|Home | About | Journals | Submit | Contact Us | Français|
Plastoglobulins (PGL) are the predominant proteins of lipid globules in the plastids of flowering plants. Genes encoding proteins similar to plant PGL are also present in algae and cyanobacteria but in no other organisms, suggesting an important role for these proteins in oxygenic photosynthesis. To gain an understanding of the core and fundamental function of PGL, the two genes that encode PGL-like polypeptides in the cyanobacterium Synechocystis sp. PCC 6803 (pgl1 and pgl2) were inactivated individually and in combination. The resulting mutants were able to grow under photoautotrophic conditions, dividing at rates that were comparable to that of the wild-type (WT) under low-light (LL) conditions (10 microeinsteins·m−2·s−1) but lower than that of the WT under moderately high-irradiance (HL) conditions (150 microeinsteins·m−2·s−1). Under HL, each Δpgl mutant had less chlorophyll, a lower photosystem I (PSI)/PSII ratio, more carotenoid per unit of chlorophyll, and very much more myxoxanthophyll (a carotenoid symptomatic of high light stress) per unit of chlorophyll than the WT. Large, heterogeneous inclusion bodies were observed in cells of mutants inactivated in pgl2 or both pgl2 and pgl1 under both LL and HL conditions. The mutant inactivated in both pgl genes was especially sensitive to the light environment, with alterations in pigmentation, heterogeneous inclusion bodies, and a lower PSI/PSII ratio than the WT even for cultures grown under LL conditions. The WT cultures grown under HL contained 2- to 3-fold more PGL1 and PGL2 per cell than cultures grown under LL conditions. These and other observations led us to conclude that the PGL-like polypeptides of Synechocystis play similar but not identical roles in some process relevant to the repair of photooxidative damage.
Members of a diverse family of polypeptides now known as plastoglobulins (PGL) (16) are present in virtually all plants, algae, and cyanobacteria, but in no other life forms. They are, in other words, restricted to organisms that carry out oxygenic photosynthesis. PGL were first described as the predominant proteins of carotenoid-containing lipid tubules and globules of pepper (7) and cucumber (44) chromoplasts, but soon thereafter, it became apparent that PGL are not confined to chromoplasts or to carotenoid-accumulating lipid globules. Rather, these polypeptides are present in all types of plastids (8, 11, 16) and may be distributed throughout the stroma (8) and/or associated with the thylakoid membranes (8, 24, 28), as well as with lipid globules.
PGL in flowering plants are thought to coat the outer surface of plastid lipid globules and related lipoprotein structures (7, 16, 44, 45), stabilizing these structures and/or preventing their coalescence, as is believed to be the function of the oleosin proteins that reside on the surface of lipid globules in the cytosolic compartment of plant cells (14). More than this, PGL may well initiate or in some way assist in the formation and assembly of lipid globules in plant plastids (3, 4). Perhaps the most persuasive evidence for this function of PGL was provided by the demonstration by Deruère et al. (7) that lipoprotein structures indistinguishable from those observed in pepper chromoplasts in vivo could be produced by combining a pepper PGL with appropriate lipids and carotenoids in vitro.
In addition to the reputed roles of PGL in the initiation and stabilization of lipid globules, there is evidence to indicate that certain PGL are involved in some way in protecting the photosynthetic apparatus, particularly photosystem II (PSII), from photooxidative damage (24, 36, 47). An important question is whether protection of PSII and the thylakoid membrane is in some way related to the role of PGL in the formation of lipid globules or is distinct from it. An emerging concept of the role of lipid globules in plant chloroplasts (3, 4, 9, 17, 37) provides a possible explanation for the photoprotective effects of PGL. In this concept the plastid lipid globules serve as a depot or site of sequestration for potentially toxic free fatty acids, carotenoids, phytol, quinones, and other lipophilic compounds that may become damaged or unbound as a consequence of photooxidative damage. Because the lipid globules remain connected to the thylakoid membrane (3), the flow of metabolites may later be reversed, with the lipid globules now serving as a reservoir of raw materials for the repair and remodeling of the thylakoid membrane.
