This Article Abstract Full Text (PDF) All Versions of this Article: 15/4/992    most recent tpc.010405v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (44) Citing Articles via Google Scholar Google Scholar Articles by Baroli, I. Articles by Niyogi, K. K. Search for Related Content PubMed PubMed Citation Articles by Baroli, I. Articles by Niyogi, K. K. Agricola Articles by Baroli, I. Articles by Niyogi, K. K.

Zeaxanthin Accumulation in the Absence of a Functional Xanthophyll Cycle Protects Chlamydomonas reinhardtii from Photooxidative Stress

Department of Plant and Microbial Biology, University of California, Berkeley, California 94720-3102

² To whom correspondence should be addressed. E-mail niyogi{at}nature.berkeley.edu; fax 510-642-4995

	Abstract

TOP Abstract INTRODUCTION RESULTS DISCUSSION METHODS References

 
Xanthophylls participate in light harvesting and are essential in protecting the chloroplast from photooxidative damage. To investigate the roles of xanthophylls in photoprotection, we isolated and characterized extragenic suppressors of the npq1 lor1 double mutant of Chlamydomonas reinhardtii, which lacks zeaxanthin and lutein and undergoes irreversible photooxidative bleaching and cell death at moderate to high light intensities. Here, we describe three suppressor strains that carry point mutations in the coding sequence of the zeaxanthin epoxidase gene, resulting in the constitutive accumulation of zeaxanthin in a range of concentrations. The presence of zeaxanthin in these strains was sufficient to prevent photooxidative damage in the npq1 lor1 background. The size of the light-harvesting antenna in the suppressors decreased in high light in a manner that was proportional to the relative content of zeaxanthin, with the strain having the most zeaxanthin showing a severe reduction in levels of the major light-harvesting complex II proteins in high light. We show that the effect of constitutive zeaxanthin on light harvesting is not the main cause of increased photoprotection, because in the absence of zeaxanthin, a strain with a smaller light-harvesting antenna showed only minor protection against photobleaching in high light. Furthermore, the zeaxanthin-accumulating suppressors were able to tolerate higher levels of exogenous reactive oxygen than their parental strain under conditions that did not affect light harvesting. Our results are consistent with an antioxidant role of zeaxanthin in the quenching of singlet oxygen and/or free radicals in the thylakoid membrane in vivo.

	INTRODUCTION

TOP Abstract INTRODUCTION RESULTS DISCUSSION METHODS References

 
Maintaining a balance between the capture and the use of light energy is essential for the survival of all oxygenic photosynthetic organisms. When absorption of light energy exceeds the capacity for photosynthesis, plants are at risk of photooxidative damage to the chloroplast (Foyer et al., 1994). Excess absorbed light energy can lead to the overreduction of components of the photosynthetic electron transport chain, and this can result in sequential one-electron transfers to ground-state oxygen, producing the superoxide anion, hydrogen peroxide, and the hydroxyl radical. In the light-harvesting antenna of photosystem II (PSII), unused light energy increases the lifetime of the singlet excited state of chlorophyll (¹Chl*), which can be converted to the triplet excited state (³Chl*) through the photophysical process of intersystem crossing. ³Chl* itself is not harmful, but it is long-lived and can transfer energy to ground-state oxygen to generate highly reactive singlet oxygen (¹O₂*). Under extreme light stress, when the quenching capacity of the chloroplast presumably has been overwhelmed, net formation of ¹O₂* has been reported (Hideg et al., 1998). ¹O₂*, superoxide, and other reactive forms of oxygen irreversibly oxidize lipids, proteins, and pigments in their immediate vicinity. Thylakoid lipids are especially susceptible to oxidative damage because of the abundance of unsaturated fatty acid side chains. Reactive oxygen attack of these lipids initiates peroxyl-radical chain reactions, which eventually can destroy the thylakoid membrane (Knox and Dodge, 1985). In the PSII reaction center, the formation of ¹O₂* under excess light is thought to be the cause of direct damage to structural protein components, which require de novo protein synthesis to be repaired (Melis, 1999).

To cope with imbalances between the absorption of excitation energy and its use, algae and plants regulate photosynthetic light harvesting and make use of a series of antioxidant molecules and enzymes to detoxify reactive oxygen species and free radicals once they have been formed (reviewed by Niyogi, 1999). Some of the xanthophylls (oxygenated carotenoids) present in the thylakoid membrane participate in both of these protective functions. In particular, zeaxanthin and lutein are necessary under high-light stress for the efficient transition of the light-harvesting complexes (LHCs) of PSII from a conformation that favors light harvesting to one that allows for thermal dissipation of part of the excess excitation energy (Niyogi et al., 1997b, 2001; Pogson et al., 1998). This thermal dissipation of excess light energy, which might involve a direct quenching of excess ¹Chl*, is measured and often referred to as nonphotochemical quenching (NPQ) of chlorophyll fluorescence (Horton et al., 1996; Müller et al., 2001).

