly at sn-1 [18,19]. Earlier PG remodeling studies used genetic manipulation of acyltransferases and desaturases to reveal the protective role of
164
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
various PG molecular species in resistance to low and high temperature stresses in both higher plants [20] and cyanobacteria [2,21,22]. In plants selective expression of acyltransferases allowed manipulation of the unsaturation of PGs. Overexpression of squash (Cucurbita moschata) and Arabidopsis G3P acyltransferases in tobacco (Nicotiana tabacum) altered the levels of saturated molecular species of PG [20,23]. Heterologous expression of the Arabidopsis gene that increased unsaturated fatty acid content of PG at the sn-1 position enhanced chilling resistance, whereas that from squash, which has strong preference to palmitoyl-acyl carrier protein, increased chilling susceptibility in transgenic tobacco [20]. The level of unsaturation of membrane glycerolipids was manipulated by modulating the genes of fatty acid desaturases both in cyanobacteria and plants [2,22,24]. The presence of diunsaturated glycerolipids in the membranes of photosynthetic organisms enhanced chilling tolerance and played a critical role in the recovery from low temperature photoinhibition [25,26]. Temperature stress-induced remodeling of lipids was demonstrated in the salt-tolerant green alga Dunaliella salina [27-29]. However, the mechanism of remodeling, which resulted in the synthesis of various molecular species with redistribution of acyl chains within individual lipid classes, has not been elucidated [30,31]. In Synechocystis PCC6803, the genes for LPA acyltransferase (sll1848) [15], CDP-DG synthase (cdsA, slr1369) [32], and PGP synthase (pgsA, sll1522) [33] involved in PG biosynthesis have recently been identified. Unlike higher plants, cyanobacteria are capable of uptaking exogenously added PG [33] therefore they can serve as ideal models for studying the role and metabolism of PG in photosynthetic organisms. The identification of the genes of PG biosynthesis opened the way for generating the ΔpgsA [33] and ΔcdsA [32] mutants of Synechocystis PCC6803 that are incapable of synthesizing PG, although they can utilize exogenously added artificial PG through uptake from the culturing medium. Therefore, the physiological roles of PG and the biochemical pathways of their biosynthesis can be studied in vivo by these mutants. For example, the effect of PG molecules with various fatty acids has been investigated on photosynthetic and other cellular functions using these biosynthetic mutants [34,35]. The intracellular content of PG can be manipulated in PG-deficient mutant cells by withdrawing exogenously supplied PG. 21-day PG depletion of Synechocystis PCC6803/ΔpgsA results in severe PG deficiency. This system provides an excellent experimental tool for monitoring the in vivo remodeling process of artificial PG molecules. In this study, we investigated the retailoring of PG in ΔpgsA mutant cells. Retailoring is a special case of remodeling, in which an exogenously supplied, non-natural PG species is converted to physiologically optimal forms. Taking advantage of a mutant that is incapable of synthesizing PG we could study the formation of essential PG molecular species derived from dioleoyl-PG [PG(18:1/18:1)] as a precursor. PG(18:1/18:1) contains a single double bond that is the minimum requirement for cell survival [36]. This PG species was added to PG-depleted ΔpgsA cells and retailoring of its fatty acyl moieties was monitored by mass spectrometry. Based on the identification of various PG species as products of the retailoring process we propose a remodeling mechanism for PG in a photosynthetic prokaryote grown under normal environmental conditions, without manipulation of the enzymes involved in this remodeling process. 2. Materials and methods 2.1. Organism and growth condition Cells of wild-type and ΔpgsA mutant strain of Synechocystis PCC6803 were grown photoautotrophically in BG11 medium [37] supplemented with 5 mM HEPES–NaOH (pH 7.5). The mutant cells were cultivated in BG11 medium supplemented with 20 μg/mL kanamycin and in the presence of 20 μM PG(18:1/18:1) (Sigma, St. Louis, MO, USA). Cells were grown at 30 °C, under continuous
illumination at the intensity of 30 μmol photons m−2 s−1. Cultures were aerated on a gyratory shaker operating at 100 rpm. PG depletion was achieved by washing the cells twice with PG-free medium, followed by cultivation in PG-free medium. The cells can survive in PG-free medium only for a limited time, since the decrease of PG content abolishes photosynthetic activity and other cellular functions [38]. Following 21 days of PG-free cultivation, we re-added PG to the medium. Recovering mutant cells, grown in the presence of PG for 1 and 21 days, are indicated as ΔpgsA+PG 1d and ΔpgsA+PG 21d. 2.2. Lipid extraction and analysis For gas chromatographic analysis, lipids were extracted from the collected cells (around OD750 = 50) by the method of Bligh and Dyer [39]. The analysis of lipids was carried out according to the method of Sato and Murata [40]. Lipids were fractionated on precoated thin layer chromatography (TLC) plates (Merck 5721) developed with CHCl3/CH3OH/28% NH4OH (65:35:5, v/v). Fifty micrograms pentadecanoic acid (15:0), an internal standard, was added to the PG band of the TLC plate. After detection of the lipids by primuline fluorescence, the appropriate area of the silica gel (PG band) was scraped and was subjected to methanolysis with 5% HCl in methanol at 85 °C for 2 h. The resulting fatty acid methyl esters of PG were analyzed with a Hewlett Packard (Palo Alto, CA, USA) HP6890 gas chromatograph equipped with a hydrogen flameionization detector. Fatty acid methyl esters were separated on a 30 m × 0.25 mm i.d. SP-2330 capillary column (Supelco, Bellefonte, PA, USA). Temperatures of the column and the flame-ionization detector were 180 and 260 °C, respectively. The relative amounts of fatty acid methyl esters were determined by comparing the areas under the peaks on the chromatogram to the area of the internal fatty acid standard (15:0). For electrospray ionization (ESI) mass spectrometry, total lipids were isolated from intact cells of cyanobacteria (around OD750 = 50) with the method of Bligh and Dyer [39]. For semi-quantitative analyses, 20 μg PG(14:0/14:0) (Avanti Polar Lipids, Alabaster, USA) was added to each sample as internal standard before lipid extraction. 