PGL of flowering plants are encoded by a family of 10 or more distinctly different genes (20). Alternatively spliced transcripts for many of the pgl genes of Arabidopsis thaliana (http://www.arabidopsis.org/index.jsp) indicate that 20 or more pgl gene products may be produced in this model plant. The relatively large number of gene products, together with probable overlaps in the functions of these products (e.g., seven different PGL have been identified in the lipid globules of A. thaliana chloroplasts [43, 48]), makes problematic the use of a classical genetic approach for studying PGL functions in plants. For example, no phenotype could be discerned for an A. thaliana mutant inactivated in a gene encoding a PGL referred to as AtFib1a (42), nor was a phenotype observed for double mutants with the expression of a second PGL-encoding gene (AtFib1b or AtFib2) “knocked down” with RNA interference in the background of the AtFib1a mutant (42).
Cyanobacteria, which typically contain only one or two genes that encode PGL-like polypeptides, are inherently more suitable than eukaryotic plants for use as experimental subjects to investigate the core and ancestral function of PGL in oxygenic photosynthetic organisms. We inactivated the two genes that encode PGL-like polypeptides in the cyanobacterium Synechocystis PCC 6803, individually and in combination, and characterized and compared the resulting mutants to the wild-type (WT) parent strain using cultures grown under low-light (LL) and moderately high-light (HL) photoautotrophic growth conditions. The phenotypes of the three mutants led us to conclude that the products of these genes have similar but non identical roles in some process that serves to protect the organism from photooxidative damage. These Synechocystis mutants for the first time provide an experimental system in which the functions of specific PGL can be examined in a background devoid of other PGL.
Escherichia coli strain XL1-Blue MRF′ (Stratagene Cloning Systems, La Jolla, CA) and Synechocystis strain PCC 6803 were grown as described previously (5, 46). Agar plates containing Synechocystis cultures were illuminated continuously with ca. 1 or 10 microeinsteins·m−2·s−1 from a GE cool white fluorescent tube, with layers of cheesecloth used to reduce the irradiance to the desired level. For liquid cultures, continuous illumination of 10 or 150 microeinsteins·m−2·s−1 was provided by metal halide lamps (Sylvania Metalarc). Inocula used for experimental cultures were taken from cultures that had been grown for at least 10 generations under the same irradiance. Care was taken to periodically dilute cultures (the optical density at 730 nm [OD730] was not allowed to exceed 0.8) in order to minimize self-shading. Samples used for analyses were taken from batch cultures in mid-log phase (OD730 between 0.3 and 0.6).
The two genes that encode PGL-like polypeptides in Synechocystis strain PCC 6803 were inactivated by insertion of antibiotic resistance cassettes into their open reading frames (Fig. (Fig.1).1). Transformation and selection of Synechocystis mutants were as earlier described (5, 46), except that selection was carried out under “light-activated heterotrophic growth” conditions (a very low irradiance of ca. 1 microeinstein·m−2·s−1 with glucose added to the medium to provide a source of reduced carbon ) in case the mutants proved to be especially light sensitive. Complete segregation of each mutant was confirmed by agarose gel electrophoresis of PCR products.
The two pgl genes of Synechocystis were modified to produce polypeptides with an epitope tag appended at the C terminus. Plasmid pBS1539 (31), containing a tandem affinity purification (TAP) tag, was obtained from Bertrand Séraphin (CNRS, Centre de Génétique Moléculaire, Gif-sur-Yvette Cedex, France). The URA3 selectable marker gene (used for selection in yeast) was removed from pBS1539 by digestion of the plasmid with SmaI and XhoI and replaced with a kanamycin resistance (Kanr) gene, which was originally obtained from Tn903. The resulting plasmid was designated pBS1539Kan (Fig. (Fig.2A).2A). Constructs containing the Synechocystis PCC 6803 sll1568 gene (encoding PGL1) or slr1024 gene (encoding PGL2) followed by and fused in frame to the TAP tag, followed by the Kanr gene, and then followed by a fragment of Synechocystis genomic DNA that lies immediately downstream of sll1568 or slr1024 in the genome (Fig. (Fig.2B),2B), were created by overlap PCR (essentially as described by Murphy et al. ) using the primers listed in Table Table1.1. PCR products were purified by agarose gel electrophoresis, recovered using a GENECLEAN kit (Bio 101, Inc., Vista, CA), precipitated with ethanol to concentrate and sterilize them, and then used to transform Synechocystis PCC 6803. Replacement of the sll1568 or slr1024 gene with the corresponding TAP-tagged construct (Fig. (Fig.2B)2B) was confirmed by PCR using the “outer” primers.