Because carotenoids are distributed in the thylakoid pigment-protein complexes in close proximity to the chlorophylls, and thus to the potential sites of ¹O₂* formation (Yamamoto and Bassi, 1996), they are the most important quenchers of electronically excited states in the thylakoid membrane. ³Chl* or ¹O₂* transfers its excitation energy to a nearby carotenoid molecule to form a carotenoid triplet that decays harmlessly to the ground state by thermal dissipation. Carotenoids also can protect against lipid peroxidation by reacting with free radicals directly (Palozza and Krinsky, 1992), forming a carotenoid radical that could be regenerated by interaction with tocopherols and ascorbate in the lipid phase of the membrane (Edge et al., 1997). The importance of carotenoids in the protection of plants from photooxidative stress is evident from the fact that mutants that lack carotenoids, or plants in which carotenoid synthesis has been inhibited, are completely unable to sustain growth in the light in the presence of oxygen (Oelmüller, 1989).

In plant and algal mutants with altered xanthophyll composition (Pogson et al., 1996, 1998; Niyogi et al., 1997b) and in experiments with purified recombinant LHCs in vitro (Bassi and Caffarri, 2000), it has been shown that some of the carotenoid binding sites in the LHC proteins are interchangeable. Despite this flexibility in the xanthophyll composition of photosynthetic organisms, it is becoming increasingly clear that the different xanthophylls vary in their ability to quench reactive intermediates and provide photoprotection (Niyogi et al., 1997b; Pogson et al., 1998). Many factors contribute to these differences, including the number of conjugated double bonds, which influences the excited-state energies of carotenoids. In addition, the polarity and stereochemistry of a carotenoid influence its position in the thylakoid membrane or in the LHC, and this in turn influences the chlorophyll-carotenoid distances that play a key role in energy transfer reactions. For this reason, the roles of the different xanthophylls in photoprotection cannot be extrapolated directly from their measured characteristics in organic solvents. In the past few years, the isolation of several mutants with specific defects in xanthophyll biosynthesis has helped clarify those individual roles (reviewed by Baroli and Niyogi, 2000).

Some of those studies have shown that lutein, the most abundant xanthophyll in the thylakoid membrane, is not essential for light harvesting and photoprotection when zeaxanthin is present. Lutein can be replaced in the LHC by other xanthophylls, such that the size of the xanthophyll pool with respect to chlorophyll remains unchanged, and lutein-deficient mutants show no apparent defect in light harvesting under laboratory conditions (Pogson et al., 1996). However, lutein deficiency has been shown to alter the molecular organization of the LHC in Arabidopsis (Lokstein et al., 2002), Chlamydomonas reinhardtii (Chunaev et al., 1991), and Scenedesmus obliquus (Bishop, 1996). Analysis of the Chlamydomonas lor1 mutant demonstrated that lutein contributes substantially to the development of NPQ (Niyogi et al., 1997b), although it is not clear whether the pigment plays a direct role in ¹Chl* quenching (Niyogi et al., 1997b) or whether the aberrant conformation of the LHC caused by the lack of lutein prevents the development of NPQ (as discussed by Lokstein et al., 2002). Lutein content increases when Chlamydomonas cells are exposed to light stress, but lutein deficiency did not cause enhanced susceptibility to light stress, at least under the relatively moderate high-light conditions tested (Niyogi et al., 1997b).

Chlamydomonas and Arabidopsis mutants that contain only the xanthophylls violaxanthin and neoxanthin are subject to photobleaching under light stress, suggesting that these xanthophylls do not perform a sufficient protective function in the chloroplast (Niyogi et al., 1997b, 2001). Among all the xanthophylls, zeaxanthin is the only one that accumulates exclusively under excess light, by deepoxidation of existing violaxanthin in the so-called xanthophyll cycle (Eskling et al., 1997) (Figure 1). It is widely thought that the main function of zeaxanthin is as a quencher of the ¹Chl* state through the process of NPQ (Müller et al., 2001). However, it has been suggested that zeaxanthin may protect from light stress by directly quenching ¹O₂* and free radicals (Havaux and Niyogi, 1999) and by making the thylakoid membrane less permeable to oxygen (Gruszecki, 1999).

View larger version (21K): [in this window] [in a new window]   Figure 1. The Xanthophyll Biosynthetic Pathway in Chlamydomonas. The reactions impaired in the npq1, npq2, and lor1 mutants are indicated.