2.3. Mass spectrometry The extracts of the biological material or calibration solutions, spiked with the PG(14:0/14:0) internal standard (20 μg/mL), were diluted 100 times with water/acetonitrile mixture (50:50) before analysis. The samples (50 μL) were directly infused at a flow rate of 5 μL/min into a hybrid linear ion trap–Fourier transform ion cyclotron resonance (FTICR) mass spectrometer with a 7-Tesla superconducting magnet (LTQ-FT, Thermo Fisher, San Jose, CA) and equipped with an ESI source operated in the negative ionization mode. The spray voltage was set to 5 kV, sheath gas flow rate to 10 units, capillary temperature to 275 °C, and the ion optics elements voltages were set by the built-in automatic tuning procedure using a solution of the PG (18:1/18:1) standard. Singly negatively charged phospholipid [M-H]- molecular ions were detected by full-scan FTICR acquisition in the 150–2000 Th mass-to-charge ratio (m/z) range. Mass resolution (m/Δm) was set to 50000 at m/z of 400 Th. Approximately 80 mass spectra were averaged per sample. For further structural characterization, MS/MS experiments were performed. The [M-H]- ions were isolated and subjected to collision-induced dissociation (CID) in the linear ion trap that was also used to detect the product ions (MS/MS spectra). The isolation width was set to 1.0 Th and the collision energy to 15%, respectively, with helium as target gas. To estimate the concentration of PG species in the biological extracts, the following simple equation was used: cA = cIS = kA ðIA = IIS Þ
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
where cA is the concentration of analyte, cIS is the concentration of internal standard [PG(14:0/14:0)] and IA and IIS are the ion intensities of analyte and the internal standard, respectively. Calibration solutions containing 1, 2, 5, 10, 20 and 50 μg/mL of PG(18:1/18:1) or PG(16:0/18:1) (Avanti Polar Lipids, Alabaster, USA) and 20 μg/mL of the internal standard were analyzed by full-scan negative-ion ESI– FTICR mass spectrometry, and the cA/cIS ratios were plotted against the IA/IIS ratios. The kPG(18:1/18:1) and kPG(16:0/18:1) constants were determined as the slopes of linear regressions with zero intercept. The constants were then used to calculate PG(18:1/18:1) (m/z 773) and PG(18:1/16:0) (m/z 747) concentrations from the observed IA/IIS ratios in the full-scan negative-ion ESI–FTICR mass spectra of biological samples. To estimate the concentration of PG (18:2/18:1) at m/z 771 and PG(18:2/16:0) at m/z 745, we used the average of kPG(18:1/18:1) and kPG(16:0/18:1) determined. 3. Results 3.1. Gas chromatographic analyses revealed remodeling of PG fatty acids PG of Synechocystis PCC6803/ΔpgsA mutant was separated and the fatty acid composition was measured by gas chromatography according to Hagio et al. [33]. The results are summarized in Table 1. The PG-deficient mutant was grown in the presence of exogenously added PG(18:1/18:1). After 1 day of culturing of the mutant cells in the presence of PG(18:1/18:1), the most abundant fatty acid species was oleic acid. In the ΔpgsA+PG 21d sample, oleic acid content decreased from 60% to 11%. At the expense of oleic acid, palmitic acid content increased from 19% to 57%, and polyunsaturated fatty acids were present in much higher amounts, as compared to ΔpgsA+PG 1d sample. The estimation of fatty acid composition of the PG in wild-type cells was done by Murata et al. [13]. This estimation was based on the fatty acid composition determined with Rhizopus phospholipase A2 (PLA2). The most abundant PG species were those which contained mainly unsaturated fatty acids at the sn-1 position (18:2/16:0; 18:1/ 16:0), whereas the ones with saturated fatty acids at the same position (18:0/16:0) were less abundant, in agreement with the results published by Okazaki and coworkers [14]. 3.2. Mass spectrometry confirms the presence of newly formed PGs The biochemical modification of artificial PG was monitored by mass spectrometry. Fig. 1 shows the negative-ion ESI–FTICR mass spectra of dioleoyl-PG (Fig. 1a) and those of the lipid extracts of mutant cells following 1 day (Fig. 1b), 12 days (Fig. 1c), 18 days (Fig. 1d) and 21 days (Fig. 1e) of culturing after the re-addition of dioleoylPG. Since after 21 days of PG depletion the mutant cells did not contain accurately detectable levels of PG, the first time point of the analysis was 1 day following re-addition. The monoisotopic [M-H]ion of dioleoyl-PG is detected at m/z 773.5339 (Fig. 1a). Apparently, several new deprotonated [M-H]- PG species (at m/z 771.5159,
Table 1 Fatty acid composition (mol%) of PG from Synechocystis PCC6803/ΔpgsA mutant cells. Strain
ΔpgsA+PG 1d ΔpgsA+PG 21d
Acyl chains in PG molecules (mol%) 16:0
16:1
18:0
18:1
18:2
18:3
19 57
2 4
10 20
60 11
7 8
2 n.d.