Culture density was ascertained by measuring the optical density at 730 nm (OD730). The OD730 is a reasonably accurate proxy for cell number (10). Total carotenoid and chlorophyll were estimated spectrophotometrically in dimethylformamide extracts using the formulae of Hirschberg and Chamovitz (13). Molar ratios of carotenoid and chlorophyll were calculated by assuming molecular weights of 893.5 for chlorophyll and 570 for carotenoids.
For analyses of individual carotenoid and chlorophyll pigments, pigment extraction and analytical high-performance liquid chromatography (HPLC) were carried out essentially as described by Cunningham et al. (6). Pigments were identified on the basis of their retention times and absorption spectra compared to those of standards (β-carotene, echinenone, chlorophyll a, and zeaxanthin) and by reference to a previous analysis of carotenoids in Synechocystis (myxoxanthophyll) (38). Amounts of individual carotenoid pigments, relative to chlorophyll, were estimated from the HPLC peak area at 465 nm for each carotenoid relative to the peak area of chlorophyll a at 663 nm. Peak areas were ascertained using the HP ChemStation for LC software (Agilent Technologies, Santa Clara, CA). cis and trans geometrical isomers were quantified separately, but the results were combined for presentation of data.
Aliquots of Synechocystis cultures in the mid-log phase of growth were harvested by centrifugation, and the pellets were resuspended in fresh growth medium so that the concentrations of chlorophyll were the same. Cell samples were mixed with an equal volume of 2× sample buffer (125 mM Tris-HCl [pH 6.8], 4% [wt/vol] SDS, 20% [vol/vol] glycerol, 0.02% [wt/vol] bromophenol blue), and β-mercaptoethanol was then added to a final concentration of 5% (vol/vol). Each sample was heated at 95°C for 5 min, centrifuged at the maximum speed for 5 min in a microcentrifuge, and then loaded onto SDS-PAGE minigels. After separation by SDS-PAGE, the proteins were transferred to Immobilon-P membranes. The membranes were probed using a rabbit peroxidase-antiperoxidase soluble complex antiserum (Sigma P-2026; Sigma-Aldrich Co.) that bound to the protein A domains of the TAP tag, and then with Supersignal West Pico chemiluminescent substrate (Pierce Biotechnology, Inc., Rockford, IL) as described by Puig et al. (30). Chemiluminescence from sample bands was quantified using a FujiFilm LAS-3000 luminescent image analyzer (FujiFilm, Stamford, CT) with Image Reader software (version 2.2).
Samples used for transmission electron microscopy were fixed, dehydrated, and embedded in Epon resin as described by Poliquin et al. (29). Some of the glutaraldehyde-fixed samples (which were not secondarily fixed with OsO4) were dehydrated and embedded in LR White resin rather than Epon, and polymerization was catalyzed by UV light at −20°C. Sections were usually stained with aqueous uranyl acetate (1%, wt/vol) and in some cases were poststained with lead citrate and then were examined using a Zeiss EM10 CA electron microscope.
Samples were taken from mid-log-phase cultures, adjusted to an equal OD730 by dilution with growth medium, as needed, and then placed in the dark for 20 min. Glycerol was added to a final concentration of 75% (vol/vol), and samples were then transferred to nuclear magnetic resonance tubes (inside diameter, 4.2 mm) and frozen in liquid nitrogen. Fluorescence spectra (at 77K) were recorded with a Perkin Elmer L5 50B fluorimeter (slit width set to 5 nm for both excitation and emission). Excitation was at 438 nm (a chlorophyll absorption maximum). Multiple sample scans (≥5) were recorded and averaged to reduce noise. Spectra were recorded for samples derived from at least three independent experiments for each strain and each growth irradiance.
Two genes in the cyanobacterium Synechocystis sp. PCC 6803 (sll1568 and slr1024 [gene designations of Kaneko et al. ) encode polypeptides similar in sequence to plant plastoglobulins (PGL). These two genes, which are here referred to as pgl1 and pgl2, were inactivated individually and in combination by insertion of antibiotic resistance cassettes into convenient restriction enzyme sites in their open reading frames (Fig. (Fig.1).1). Three mutants were created: a Δpgl1 (sll1568::Kanr) mutant, a Δpgl2 (slr1024::Kanr) mutant, and a Δpgl1 Δpgl2 (sll1568::Specr slr1024::Kanr) double mutant, which for brevity is referred to as the Δpgl1+2 mutant. Complete segregation of each mutant was obtained (Fig. (Fig.1,1, lower right, and data not shown), demonstrating that the products of the pgl genes are not essential to the organism under the low-light photoheterotrophic growth conditions that were employed for selection and segregation of the mutants.