A fruitful strategy to identify interacting components of a complex biological system for which mutants are available is to isolate extragenic mutations that modify (by either enhancing or suppressing) a mutant phenotype (Prelich, 1999). The npq1 lor1 double mutant of Chlamydomonas is particularly suitable for extragenic suppressor analysis. The individual mutations are very stable and likely null alleles, and the bleaching phenotype imposes a strong selective pressure for suppressors that survive in high light. At least three types of suppressor mutations can be envisioned. Mutations that affect xanthophyll metabolism may lead to the accumulation of precursors not found normally in the chloroplast that are able to quench reactive oxygen more efficiently than the epoxidized xanthophylls violaxanthin and neoxanthin. Another type of suppressor mutation could cause the increased accumulation of nonxanthophyll antioxidants, which may fulfill the antioxidant role normally performed by lutein or zeaxanthin. A third kind of suppressor mutation could interfere with LHC assembly, decreasing the antenna size and preventing the formation of excess ¹Chl* under light stress (Hippler et al., 2000).

Here, we describe suppressors that affect xanthophylls and present evidence that constitutive accumulation of zeaxanthin, independent of the function of the xanthophyll cycle and in the absence of lutein, is sufficient to prevent photobleaching under light or other oxidative stress conditions in Chlamydomonas.

	RESULTS

TOP Abstract INTRODUCTION RESULTS DISCUSSION METHODS References

 
Isolation of Suppressors of npq1 lor1
Our physiological experiments were performed on Chlamydomonas cells grown photoautotrophically in minimal medium under two light conditions, 50 µmol·m^-2·s^-1 (low light [LL]) and 500 µmol·m^-2·s^-1 (high light [HL]). It is known that in HL, the wild type shows no signs of light stress; rather, it exhibits an acclimation response characterized by an increased number of photosynthetic units, an increased PSII:PSI stoichiometry (Neale and Melis, 1986), and a greater degree of xanthophyll deepoxidation (Niyogi et al., 1997b) relative to LL conditions. Figure 2 shows the growth phenotypes of the wild type and the zeaxanthin- and lutein-deficient npq1 lor1 double mutant in minimal medium under the two light conditions. Both strains grew in LL, but the double mutant showed no growth in HL, in contrast to the wild type. The HL sensitivity of npq1 lor1 was rescued partially by growth under a lower oxygen tension (data not shown). Growth in acetate-containing medium, however, did not prevent cell death, although bleaching was retarded (data not shown). These results suggest that excess reactive oxygen species, such as ¹O₂*, that are generated in the npq1 lor1 double mutant in HL affect essential cellular processes besides photosynthesis.

View larger version (37K): [in this window] [in a new window]   Figure 2. Growth of Chlamydomonas Strains in Minimal Medium under Continuous Light. (A) Cells grown in LL (50 µmol·m-2·s-1) for 7 days. (B) Cells grown in HL (500 µmol·m-2·s-1) for 7 days.

Three independently isolated suppressors, which exemplify varying degrees of constitutive accumulation of zeaxanthin, were selected for further genetic and physiological characterization. As shown below, these suppressed strains were found to contain mutated alleles of the NPQ2 gene; therefore, they are called npq2-2 npq1 lor1, npq2-3 npq1 lor1, and npq2-4 npq1 lor1, in decreasing order of relative zeaxanthin accumulation. As shown in Figure 2, all three suppressed strains grew at approximately wild-type rates in both LL and HL under photoautotrophic conditions.

Pigment Content and Photosynthetic Characteristics of Suppressors
Figure 3 compares the relative content of -carotene–derived xanthophylls in wild-type and mutant strains grown under either LL or HL conditions. When the wild-type strain was grown in HL, the content of deepoxidized xanthophylls increased significantly compared with that in LL, such that the xanthophyll cycle deepoxidation state ([zeaxanthin + antheraxanthin]/[violaxanthin + antheraxanthin + zeaxanthin], [Z+A]/[V+A+Z]) increased from 0.11 ± 0.01 in LL to 0.62 ± 0.10 in HL (Figures 3A and 3B). Also in the HL-grown wild type, there was a doubling of the total xanthophyll cycle pool relative to chlorophyll a (Chl a) (Figure 3A). The double mutant npq1 lor1 showed a deepoxidation state of 0.20 ± 0.02 under continuous LL, which is greater than in the LL-grown wild type (Figure 3B). This increased deepoxidation state in npq1 lor1 is not attributable to the light-induced activity of violaxanthin deepoxidase, which is blocked by the npq1 mutation, but is attributable to competition between LHC proteins and zeaxanthin epoxidase for the binding of zeaxanthin and antheraxanthin (I. Baroli and K.K. Niyogi, unpublished results). The greater xanthophyll cycle pool size in npq1 lor1 (Figure 3A) is the result of increased metabolic flux into the -carotene branch caused by the lor1 mutation (Niyogi et al., 1997b).

View larger version (25K): [in this window] [in a new window]   Figure 3. Pigment Characteristics of Chlamydomonas Strains. (A) Relative content of xanthophyll-cycle pigments. (B) Deepoxidation state, (Z+A)/(V+A+Z). (C) Relative content of -carotene. Cells were grown on plates in minimal medium under either continuous LL or HL. Values shown are means of three to five independent experiments. In (A), the standard deviation was 15% of the means. In (B) and (C), the error bars represent standard deviations.