Synechocystis PCC6803/ΔpgsA cells were grown in PG-supplemented medium for 1 or 21 days (ΔpgsA+PG 1d and ΔpgsA+PG 21d, respectively). 16:0, hexadecanoic acid (palmitic acid); 16:1, hexadecenoic acid (palmitoleic acid); 18:0, octadecanoic acid (stearic acid); 18:1, 9-octadecenoic acid (oleic acid); 18:2, 9,12-octadecadienoic acid (linoleic acid); 18:3, 9,12,15-octadecatrienoic acid (α-linolenic acid); n.d., not detected.
165
747.5164 and 745.5004) appeared after 12 days of culturing in the presence of the added synthetic PG (Fig. 1c). Accurate m/z values (Fig. 1e) of the molecular ions observed in the ESI-FTICR mass spectra (Table 2) permitted the verification of the composition of these new PG forms and, consequently, guided the identification of their accurate chemical structure. In all cases, the mass accuracy was within 5 ppm. Following 12-day culturing (Fig. 1d and e) new PG molecular species did not emerge, however, the amount of the newly formed PG species continuously increased during the 21-day culturing period with concomitant decrease in the dioleoyl-PG content (Fig. 1e). Structural identification of artificial and reformed PG molecules was performed by low-energy CID-MS/MS of [M-H]- ions observed in the full-scan mass spectrum displayed in Fig. 1e (ΔpgsA+PG 21d sample). Also relying upon earlier results implicating the preference of acyl-lipid desaturases to introduce double bonds into the sn-1 fatty acid of PG in cyanobacteria [18,19], we could infer fatty acyls bound to the sn-1 and sn-2 positions by detailed MS/MS analyses using the results of earlier studies by Hsu and Turk [41] on CID of [M-H]- ions of various synthetic PG species. The features they identified permitted the reliable assignments of the fatty acid substituents and their position in the glycerol backbone. In particular, we deduced structural information from major fragment ions that were characteristic for the individual PG molecules and arose from a neutral loss of free fatty acid substituents [M-HRxCO2H]- or [M-H-R'xCH = C = O]- ketenes, by consecutive losses of the fatty acid and the glycerol head group [M-HRxCOOH-C3H6O2]-, and by formation of fatty carboxylate anions [RxCO-2], where x = 1, 2 and Rx = R'xCH2. We also utilized the feature that the abundances of the ions arising from neutral loss of the sn-2 substituent as a free fatty acid [M-H-R2CO2H]- or as a ketene [M-H-R'2CH = C = O]- were generally greater than those of the product ions from the analogous losses at sn-1 [41]. This was due to the nucleophilic attack of the anionic phosphate site on the C-1 or the C-2 of the glycerol, to which the carboxylates attached, to expel sn-1 (R1CO-2) or the sn-2 (R2CO-2) carboxylate anion with preference to R2COO- over R1COO[41]. Identification of the major isomers by this rule of Hsu and Turk [41] did not, however, exclude the presence of the other isobaric isomers in small amounts. Fig. 2a shows the MS/MS spectrum of the added synthetic PG(18:1/18:1) as reference. We indeed observed fragment ions at m/z 491, 509 and 417, reflecting neutral loss of the 18:1 fatty acyl substituent as an acid, neutral loss of a ketene, and consecutive losses of the fatty acid and glycerol head group, respectively. A single carboxylate anion (m/z 281) appeared, because the fatty acyl constituent was the same in both the sn-1 and sn-2 positions of PG(18:1/18:1). 3.3. Identification of PG(18:2/18:0) The product–ion spectrum of the m/z 773 [M-H]- ions of the lipid extract obtained 21 days after PG(18:1/18:1) re-addition is shown in Fig. 2b. In addition to the fragment ions observed for PG (18:1/18:1), the MS/MS spectrum also contained additional ions at m/z 493 and 489 reflecting neutral losses of 18:2 and 18:0 fatty acids, respectively. Ions at m/z 419 and 415 were derived from the consecutive losses of fatty acid and the glycerol head group. The ions at m/z 511 and 507 corresponded to neutral losses as ketenes. The carboxylate anions were m/z 279 and 283, respectively. The spectrum showed that ions m/z 489, 415, 507 and 283 were more abundant than ions at m/z 493, 419, 511 and 279. These fragment ions confirmed the presence of isomeric PG species, and the intensities suggested that the 18:2 fatty acid residues were located predominantly in the sn-1 position, while the 18:0 residues were predominantly in the sn-2 position. Thus the MS/MS spectrum also indicated presence of PG(18:2/18:0), in addition to the added artificial PG(18:1/18:1).
166
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
Fig. 1. The “phosphatidylglycerol range” (showing the [M-H]- ions of this class of lipids) in the negative-ion ESI-FTICR mass spectra of (a) artificial synthetic PG(18:1/18:1) (b), ΔpgsA+PG 1d (c), ΔpgsA+PG 12d (d), ΔpgsA+PG 18d and (e), ΔpgsA+PG 21d lipid extracts.