When grown photoautotrophically under a relatively low irradiance of continuous white light (10 microeinsteins·m−2·s−1) (LL conditions), the growth rates of the Δpgl mutants were indistinguishable from that of the wild-type (WT) parent strain. For cultures grown under moderately high-irradiance conditions (150 microeinsteins·m−2·s−1) (HL conditions), the growth rates of the Δpgl mutants were somewhat less than that of the WT, with division times of ca. 13 h for the Δpgl1+2 mutant, 11 h for the Δpgl1 mutant, 10 h for the Δpgl2 mutant, and 9 h for the WT. Differences in the appearance of the Δpgl mutant cultures were quite evident for cultures grown under HL conditions. Cultures of the Δpgl1 mutant and, even more, of the Δpgl1+2 mutant were a pronounced yellowish green (Fig. (Fig.3A).3A). Cultures of the Δpgl2 mutant, in contrast, were a darker green than WT cultures (not shown).
Amounts of chlorophyll and carotenoid pigments in mid-log-phase cultures of the Synechocystis WT and Δpgl mutants were ascertained using cultures that had been acclimated to and grown under LL or HL photoautotrophic growth conditions. When plotted relative to the chlorophyll content, the carotenoid content of the Δpgl mutants did not differ significantly from that of the WT for cultures grown under LL (Fig. (Fig.3B).3B). The carotenoid-to-chlorophyll ratio for WT cultures grown under HL conditions was essentially the same as that for WT cultures grown under LL. In contrast, each of the three Δpgl mutants contained significantly more carotenoid per unit of chlorophyll (mol/mol) under HL than under LL, with the Δpgl1+2 mutant containing, on average, three times more carotenoid per unit of chlorophyll in cultures grown under HL conditions (Fig. (Fig.3B3B).
The higher ratios of carotenoid to chlorophyll for Δpgl mutants grown under HL resulted primarily from increased amounts of myxoxanthophyll, with smaller contributions from zeaxanthin and β-carotene (Fig. (Fig.3C).3C). HL-grown Δpgl1+2 mutant cultures, for example, contained nearly six times more myxoxanthophyll per unit of chlorophyll than did LL-grown Δpgl1+2 mutant cultures and more than seven times more than HL-grown WT cultures. In contrast to the findings for the other carotenoid pigments, amounts of echinenone, relative to chlorophyll, did not differ significantly for HL- versus LL-grown cultures or for the WT versus Δpgl mutant cultures.
When the amounts were expressed relative to the culture density (as measured by the OD730, a good proxy for cell number ), some significant differences in the carotenoid and chlorophyll pigments of the Δpgl mutants versus the WT could be discerned even for cultures grown under LL. The Δpgl1+2 double mutant contained less chlorophyll and carotenoid per unit of culture density than did WT cultures grown under LL (Fig. (Fig.3E),3E), whereas the Δpgl2 mutant contained less total carotenoid (Fig. (Fig.3E).3E). For HL-grown cultures, it can be seen that the higher carotenoid-to-chlorophyll ratios of the Δpgl mutants, relative to those of the WT (Fig. (Fig.3B),3B), resulted primarily from a reduction in the cell content of chlorophyll rather than from an increase in the total amount of carotenoids (Fig. (Fig.3E).3E). Under HL, the Δpgl1+2 mutant had only about 40% of the chlorophyll per unit of OD730 in LL-grown Δpgl1+2 cultures and about one-half of the chlorophyll in HL-grown WT cells.