The xanthophyll composition of the suppressors suggested that these strains are partially impaired, to different degrees, in the activity of zeaxanthin epoxidase (Figure 1). The npq2-2 npq1 lor1 strain had undetectable levels of neoxanthin and violaxanthin, and the deepoxidation state was 1.0 irrespective of growth conditions (Figure 3B), indicating a severe block in zeaxanthin epoxidation. The block was not complete in this strain, though, because low levels of antheraxanthin accumulated in both light conditions (Figure 3A). The other two suppressor strains, npq2-3 npq1 lor1 and npq2-4 npq1 lor1, showed intermediate levels of xanthophyll epoxidation with respect to the parental strain and the more extreme case of npq2-2 npq1 lor1. They showed some change in deepoxidation state when grown in HL, but they maintained high levels of zeaxanthin and antheraxanthin under both light conditions (Figure 3B).

Table 1 compares the photosynthetic performances of the wild type and the strains with xanthophyll deficiencies. The wild-type strain showed the expected response to light (Neale and Melis, 1986), with a slightly increased Chl a/Chl b ratio, twofold lower cellular chlorophyll content, and decreased photosynthetic efficiency in HL compared with LL. All of these changes are thought to reflect an acclimation of the photosynthetic apparatus to the ambient irradiance (MacIntyre et al., 2002). Compared with the wild type under each light condition, the suppressors showed alterations in cellular chlorophyll content and in Chl a/Chl b ratios. The degree of change was approximately proportional to the xanthophyll deepoxidation index. The npq2-2 npq1 lor1 and npq2-3 npq1 lor1 suppressors showed significantly decreased chlorophyll content in both LL and HL. In npq2-4 npq1 lor1, the strain that retained higher levels of violaxanthin, the chlorophyll content was similar to that of the wild type in both light conditions. All three suppressor strains had an increased Chl a/Chl b ratio, and this was more evident in HL, with an almost twofold increase in the ratio between LL and HL conditions. Because the ratio of the two chlorophylls did not change significantly in the HL-grown lor1 and npq1 single mutants (Niyogi et al., 1997b), we conclude that the extreme Chl a/Chl b values observed in HL in the triple mutants are caused by the lack of violaxanthin and/or neoxanthin, which may make the LHC polypeptides more unstable. The Chlamydomonas npq2 lor1 mutant has been shown to have a smaller PSII light-harvesting antenna than the wild type (Polle et al., 2001), and we observed the same effect in the suppressors (see below).

When grown in LL, the suppressors with high degrees of zeaxanthin accumulation (npq2-2 npq1 lor1 and npq2-3 npq1 lor1) showed a significant decrease in the efficiency of photochemical conversion at PSII, measured as the chlorophyll fluorescence parameter F_v/F_m (maximum photochemical efficiency of PSII in the dark-adapted state; Table 1). The suppressor npq2-4 npq1 lor1, which has a xanthophyll cycle pigment content more similar to that of the wild type, showed wild-type values of F_v/F_m. Induction of NPQ during the exposure of LL-grown cells to 1300 µmol·m^-2·s^-1 remained aberrant in the npq2 npq1 lor1 strains (data not shown), consistent with measurements of NPQ in the zeaxanthin-accumulating npq2-1::ARG7 mutant of Chlamydomonas (Niyogi et al., 1997a).

Genetic Characterization of the Zeaxanthin-Accumulating Suppressors
To determine the genetic basis of the suppression phenotype in the zeaxanthin-accumulating strains, we backcrossed these mutants to the corresponding npq1 lor1 strain of the opposite mating type (Table 2) and analyzed progeny from dissected tetrads for their resistance to HL and their pigment composition. Figure 4 shows an example of tetrad analysis in a backcross of npq2-3 npq1 lor1 mt- and npq1 lor1 mt+. In crosses of all of the suppressed strains, the HL-bleaching phenotype segregated with a 2:2 ratio, demonstrating that the suppression of HL sensitivity results from a single nuclear mutation. HPLC analysis of the meiotic progeny showed perfect cosegregation of the HL resistance phenotype and the zeaxanthin accumulation phenotype (Table 2).

View larger version (58K): [in this window] [in a new window]   Figure 4. Genetic Analysis of a Suppressor of the HL Sensitivity of npq1 lor1. An example of tetrad analysis from a backcross between the parental strain npq1 lor1 arg7-1 mt+ and npq2-3 npq1 lor1 arg7-8 mt- (sup). Progeny from four tetrads were grown under continuous LL (50 µmol·m-2·s-1) or HL (500 µmol·m-2·s-1) in minimal medium for 7 days. Each vertical column represents a tetrad (labeled a–d). Both parental strains grown under the same conditions are shown as controls.