3.4. Identification of PG(18:2/18:1) and its isoform, PG(18:3/18:0) The product–ion spectrum of the [M-H]- ions of m/z 771 (Fig. S1) showed ions at m/z 491 and 489 reflecting neutral loss of 18:2 and 18:1 fatty acids, respectively. The m/z 417, 415 ions arose from the consecutive losses of fatty acid and the glycerol head group. The spectrum also contained ions at m/z 509 and 507, reflecting neutral loss of the fatty-acyl substituents as ketenes. The abundance of the m/z 489 ion was greater than that of the m/z 491 ion, and the abundance of the m/z 415 ion was higher than that of 417. The abundance of the m/z 507 ion was also greater than that of the m/z 509 ion. The abundance of the carboxylate anion at m/z 281 was more abundant than the
corresponding ion arising from the similar loss at m/z 279. From the fragment ion intensities we inferred that the 18:2 and 18:1 fatty acids were esterified to the sn-1 and the sn-2 positions of the glycerol backbone, respectively. In this MS/MS spectrum, obtained from m/z 771 as precursor, we also identified other ions of low intensity. However, the abundances at m/z 487, 413, 505 and 283 were clearly greater than those of m/z 493, 419, 511 and 277. Therefore, we attributed these fragment ions to losses of 18:3 and 18:0 fatty acyl groups from the sn-1 and sn-2 positions, respectively. Altogether, the MS/MS spectrum of m/z 771 (Fig. S1) indicated the presence of both PG(18:2/18:1) and PG(18:3/18:0). 3.5. Identification of PG(18:1/16:0) and its isoform PG(16:1/18:0)
Table 2 Molecular ions (m/z) of putative PG species. Theoretical (m/z)
Detected (m/z)
Error (ppm)
Number of carbon (C) atoms
Number of desaturation
Putative PG species (sn-1/sn-2)
773.5338
773.5325
−1.7
C42
2
771.5201
771.5164
−4.8
C42
3
747.5182
747.5167
−1.9
C40
1
745.5025
745.5007
−2.4
C40
2
18:1/18:1 18:2/18:0 18:2/18:1 18:3/18:0 18:1/16:0 16:0/18:1 16:1/18:0 18:2/16:0 16:1/18:1
Italics indicate the most abundant PG molecular species for the given m/z.
Fig. S2 shows the MS/MS spectrum of the [M-H]- ions of m/z 747. Ions at m/z 465 and 491, reflecting neutral losses of 18:1 and 16:1 fatty acids, were produced. The m/z 391 and 417 ions arose from the consecutive losses of fatty acid and the glycerol head group. Other abundant ions at m/z 483 and 509 reflect neutral loss of the fatty acyl substituent as a ketene. We observed that the abundance of the m/z 491 ion was greater than that of the m/z 465 ion, the abundance of the m/z 417 ion was higher than that of the m/z 391 ion, and the abundance of the m/z 509 ion is also greater than that of the m/z 483 ion. The carboxylate anion at m/z 255 was less abundant than the corresponding ion arising from the similar loss at m/z 281. The relative abundance of these fragment ions indicated that the 18:1 and 16:0 fatty acids were derived from the sn-1 and sn-2 positions, respectively. In this m/z 747 ion spectrum we could also identify
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
167
Fig. 2. The CID product–ion spectra of (a), [M-H]- of artificial PG(18:1/18:1) standard (precursor: m/z 773), (b), [M-H]- of PG(18:1/18:1) and its isomer PG(18:2/18:0) (precursor: m/z 773) extracted from ΔpgsA+PG 21 d.
other ions, albeit with very low abundance. Ion abundances from the loss of fatty-acyl constituents at m/z 463, 389, 481 and 283 were greater than those at m/z 493, 419, 511 and 253, which reflected the loss of 16:1 and 18:0 fatty-acyl constituents from the sn-1 and sn-2 positions of the glycerol backbone, respectively. Therefore, the MS/ MS spectrum of m/z 747 (Fig. S2) indicated the presence of both PG (18:1/16:0) and PG(16:1/18:0). 3.6. Identification of PG (18:2/16:0) and its isoform PG(16:1/18:1) When the parent [M-H]- ions at m/z 745 were fragmented, the product ions permitted the identification and position assignment of fatty-acyl groups attached to the glycerol backbone, and isomers such as PG(18:2/16:0) and PG(16:1/18:1) could be easily distinguished (Fig. S3). MS/MS product ions at m/z 465 and 489 reflected neutral loss of 18:2 and 16:0 fatty acids. The m/z 391, 415 product ions arose from the consecutive losses of fatty acid and the glycerol head group. Other abundant ions at m/z 483 and 507 reflected neutral losses of the fatty-acyl substituents as ketenes. The abundance of the m/z 489 ion was greater than that of the m/z 465 ion, the abundance of the m/z 415 ion was higher than that of m/z 391. The abundance of the m/z 507 ion was also greater than that of the m/z 483 ion. The carboxylate anion at m/z 255 was more abundant than the corresponding ion