The pgl1 and pgl2 genes of Synechocystis were modified so as to produce polypeptides with a C-terminal epitope tag (the TAP tag described by Rigaut et al. ). This tag enabled a quantitative estimation of the amounts of the PGL1 and PGL2 proteins in cells of Synechocystis. Mid-log-phase HL-grown cultures were found to contain ca. five times more PGL1, relative to the amount of chlorophyll, than did LL-grown cultures (Fig. (Fig.4).4). The level of PGL2 was ca. 3-fold higher under HL versus LL conditions (data not shown). Because HL-grown cultures contained ca. 60% of the amount of chlorophyll in cells of LL-grown cultures (Fig. (Fig.3E),3E), the increases in PGL1 and PGL2 under HL were ca. 3-fold for PGL1 and 2-fold for PGL2 when the data were expressed on a per cell basis.
Cells from Synechocystis WT and Δpgl mutant cultures produced the expected 77K fluorescence emission maxima for photosystem I (PSI) (a broad peak at ca. 722 nm) and photosystem II (PSII) (peaks at ca. 685 and 695 nm) when excited with light absorbed primarily by chlorophyll (Fig. (Fig.5).5). Fluorescence emission from PSI, relative to that from PSII, typically diminishes for photoautotrophically grown cells that are moved from low- to high-irradiance conditions (10, 12), and such was observed in this study, with a ca. 50% decline for the WT strain (not shown). For both LL- and HL-grown cultures, there was much less emission at 722 nm relative to the peaks at 685 and 695 nm (Fig. (Fig.55 and data not shown) for fluorescence emission spectra produced by Δpgl1+2 mutant cells than was observed in the fluorescence emission spectra produced by WT cells, indicating a lower ratio of PSI to PSII reaction centers. The emission at 722 nm for the Δpgl1 and Δpgl2 mutants also was reduced compared to that for the WT under HL conditions (and for Δpgl2, but not for Δpgl1, also under LL conditions [data not shown]), but the reduction was not nearly as great as that for the Δpgl1+2 mutant.
Cells of the three Δpgl mutants of Synechocystis differed from the WT in their ultrastructure (Fig. (Fig.6).6). One or more enlarged vacuolated bodies were observed in many thin sections of the Δpgl1 mutant (Fig. (Fig.6B,6B, arrowhead). Similar bodies, which may be indicative of some type of non-electron-dense or nonstaining material, were also seen in sections of WT cells, but much less frequently than in sections of Δpgl1 mutant cells.
The morphological consequences of pgl gene inactivation were more conspicuous in thin sections of Δpgl2 and Δpgl1+2 mutant cells. Large vacuolated bodies that were partially filled with a fibrous, electron-dense material were observed in thin sections of cells of these mutants (Fig. 6C and D, black arrows). The electron-dense material in these vacuolated bodies was visible whether sections were not stained, secondarily stained with uranyl acetate alone, or stained with uranyl acetate in combination with lead citrate, suggesting that this fibrous material was inherently electron dense.
Vacuolated bodies in thin sections of the Δpgl1+2 mutant, but not of the Δpgl2 mutant, quite often contained nonfibrous accretions adjacent to the fibrous material more typically observed in these bodies (Fig. 6D and E, white arrows), indicating an accumulation of a second type of electron-dense material. It is noteworthy that vacuolated bodies containing fibrous and nonfibrous material were observed in cells of Δpgl1+2 mutant cultures grown under LL conditions as well as under HL conditions.
Cultures of Synechocystis PCC 6803 grown under a relatively low irradiance and then exposed to much higher irradiance exhibit certain characteristic phenotypic changes as cells acclimate to the new conditions. Among these changes are a reduction in cell pigmentation as the numbers of the photosynthetic reaction centers and the sizes of their associated antennae are reduced, transient or stable increases in the amounts of certain carotenoids, particularly myxoxanthophyll (10, 23, 34), and a reduction in the PSI/PSII ratio (10, 12). The three Δpgl mutants of Synechocystis exhibited this same suite of phenotypic changes under an irradiance lower than that required to produce it in cultures of the WT. Inactivation of pgl1 and pgl2 individually gave rise to roughly comparable phenotypes, but PGL1 and PGL2 must not have exactly the same function in cells of Synechocystis because the unusual electron-dense inclusion bodies that were observed in the Δpgl2 mutant (Fig. (Fig.6C)6C) were not observed in the Δpgl1 mutant.