For the other two suppressors, we tested allelism by complementation analysis with the npq2-2 allele. Stable vegetative diploid strains were recovered as Arg prototrophs after crosses between npq2-2 npq1 lor1 arg7-1 mt+ and either npq2-3 npq1 lor1 arg7-8 mt- or npq2-4 npq1 lor1 arg7-8 mt-. As shown in Table 3, HPLC analysis of the pigment content of the diploid strains showed that the characteristic accumulation of violaxanthin of the initial npq1 lor1 strain was not recovered, demonstrating that there was no complementation of the zeaxanthin epoxidation defect in the diploid strains. Therefore, we conclude that the three suppressor strains carry mutated alleles of the NPQ2 gene that confer different degrees of zeaxanthin epoxidation.

The Molecular Nature of the npq2 Mutations
The four mutant alleles of npq2 (npq2-1::ARG7 and the three npq2 suppressors of npq1 lor1) showed a defect in the epoxidation of zeaxanthin to antheraxanthin and violaxanthin, so we sought to determine whether the NPQ2 gene corresponds to the structural gene that encodes the chloroplast-localized enzyme zeaxanthin epoxidase. Chlamydomonas ESTs with similarity to known zeaxanthin epoxidase genes from plants (Marín et al., 1996) were found in the GenBank database, and full-length cDNA sequences were obtained for genes from Chlamydomonas and from the marine Chlamydomonas strain W80 (Miyasaka et al., 2000). The deduced amino acid sequence of the open reading frame in the Chlamydomonas cDNA clone consisted of 763 amino acid residues (Figure 5) with a predicted molecular mass of 81 kD. The cDNA from strain W80 exhibited 66.7% identity to the Chlamydomonas sequence at the DNA level, and it contained an open reading frame of 727 amino acids (Figure 5).

View larger version (98K): [in this window] [in a new window]   Figure 5. Alignment of the Deduced Amino Acid Sequences of Zeaxanthin Epoxidase Precursor Proteins from Chlamydomonas and Selected Plant Species. Residues that are identical in at least three of the sequences are shaded in black. The long monooxygenase domain, which contains the putative ADP and flavin adenine dinucleotide binding sites, is shown with a solid underline, the two lipocalin conserved motifs are shown with a dotted underline, and the FHA (Forkhead-associated) domain is shown with a dashed underline. The mutation sites in the npq2-3 and npq2-4 alleles are indicated with asterisks. The mutation in npq2-2 causes a change in the translation initiation codon.

Genomic DNAs from the wild type and the npq2 mutants were analyzed for restriction fragment length polymorphisms (RFLPs) that affect the zeaxanthin epoxidase gene by DNA gel blot hybridization. The npq2-1::ARG7 mutation was isolated after insertional mutagenesis with plasmid DNA containing the argininosuccinate lyase (ARG7) gene and was shown to be caused by the insertion of the ARG7 plasmid at a single locus (Niyogi et al., 1997a). Figure 6 shows that npq2-1::ARG7 exhibited RFLPs with two different restriction enzymes, suggesting that it does contain a DNA insertion in the zeaxanthin epoxidase gene. By contrast, the other three npq2 alleles present in the zeaxanthin-accumulating suppressors showed no RFLPs with any of the enzymes tested compared with wild-type DNA (Figure 6). These results suggest that in the suppressors, the spontaneously arising mutations in the zeaxanthin epoxidase gene probably are not caused by transposon insertions or by large deletions or rearrangements in the coding region of the gene and that the mutant phenotypes may be attributable to point mutations.

View larger version (61K): [in this window] [in a new window]   Figure 6. DNA Gel Blot Analysis of Wild-Type Chlamydomonas, the npq1 lor1 Mutant, and Strains Carrying Mutations in the NPQ2 Gene. Genomic DNA was digested with restriction enzymes and probed with a 1.8-kb fragment corresponding to the 5' portion of the zeaxanthin epoxidase cDNA. (A) DNA digested with PstI. (B) DNA digested with BamHI and SpeI.

View larger version (24K): [in this window] [in a new window]   Figure 7. Structure of the Chlamydomonas Zeaxanthin Epoxidase Gene and Positions of Point Mutations. (A) Scheme of the zeaxanthin epoxidase locus in Chlamydomonas. The gene has 10 exons (indicated in black) and 9 introns (indicated in white). The 5' and 3' untranslated regions are indicated by the thinner boxes. The positions of the three point mutations in the npq2 suppressors are indicated by arrows. (B) Sequence of the zeaxanthin epoxidase gene in the regions surrounding the point mutations showing the amino acid changes.