arising from a similar fragmentation process yielding m/z 279 suggesting that 18:2 fatty acid was esterified to the sn-1, while 16:0 fatty acid to the sn-2 position, corresponding to PG(18:2/16:0). In this MS/MS spectrum from m/z 745 as the precursor ion (Fig. S3), we also noticed other ions corresponding to the loss of 16:1 and 18:1 fattyacyl substituents. The relative abundances of the ions at m/z 463, 389, 481, and 281 were greater than those derived from ions at m/z 491, 509, 417 and 253, respectively. Therefore, they indicated the presence of both PG(18:2/16:0) and PG(16:1/18:1). 3.7. Semi-quantitative mass spectrometric analysis of PG molecular species To follow the dynamics of PG retailoring, we made semiquantitative analyses of PG molecular species. We used an internal standard, PG(14:0/14:0), because it was not a natural phospholipid and, moreover, its molecular ion (m/z 665) did not overlap with those of the naturally occurring molecular species in the negative-ion ESI mass spectra. Relative ion intensities obtained from full-scan FTICR measurements were used for quantitation, therefore, we could not distinguish isobaric PG isomers. Based on comparing the relative ion intensities of the m/z 773, 771, 747 and 745 ions, we observed a gradual decrease of PG(18:1/18:1) (m/z 773) during the culturing
168
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
period (1, 14 and 21 days) (Fig. 3a). At the same time, the concentration of other PG species at m/z 771 (Fig. 3b), 747 (Fig. 3c) and 745 (Fig. 3d) increased at the expense of m/z 773, indicating remodeling of the PG species. The concentrations of PG species were also estimated using calibration coefficients. PG(18:1/18:1) decreased from 1.77 to 0.77 μM, together with simultaneous increase in the concentration of the other three most abundant PG molecular species (Table 3). The tendency was similar to that shown in Fig. 3. 4. Discussion GC analyses indicated retailoring of PG(18:1/18:1) in Synechocystis PCC6803/ΔpgsA, in agreement with our previous results obtained with the PAL/ΔcdsA mutant [42]. This exogenous PG was reformed by the cyanobacteria according to their functional needs. In cells, natural PG molecular species were synthesized which provided essential PG molecules for cellular functions. Our earlier studies highlighted the importance of PG(18:2/16:0), revealing that this molecular species is needed for resistance to low temperature photoinhibition [24,43]. This can explain why the remodeling proceeds mainly to the production of PG(18:2/16:0). Whereas our earlier GC results suggested remodeling, accurate and site-specific information has been provided by the mass spectrometric analyses performed. For the reliability of our results it was important verifying that the Synechocystis PCC6803/ΔpgsA cells remained fully viable during the 21-day PG depletion period. Earlier we demonstrated that, following PG re-addition, the photosynthetic activity of the cultures fully recovered from the PG starvation-related decrease [33,44]. Furthermore, the viability of the cells was confirmed in each stage of the experiments by determining the numbers of colony-forming units (data not shown). The results of mass spectrometric analyses, together with earlier evidence about the preference of acyl-lipid desaturases to introduce double bonds into the sn-1 fatty acid of PG in cyanobacteria [18,19], suggested the biosynthetic scheme shown in Fig. 4a. Starting from the
Table 3 Temporal changes in molar concentrations of PG species. Strain
Molecular species 18:1/16:0 16:0/18:1 16:1/18:0
18:2/16:0 16:1/18:1
Molecular ion (m/z) 773 771
747
745
Molar concentration (μM) 1.77 0.12 0.77 1.47
0.17 0.67
n.d. 0.88
18:1/18:1 18:2/18:0
ΔpgsA+PG 1d ΔpgsA+PG 21d
18:2/18:1 18:3/18:0
Samples are denoted as in Table 1. Italics indicate the most abundant PG molecular species; n.d., not detected.
added PG(18:1/18:1), the first possible enzymatic reaction is 1 (indicated with solid arrow), which involves an acyl-lipid desaturation that results in the production of linoleoyl from oleoyl residue. The acyl-lipid desaturase (Fig. 4b) that catalyses this reaction is a wellcharacterized enzyme [18,19]. In cyanobacteria lipid desaturation is carried out exclusively by acyl-lipid desaturases [45]. Routes 2, 3, 4 and 8 (Fig. 4a) (open arrows) represent synthetic routes that require at least two separate enzyme reactions. The first step is the cleavage of fatty acids, either at sn-2 or sn-1 positions, which need PLA2 or phospholipase A1 (PLA1), respectively. An alternative biosynthetic route could be mediated by acyl-hydrolases [46,47]. These processes result in the formation of PG species containing fatty acids other than 18:1, such as 16:0 and 18:0. For the enzyme processes in 3 and 4 the cells need PLA2 (Fig. 4b) that can eliminate an acyl group from the sn2 position. Reactions 2 and 8 requires PLA1 (Fig. 4b) that can cleave the ester bound at the sn-1 position. Acyl-hydrolases can remove fatty acids from both the sn-1 and sn-2 positions [48]. In Synechocystis PCC6803 phospholipases, acyl-hydrolases and their genes have not yet been identified. Sequence homology search could not provide positive results, suggesting that if there are phospholipases and acylhydrolases in Synechocystis PCC6803, these are substantially different
Fig. 3. Analyses of PG species, changes in ion intensities: (a), PG(18:1/18:1) (m/z 773); (b), PG(18:2/18:1), (m/z 771); (c), PG(18:1/16:0) (m/z 747); and (d), PG(18:2/16:0) (m/z 745), this molecular species was not detectable after one day re-addition of PG.