The Δpgl1+2 mutant was especially sensitive to the light environment, with significant differences, relative to the WT, in pigmentation, ultrastructure, and PSI/PSII ratio under only 10 microeinsteins of light·m−2·s−1 (ca. 1/10 the irradiance needed to saturate photosynthesis ). Myxoxanthophyll, an elevated level of which is considered symptomatic of exposure to high light stress (10, 23, 34, 35), was ca. 3-fold higher, relative to the WT, for cultures of the Δpgl1+2 mutant grown under an irradiance of only 20 microeinsteins·m−2·s−1 (Fig. (Fig.3D3D).
The occurrence of genes that encode PGL-like polypeptides in virtually all cyanobacteria, algae, and plants implies a fundamental function for these polypeptides in oxygenic photosynthesis. The viability of the Synechocystis Δpgl1+2 mutant under phototrophic growth conditions makes clear that these polypeptides are not required for photosynthesis per se. The light-sensitive phenotypes of the various Synechocystis Δpgl mutants instead suggest a role for these polypeptides in some process relevant to the prevention and/or repair of photooxidative damage to the thylakoid membrane. It is perhaps telling that the few cyanobacteria with genomes that appear to lack genes encoding PGL-like polypeptides (Prochlorococcus strains MIT9211, MIT9303, MIT9313, and CCMP1375 [also known as strain SS120] and Synechococcus strain CC9605) typically reside in low-light environments and grow best under a relatively low irradiance (this is true at least for the four Prochlorococcus strains [25, 33]).
Photooxidative damage to the photosynthetic apparatus, particularly to the reaction centers of photosystem II, is a routine and inescapable consequence of oxygenic photosynthesis (22, 41) that occurs even under very-low-light conditions for Synechocystis (1). Cells of Synechocystis, as do those of other oxygenic photosynthetic organisms, have numerous ways of preventing, minimizing, and/or repairing photooxidative damage (21, 22, 26). These various mechanisms of protection and repair are unable to fully compensate for the absence of the PGL1 and PGL2 polypeptides in cells of Synechocystis, even for cultures grown under a relatively low irradiance.
The PGL-like polypeptides of cyanobacteria are exceedingly diverse in their amino acid sequences and cluster into five distinctly different groups (Fig. (Fig.7).7). Most cyanobacterial genomes contain one or two genes that encode PGL-like polypeptides, with Trichodesmium IMS101 and two Cyanothece species being exceptional in having three such genes in their genomes (Fig. (Fig.7).7). Many cyanobacteria with a gene that encodes a PGL of group I (including Synechocystis PGL1) also have a gene that encodes a PGL of group II (e.g., Synechocystis PGL2) and/or group IV, but no cyanobacterial genome so far available contains only genes for group II and/or group IV polypeptides. Similarly, certain cyanobacteria that contain a gene encoding a PGL of group III also have a gene encoding a PGL of group V, although there are species with only a gene for a group III or group V polypeptide.
Most of what is known about the functions of PGL has come from studies on the polypeptides of a few flowering plants. The 14 PGL of A. thaliana are included in the neighbor-joining tree of cyanobacterial PGL-like polypeptides that is shown in Fig. Fig.77 in order to facilitate a comparison of cyanobacterial and plant PGL. An alignment of the C-terminal portions of the amino acid sequences of the A. thaliana and Synechocystis PGL also is informative in this regard (Fig. (Fig.8).8). From Fig. Fig.77 and and88 it can be concluded that the Synechocystis PGL (as well as other cyanobacterial PGL of groups I, II, and III) are most similar in sequence to the A. thaliana PGL referred to as AtFib1a, AtFib1b, and AtFib2. The two Synechocystis PGL are, in fact, more similar to these three A. thaliana PGL than are most of the other PGL of A. thaliana.
A. thaliana AtFib1a, AtFib1b, and AtFib2 are among the most abundant PGL in the plastid lipid globules of A. thaliana chloroplasts (28). Most other plant PGL for which experimental information is available (including the fibrillin of Capsicum annuum , Solanum demissum C40.4 , Cucumis sativa CHRC , Pisum sativum PG1 , and Brassica rapa PAP1, PAP2, and PAP3 ) are highly similar in sequence to A. thaliana AtFib1 or AtFib2 and, in addition, have been shown to be constituents of plastid lipid globules.