View larger version (22K): [in this window] [in a new window]   Figure 8. Lipid Peroxidation in Suppressors of npq1 lor1 after Transfer to High Light. Exponentially growing liquid cultures were transferred from LL to HL conditions at time 0. Lipid peroxides were measured as thiobarbituric acid–reactive substances (TBARS). The results are shown relative to the initial culture content of TBARS, which was similar in all strains tested. Data are from three independent cultures. Error bars represent standard deviations and are not shown if smaller than the symbols.

Immunoblot analysis of the npq2 npq1 lor1 suppressor strains showed that the amounts of LHCII polypeptides were affected in LL in a manner that correlated with the relative content of zeaxanthin (Figure 9). That is, higher deepoxidation states of the xanthophyll cycle pool were associated with lower contents of LHC polypeptides. This effect was more extreme when the cells were grown under continuous HL, such that the content of the major LHC polypeptides was reduced to <20% of wild-type levels in the npq2-2 npq1 lor1 strain. These decreased levels of LHC polypeptides were consistent with the increased Chl a/Chl b ratios observed in the suppressors (Table 1).

View larger version (78K): [in this window] [in a new window]   Figure 9. Protein Analysis of Chlamydomonas Whole Cell Extracts. Cells were grown for several generations under continuous LL or HL conditions, except for mutants npq1 lor1 and cbn1 npq1 lor1, which were grown only in LL. Lanes were loaded with an equal number of cells. (A) Immunoblot analysis of the LHC of PSII. The major LHC polypeptides of PSII (LHCII) were visualized with the anti-P17 antibody (Bassi and Wollman, 1991). (B) Coomassie blue staining of a protein gel identical to the one used in (A) for immunoblot analysis.

View larger version (32K): [in this window] [in a new window]   Figure 10. Effect of Light-Harvesting Antenna Size on the HL Sensitivity of npq1 lor1. The cbn1 mutation, which causes an impairment in Chl b biosynthesis, was introduced in the npq1 lor1 genetic background, and the resulting strain was tested for survival in HL. The wild-type and npq1 lor1 strains shown here are the same strains shown in Figure 1 and are included for comparison. (A) Cells grown in minimal medium under continuous LL (50 µmol· m-2·s-1). (B) Cells grown in minimal medium under continuous HL (500 µmol· m-2·s-1).

View larger version (49K): [in this window] [in a new window]   Figure 11. Response of Wild-Type and Mutant Strains of Chlamydomonas to Exogenously Generated Singlet Oxygen. (A) Cells grown in minimal medium under continuous LL. (B) Cells grown in minimal medium containing 3.8 µM rose bengal under continuous LL. Because of strain variability in the response to rose bengal, each of the zeaxanthin-accumulating suppressors is shown with the corresponding parental strain as a control. Strain npq1 lor1 arg7-8 is the parent of the suppressor carrying the npq2-2 allele, and strain npq1 lor1 arg7-1 is the parent of the suppressors carrying the npq2-3 and npq2-4 alleles. Each row represents a 10-fold serial dilution of cells at the time of plating. The total number of cells plated is indicated at left.

	DISCUSSION

TOP Abstract INTRODUCTION RESULTS DISCUSSION METHODS References

In contrast to Chlamydomonas, the lut2 aba1 mutant of Arabidopsis, which has a pigment composition analogous to that of the Chlamydomonas npq2-2 npq1 lor1 mutant described here, showed reduced seedling viability, poor growth, and a virescent phenotype (Pogson et al., 1998). In a developing plant seedling, the lack of specific xanthophylls may affect the assembly of the light-harvesting antenna during the greening process, thereby altering the further development of the plant. Chlamydomonas cells are capable of chlorophyll synthesis in the dark and do not undergo greening when grown in continuous illumination, so the effect of altered xanthophyll composition might not be as apparent. Also, because violaxanthin and neoxanthin can be precursors of abscisic acid (Schwartz et al., 1997), the phenotype of the double mutant lut2 aba1 could be attributable in part to the lack of this hormone rather than to a direct effect of the lack of violaxanthin and neoxanthin, although the delayed greening of the Arabidopsis lut2 aba1 mutant was not rescued by exogenous abscisic acid (Pogson et al., 1998). As in the Chlamydomonas mutants, Arabidopsis plants that lack the major xanthophylls had increased Chl a/Chl b ratios and decreased chlorophyll content.

Most of the zeaxanthin-accumulating suppressors also accumulated antheraxanthin, but no strains that accumulated only antheraxanthin were recovered in our screen. These results are consistent with a defect in zeaxanthin epoxidase, which catalyzes the epoxidation of both -rings of zeaxanthin to produce sequentially antheraxanthin and violaxanthin (Eskling et al., 1997). Zeaxanthin epoxidase is located on the stromal side of the thylakoid membrane, and it is thought that some flexibility must exist in the membrane orientation of the substrates violaxanthin and antheraxanthin for the enzyme to complete the epoxidation of both -rings. Zeaxanthin epoxidase activity is almost completely absent in the suppressor that carries the npq2-2 allele. The other two alleles in the suppressors, npq2-3 and npq2-4, allow for incomplete epoxidation of zeaxanthin.