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
169
Fig. 4. Putative biosynthetic scheme for the remodeling of PG in Synechocystis PCC6803/ΔpgsA mutant cells. (a) Numbers indicate enzymatic reactions, bold characters denote reformed fatty acid moieties. The chemical reactions, replacing and desaturating fatty acids, are indicated by open and solid arrows, respectively. (b) “'” and “⁎” symbols indicate the replacement and desaturation of fatty acids at the sn-1 and sn-2 positions, respectively. Reformed fatty acids and known biosynthetic enzymes are highlighted by bold characters.
from those known from higher plants. Acyltransferase enzymes (Fig. 4b) with specific acyl-carrier characteristics are required for new fatty acids esterified to PG. Lysoglycerophospholipid acyltransferase (LGP acyltransferase), a mainly sn-2-specific 16:0 acyltransferase, has been identified in cyanobacteria [15], which explains our finding that one of the main products of remodeling is PG(18:1/16:0) that could be desaturated further to PG(18:2/16:0) (Fig. 4a, route 3). The replacement of fatty acids is followed by desaturation (Fig. 4a, routes 5, 6, 7 and 9) (bold arrows). As a possible alternative mechanism (Fig. 4a, route 2) we could mention the exchange of acyl groups of PG with those of other lipids. In cyanobacteria there are acyltransferases, such as acyl-ACP: lyso-MGDG acyltransferase [49], therefore, 18:2 can be also transferred from MGDG to the sn-1 position of lyso-PG synthesized by acyl-hydrolase or PLA1 (Fig. 4b). One day after the re-addition of PG(18:1/18:1), the most abundant molecular species was still the synthetic lipid, representing more than 90% of the overall PG content. However, some minor PG derivatives containing 18:2 and 16:0 acyl chains could also be observed as low-abundance components, and were retained after PG depletion, when the cells needed specific residual PG for survival. The products of desaturation emerged 1 day after the re-addition of dioleoyl-PG. After 2 weeks of culturing in the presence of dioleoyl-PG, more than 50% of total PG contained 18:2, which was the most abundant polyunsaturated fatty acid. The 18:2 was esterified to the sn-1 position. Twenty-one days after re-addition of dioleoyl-PG, a rather significant accumulation of 16:0 occurred. This fatty acid esterified mainly the sn-2 position of PG. The above enzymatic reactions (Fig. 4b) are capable of remodeling lipid molecules, providing an important regulatory machinery for the adaptation to environmental changes [2,7]. The proposed biosynthetic scheme is based on the known processes of lipid synthesis in cyanobacteria and higher plants [35,50,51] although our results suggest the presence of yet unknown synthetic steps in cyanobacteria. In conclusion, in vivo retailoring of exogenously added dioleoyl-PG was found to produce several native PG species that are required for the survival of Synechocystis PCC6803/ΔpgsA cells. We propose that PG retailoring in Synechocystis PCC6803 is mediated by PLA1, PLA2 and acyl-hydrolases which are necessary to remove fatty acids from the glycerol backbone, lipid acyltransferases needed for constructing new fatty acid residues esterified to the glycerol backbone, and acyllipid desaturases that create the physiologically required lipid
molecules. Our data provide evidence that cyanobacteria possess the enzymatic machinery capable of PG remodeling. Acknowledgements This research was supported by the Hungarian Science Foundation (OTKA: grants K 60109 and K 68692 to Z.G.). Laszlo Prokai is the Robert A. Welch Professor of the University of North Texas Health Science Center (grant BK-0031). P.F. thanks the Ministry of Education of the Czech Republic (MSM6198959216). The authors are grateful to Anna Sallai for her technical assistance. We are thankful to Dr. Miklós Szekeres (Institute of Plant Biology, Biological Research Center of the Hungarian Academy of Sciences, Szeged) for critical reading of the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbalip.2009.10.009. References [1] J.L. Harwood, A.L. Jones, H.J. Perry, A.J. Rutter, K.L. Smith, M. Williams, Changes in plant lipids during temperature adaptation, in: A.R. Cossins (Ed.), Temperature Adaptation of Biological Membranes, Portland Press, London, 1994, pp. 107–118. [2] H. Wada, N. Murata, Temperature-induced changes in the fatty-acid composition of the cyanobacterium, Synechocystis PCC6803, Plant Physiol. 92 (1990) 1062–1069. [3] J.L. Harwood, Involvement of chloroplast lipids in the reaction of plants submitted to stress, in: P.A. Siegenthaler, N. Murata (Eds.), Advances in Photosynthesis, Lipids in Photosynthesis: Structure, Function and Genetics, vol. 6, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998, pp. 287–302. [4] P.A. Siegenthaler, Molecular organization of acyl lipids in photosynthetic membranes of higher plants, in: P.A. Siegenthaler, N. Murata (Eds.), Advances in Photosynthesis, Lipids in Photosynthesis: Structure, function and genetics, vol. 6, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998, pp. 119–144. [5] N. Murata, I. Nishida, Lipids in blue-green algae (cyanobacteria), in: P.K. Stumpf (Ed.), The Biochemistry of Plants, Academic Press, Orlando, 1987, pp. 315–347. [6] N. Sato, N. Murata, Y. Miura, N. Ueta, Effect of growth temperature on lipid and fatty-acid compositions in the blue-green-algae, Anabaena variabilis and Anacystis nidulans, Biochim. Biophys. Acta 572 (1979) 19–28. [7] N. Murata, N. Sato, N. Takahashi, Y. Hamazaki, Compositions and positional distributions of fatty-acids in phospholipids from leaves of chilling-sensitive and chilling-resistant plants, Plant Cell Physiol. 23 (1982) 1071–1079.