There is evidence that AtFib1a and closely related PGL in other flowering plants are involved in some way in protecting the photosynthetic apparatus, specifically PSII, from photooxidative damage (24, 36, 47). Monte et al. (24) reported an association of a potato PGL with an antenna complex of PSII and found that antisense suppression of the gene encoding this protein resulted in stunted plants. Most intriguing, they observed a reduction in nonphotochemical quenching of chlorophyll fluorescence in the antisense plants, leading them to suggest that this PGL is involved in the dissipation of excess absorbed light energy. Yang et al. (47) also concluded that there is a photoprotective effect for a specific PGL on PSII. A. thaliana plants with AtFib1a downregulated by antisense suppression showed increased sensitivity to photoinhibition, whereas overexpression of the gene encoding AtFib1a led to an enhanced phototolerance. More recently, a C. annuum PGL (the “fibrillin” of Deruère et al. ) was shown to protect isolated thylakoid membranes against uncouplers and to delay disassembly of the photosynthetic membrane during a developmental transition from chloroplast to chromoplast in tomato fruits (36).
The plant PGL that are most closely related to PGL1 and PGL2 of Synechocystis PCC 6803 are bona fide constituents of plastid lipid globules yet have, as well, been implicated in photoprotection. Lipid globules have been reported to be present in cells of Synechocystis and other cyanobacteria, albeit at relatively low levels in actively growing cultures (ca. 20 globules per cell in Synechocystis PCC 6803, localized predominantly at the periphery of the cell between the cytoplasmic and thylakoid membranes, according to van de Meene et al. ). Do Synechocystis PGL1 and/or PGL2 reside in the lipid globules in the cells of this cyanobacterium? Are there any lipid globules in cells of the Δpgl1+2 mutant? Is the formation of lipid globules relevant or integral to the photoprotective functions of Synechocystis PGL1 or PGL2?
We employed immunological methods to localize PGL1 and PGL2 in cells of Synechocystis. First, thin sections of the Synechocystis mutants that were engineered to produce epitope-tagged PGL1 and PGL2 (Fig. (Fig.4)4) were probed with an antiserum specific for the epitope tag. Second, thin sections of the Synechocystis WT were probed with affinity-purified antisera specific for the PGL1 and PGL2 polypeptides. These experiments did not indicate a particulate (i.e., in lipid globules) or membrane localization for the Synechocystis PGL but rather a more dispersed distribution throughout the cell (not shown).
The second question still remains to be answered. Lipid globules could not be identified in thin sections with any certainty, and the paucity of lipid globules in Synechocystis (a single thin section of the WT strain would be expected to contain at most a single lipid globule) makes it quite difficult to verify their absence. Determination of the presence of lipid globules in the Δpgl1+2 mutant requires alternative approaches.
We favor the hypothesis that the ancestral and primary function of polypeptides of the plastoglobulin family was to bind and remove or sequester within the photosynthetic membrane various damaged and/or unneeded and potentially toxic fatty acids, lipids, proteins, quinones, phytol, and other lipid-soluble molecules. The formation of lipid globules may be an inherent aspect of this core and fundamental function or may instead reflect a secondary or derived (albeit valuable and important) role for PGL in the plastids of plants, one that has made possible the proliferation of carotenoid-containing lipid globules in chromoplasts, the accumulation of oil-bearing lipid bodies in elaioplasts, and the disassembly of the thylakoid membrane and recapture of valuable nutrients during senescence.
The data presented here make it clear that PGL-like polypeptides are not essential for photoautotrophic growth of Synechocystis PCC 6803, but the in absence has consequences that are indicative of roles for these polypeptides in protection and/or repair of the photosynthetic membrane. The Synechocystis mutants in which one or both pgl genes are inactivated provide a unique and powerful experimental system that, for the first time, enables investigation of the functions of individual PGL in a background devoid of other PGL that might have overlapping or redundant functions.
This work was supported in the initial stages by grant MCB-0316448 from the National Science Foundation to F.X.C. and E.G. A.B.T. and C.P. acknowledge the support of the University of Maryland Department of Cell Biology and Molecular Genetics Honors Program. C.P. is grateful for support provided by the Howard Hughes Medical Institute Undergraduate Research Fellowship Program of the College of Chemical & Life Sciences at the University of Maryland.
We thank Susan Golden (University of California at San Diego), Bertrand Séraphin (Centre National de la Recherche Scientifique, Gif-sur-Yvette Cedex, France), and Wim Vermaas (Arizona State University) for providing plasmids or cultures used in this work.
Published ahead of print on 14 January 2010.