By determining the molecular basis for the defects in four npq2 alleles, we showed that NPQ2 is synonymous with the zeaxanthin epoxidase structural gene in Chlamydomonas. Furthermore, the severity of the epoxidation defects in these alleles can be explained by the nature of the mutation in each case. The reference allele, npq2-1::ARG7, contains an inactivating DNA insertion in the zeaxanthin epoxidase gene (Figure 6) that cosegregates with ARG7 in a cross (Niyogi et al., 1997a). The npq2-1::ARG7 mutation completely blocks zeaxanthin epoxidation. The npq2-2 mutation, which allows the synthesis of very low levels of antheraxanthin, affects the predicted initiation codon of the zeaxanthin epoxidase mRNA, changing it from AUG to GUG (Figure 7). In the absence of an alternative in-frame Met codon, it is likely that a very low level of translation initiation occurs, resulting in very low levels of zeaxanthin epoxidase enzyme expression. Indeed, site-directed mutagenesis of the start codon has been used to attenuate the expression of essential chloroplast genes in Chlamydomonas (Chen et al., 1995; Majeran et al., 2000). The mutations in npq2-3 and npq2-4 affect the conserved monooxygenase domain of zeaxanthin epoxidase (Figure 5), leading to partial losses of function most likely by affecting enzyme kinetics or accumulation. Consistent with the observed deepoxidation states in vivo (Figure 3B), substitution of Asp instead of Gly in npq2-3, which introduces a negative charge into an otherwise neutral to positively charged region, might be expected to impair function to a greater extent than the more conservative replacement of a nearby Phe by Leu in npq2-4. A more thorough structure-function analysis of zeaxanthin epoxidase could be performed by sequencing additional independent npq2 alleles recovered in our suppressor screen.

Suppressors with an npq2 mutant phenotype were recovered in our screen with high frequency (4 x 10^-5). Results of a fluctuation test suggest that most of the suppressor mutations are induced by the HL treatment, rather than arising spontaneously in the LL culture before selection (B. Gutman and K.K. Niyogi, unpublished results). Of nine npq2 suppressor strains that were tested, none was found to carry insertions of the Chlamydomonas transposons Tcr1, Tcr2, or Tcr3 in the zeaxanthin epoxidase gene (I. Baroli and K.K. Niyogi, unpublished results). Thus, higher rates of transposon insertions at the NPQ2 locus do not appear be the reason for the high frequency of suppressors. However, we cannot exclude the possibility that the NPQ2 locus may be a "hot spot" for point mutations. Zeaxanthin epoxidase is a large protein with two conserved cofactor binding domains and regions involved in protein–protein interactions, and point mutations that cause even a minor impairment of enzyme function could lead to levels of constitutive zeaxanthin accumulation that are sufficient to allow the growth of npq1 lor1 in HL.

How Does Zeaxanthin Accumulation Protect Cells from Photooxidation?
Zeaxanthin has multiple functions in the chloroplast that could explain the recovery of npq2 suppressors of npq1 lor1. Zeaxanthin is well known for its role in the quenching of ¹Chl* and the thermal dissipation of excess energy (NPQ) (Müller et al., 2001). With its 11 conjugated double bonds, zeaxanthin also is an efficient quencher of ³Chl*, ¹O₂*, and free radicals, and in the case of mutant strains that lack other xanthophylls, it must fulfill a structural role in the LHC as well.

The double mutant npq1 lor1 has greatly reduced levels of NPQ, so one conceivable way in which the constitutive accumulation of zeaxanthin could increase photoprotection in the npq1 lor1 mutant background is by increasing the rate of thermal dissipation. However, it is unlikely that the bleaching phenotype of npq1 lor1 in HL is caused by the lack of NPQ in the first place (Niyogi et al., 1997b). The npq5 mutant of Chlamydomonas also shows a severe lack of NPQ but has nearly wild-type xanthophyll composition (Elrad et al., 2002), and it does not exhibit photobleaching under our HL conditions (Niyogi et al., 1997b), suggesting that maintenance of a high NPQ is not necessary for survival in HL. Paradoxically, in the npq2 suppressors, we observed restoration of NPQ to almost wild-type levels in the strain with the lowest content of zeaxanthin and the highest content of violaxanthin, npq2-4 npq1 lor1 (data not shown). Consistent with measurements in the npq2-1::ARG7 single mutant (Niyogi et al., 1997a), the extent of NPQ was not increased in npq2-2 npq1 lor1 with respect to the parental strain npq1 lor1, although this suppressor showed normal growth in HL (Figure 2). Together, these data suggest that the restoration of NPQ is not the main mechanism by which zeaxanthin (and antheraxanthin) protect the suppressor strains from photobleaching. This notion is consistent with results from Arabidopsis plants with an increased zeaxanthin content through overexpression of -carotene hydroxylase, which show enhanced tolerance to light stress under HL, although their NPQ remains unchanged (Davison et al., 2002).