170
H. Laczko-Dobos et al. / Biochimica et Biophysica Acta 1801 (2010) 163–170
[8] N. Murata, Molecular-species composition of phosphatidylglycerols from chillingsensitive and chilling-resistant plants, Plant Cell Physiol. 24 (1983) 81–86. [9] P.G. Roughan, Phosphatidylglycerol and chilling sensitivity in plants, Plant Physiol. 77 (1985) 740–746. [10] Y. Tasaka, I. Nishida, S. Higashi, T. Beppu, N. Murata, Fatty-acid composition of phosphatidylglycerols in relation to chilling sensitivity of woody-plants, Plant Cell Physiol. 31 (1990) 545–550. [11] A. Ohta, K. Waggoner, A. Radominskapyrek, W. Dowhan, Cloning of genes involved in membrane lipid synthesis—effects of amplification of phosphatidylglycerophosphate synthase in Escherichia coli, J. Bacteriol. 147 (1981) 552–562. [12] H. Wada, N. Murata, Membrane lipids in cyanobacteria, in: P.A. Siegenthaler, N. Murata (Eds.), Advances in Photosynthesis, Lipids in Photosynthesis: Structure, Function and Genetics, vol. 6, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1998, pp. 65–81. [13] N. Murata, H. Wada, Z. Gombos, Modes of fatty-acid desaturation in cyanobacteria, Plant Cell Physiol. 33 (1992) 933–941. [14] K. Okazaki, N. Sato, N. Tsuji, M. Tsuzuki, I. Nishida, The significance of C16 fatty acids in the sn-2 positions of glycerolipids in the photosynthetic growth of Synechocystis sp PCC6803, Plant Physiol. 141 (2006) 546–556. [15] D. Weier, C. Muller, C. Gaspers, M. Frentzen, Characterisation of acyltransferases from Synechocystis sp PCC6803, Biochem. Biophys. Res. Commun. 334 (2005) 1127–1134. [16] N. Sato, N. Murata, Lipid biosynthesis in the blue-green-alga, Anabaena variabilis. 1. Lipid Classes, Biochim. Biophys. Acta 710 (1982) 271–278. [17] N. Sato, N. Murata, Lipid biosynthesis in the blue-green-alga, Anabaena variabilis. 2. Fatty-acids and lipid molecular-species, Biochim. Biophys. Acta 710 (1982) 279–289. [18] N. Murata, H. Wada, Acyl-lipid desaturases and their importance in the tolerance and acclimatization to cold of cyanobacteria, Biochem. J. 308 (1995) 1–8. [19] H. Wada, N. Murata, Synechocystis PCC6803 mutants defective in desaturation of fatty-acids, Plant Cell Physiol. 30 (1989) 971–978. [20] N. Murata, Q. Ishizakinishizawa, S. Higashi, H. Hayashi, Y. Tasaka, I. Nishida, Genetically engineered alteration in the chilling sensitivity of plants, Nature 356 (1992) 710–713. [21] T. Sakamoto, D.A. Los, S. Higashi, H. Wada, I. Nishida, M. Ohmori, N. Murata, Cloning of omega-3 desaturase from cyanobacteria and its use in altering the degree of membrane-lipid unsaturation, Plant Mol. Biol. 26 (1994) 249–263. [22] H. Wada, Z. Gombos, T. Sakamoto, N. Murata, Genetic manipulation of the extent of desaturation of fatty-acids in membrane-lipids in the cyanobacterium Synechocystis PCC6803, Plant Cell Physiol. 33 (1992) 535–540. [23] F.P. Wolter, R. Schmidt, E. Heinz, Chilling sensitivity of Arabidopsis-thaliana with genetically engineered membrane-lipids, EMBO J. 11 (1992) 4685–4692. [24] Y. Tasaka, Z. Gombos, Y. Nishiyama, P. Mohanty, T. Ohba, K. Ohki, N. Murata, Targeted mutagenesis of acyl-lipid desaturases in Synechocystis: evidence for the important roles of polyunsaturated membrane lipids in growth, respiration and photosynthesis, EMBO J. 15 (1996) 6416–6425. [25] Z. Gombos, E. Kanervo, N. Tsvetkova, T. Sakamoto, E.M. Aro, N. Murata, Genetic enhancement of the ability to tolerate photoinhibition by introduction of unsaturated bonds into membrane glycerolipids, Plant Physiol. 115 (1997) 551–559. [26] N. Murata, Low-temperature effects on cyanobacterial membranes, J. Bioenerg. Biomembr. 21 (1989) 61–75. [27] D.V. Lynch, G.A. Thompson, Retailored lipid molecular-species—a tactical mechanism for modulating membrane-properties, Trends Biochem. Sci. 9 (1984) 442–445. [28] D.V. Lynch, G.A. Thompson, Microsomal phospholipid molecular-species alterations during low-temperature acclimation in Dunaliella, Plant Physiol. 74 (1984) 193–197. [29] D.V. Lynch, G.A. Thompson, Chloroplast phospholipid molecular-species alterations during low-temperature acclimation in Dunaliella, Plant Physiol. 74 (1984) 198–203.
[30] I. Nishida, N. Murata, Chilling sensitivity in plants and cyanobacteria: the crucial contribution of membrane lipids, Annu. Rev. Plant Physiol. Plant Mol. Biol. 47 (1996) 541–568. [31] P.G. Roughan, C.R. Slack, Cellular organization of glycerolipid metabolism, Annu. Rev. Plant Physiol. Plant Mol. Biol. 33 (1982) 97–132. [32] N. Sato, M. Hagio, H. Wada, M. Tsuzuki, Requirement of phosphatidylglycerol for photosynthetic function in thylakoid membranes, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 10655–10660. [33] M. Hagio, Z. Gombos, Z. Varkonyi, K. Masamoto, N. Sato, M. Tsuzuki, H. Wada, Direct evidence for requirement of phosphatidylglycerol in photosystem II of photosynthesis, Plant Physiol. 124 (2000) 795–804. [34] I. Domonkos, H. Laczkó-Dobos, Z. Gombos, Lipid-assisted protein-protein interactions that support photosynthetic and other cellular activities, Prog. Lipid Res. 47 (2008) 422–435. [35] H. Wada, N. Murata, The essential role of phosphatidylglycerol in photosynthesis, Photosynth. Res. 92 (2007) 205–215. [36] I. Sakurai, M. Hagio, Z. Gombos, T. Tyystjarvi, V. Paakkarinen, E.M. Aro, H. Wada, Requirement of phosphatidylglycerol for maintenance of photosynthetic machinery, Plant Physiol. 133 (2003) 1376–1384. [37] M.M. Allen, Simple conditions for growth of unicellular blue-green algae on plates, J. Phycol. 4 (1968) 1–4. [38] I. Domonkos, P. Malec, A. Sallai, L. Kovacs, K. Itoh, G.Z. Shen, B. Ughy, B. Bogos, I. Sakurai, M. Kis, K. Strzalka, H. Wada, S. Itoh, T. Farkas, Z. Gombos, Phosphatidylglycerol is essential for oligomerization of photosystem I reaction center, Plant Physiol. 134 (2004) 1471–1478. [39] E.G. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. J. Biochem. Physiol. 37 (1959) 911–917. [40] N. Sato, N. Murata, Membrane lipids, Methods Enzymol. 167 (1988) 251–259. [41] F.F. Hsu, J. Turk, Studies on phosphatidylglycerol with triple quadrupole tandem mass spectrometry with electrospray ionization: fragmentation processes and structural characterization, J. Am. Soc. Mass Spectrom. 12 (2001) 1036–1043. [42] H. Laczko-Dobos, B. Ughy, S.Z. Toth, J. Kornenda, O. Zsiros, I. Domonkos, A. Parducz, B. Bogos, M. Komura, S. Itoh, Z. Gombos, Role of phosphatidylglycerol in the function and assembly of Photosystem II reaction center, studied in a cdsAinactivated PAL mutant strain of Synechocystis sp. PCC6803 that lacks phycobilisomes, Biochim. Biophys. Acta 1777 (2008) 1184–1194. [43] Z. Gombos, H. Wada, N. Murata, The recovery of photosynthesis from lowtemperature photoinhibition is accelerated by the unsaturation of membranelipids—a mechanism of chilling tolerance, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 8787–8791. [44] Z. Gombos, Z. Varkonyi, M. Hagio, M. Iwaki, L. Kovacs, K. Masamoto, S. Itoh, H. Wada, Phosphatidylglycerol requirement for the function of electron acceptor plastoquinone Q(B) in the photosystem II reaction center, Biochemistry 41 (2002) 3796–3802. [45] S. Higashi, N. Murata, An in vivo study of substrate specificities of acyl-lipid desaturases and acyltransferases in lipid synthesis in Synechocystis PCC6803, Plant Physiol. 102 (1993) 1275–1278. [46] A.R. Matos, A.T. Pham-Thi, Lipid deacylating enzymes in plants: old activities, new genes, Plan Physiol. Biochem. 47 (2009) 491–503. [47] G.A. Mignery, C.S. Pikaard, D.J. Hannapel, W.D. Park, Isolation and sequenceanalysis of cDNAs for the major potato-tuber protein, patatin, Nucleic Acids Res. 12 (1984) 7987–8000. [48] X.M. Wang, Lipid signaling, Curr. Opin. Plant Biol. 7 (2004) 329–336. [49] H.H. Chen, A. Wickrema, J.G. Jaworski, Acyl-acyl-carrier protein: lysomonogalactosyldiacylglycerol acyltransferase from the cyanobacterium Anabaena variabilis, Biochim. Biophys. Acta 963 (1988) 493–500. [50] X. Wang, Plant phospholipases, Annu. Rev. Plant Physiol. Plant Mol. Biol. 52 (2001) 211–231. [51] W. Yang, S.P. Devaiah, X. Pan, G. Isaac, R. Welti, X. Wang, AtPLAI is an acyl hydrolase involved in basal jasmonic acid production and Arabidopsis resistance to Botrytis cinerea, J. Biol. Chem. 282 (2007) 18116–18128.