Zeaxanthin accumulation by the npq2 suppressors might confer photoprotection indirectly by resulting in a decrease in the size of the light-harvesting antenna that feeds light energy into PSII. In general, a lower content of chlorophyll in the chloroplast might be expected to decrease the levels of ³Chl* and ¹O₂* formed in HL. Indeed, when grown in HL, the npq2 suppressors showed lower cellular chlorophyll content and decreased accumulation of LHCII polypeptides; thus, it is possible that part of the observed decreased sensitivity to photodamage in these mutants is caused by a decrease in the photosensitized production of ¹O₂* and free radicals. By introducing the cbn1 mutation in the npq1 lor1 background, we observed that decreased antenna size, in the absence of zeaxanthin and antheraxanthin accumulation, does allow for some degree of protection from photodamage (Figures 9 and 10). However, the cbn1 npq1 lor1 strain showed partial bleaching at 500 µmol·m^-2·s^-1 and was completely unable to grow at higher light intensities, whereas all of the npq2 npq1 lor1 strains grew at wild-type rates. We conclude that only partial protection is afforded by a decrease in antenna size in the absence of the protective xanthophylls zeaxanthin and antheraxanthin and that the smaller LHC is not sufficient to explain the recovery of npq2 suppressors in our screen.

The decreased accumulation of LHCII proteins in the npq2 npq1 lor1 strains in HL could have multiple underlying causes. It is conceivable that the abnormal xanthophyll composition might interfere with the synthesis, assembly, and/or stability of the LHCs. Experiments performed in vitro have demonstrated that it is possible to reconstitute LHC proteins with chlorophylls and with either violaxanthin or zeaxanthin in place of lutein, but the yield of LHCII-zeaxanthin complexes is substantially lower than that of either LHCII-violaxanthin or LHCII-lutein complexes (Croce et al., 1999; Hobe et al., 2000), consistent with a possible assembly defect in the npq2 suppressors in vivo. The lack of lutein caused by the lor1 mutation is known to affect LHCII organization in Chlamydomonas (Chunaev et al., 1991), and we have observed a loss of LHCII proteins in the HL-grown lor1 single mutant (I. Baroli and K.K. Niyogi, unpublished results) similar to that in the npq2-2 npq1 lor1 strain (Figure 9). This finding suggests that the absence of lutein, rather than the accumulation of abnormally high levels of zeaxanthin, is responsible for the loss of LHCII proteins in HL. However, despite the absence of lutein, the npq2-3 npq1 lor1 and npq2-4 npq1 lor1 strains in HL exhibited LHCII protein levels that were inversely related to the severity of the block in zeaxanthin epoxidation. In fact, LHCII levels approached wild-type levels in the HL-grown npq2-4 npq1 lor1 strain (Figure 9). It appears that in the absence of lutein, it is the amount of accumulated antheraxanthin that is correlated with the accumulation of LHCII proteins in HL, consistent with the close structural and conformational similarity between lutein and antheraxanthin. The loss of the LHCs in the npq2-2 npq1 lor1 strain also might be attributable in part to decreased photoprotection within the LHCs, because the LHCII-zeaxanthin complex reconstituted in vitro was found to be more sensitive to photooxidation than the wild-type complex (Formaggio et al., 2001).

However, the protective effect of zeaxanthin accumulation is enhanced outside of the LHCs, because the npq2 suppressors were more resistant than the npq1 lor1 starting strain to ¹O₂* generated by an exogenous photosensitizer, rose bengal (Figure 11). Rose bengal generates ¹O₂* throughout the growth medium and within the cell by a type-II photosensitization reaction similar to the interaction between ³Chl* and ground-state O₂. The resistance of the npq2 suppressors to rose bengal–induced photooxidation was apparent in LL, in which the effect of zeaxanthin accumulation on LHC protein levels was minor (Figures 9 and 11). This result strongly suggests that zeaxanthin could protect Chlamydomonas from photooxidation in HL by quenching ¹O₂ and/or free radicals directly in the lipid phase of the thylakoid membrane. Alternatively, zeaxanthin might have an unidentified beneficial effect on lipid peroxide metabolism. The high zeaxanthin/chlorophyll ratio and the small antenna in the suppressors, especially the npq2-2 npq1 lor1 strain grown in HL, suggests that Chlamydomonas chloroplast membranes can maintain a relatively high concentration of free zeaxanthin in the lipid phase. In fact, there is evidence for a role of zeaxanthin as a stabilizer of thylakoid membrane function and structure under stress conditions (Havaux, 1998). A role of zeaxanthin in quenching ¹O₂ is consistent with the enhanced protection of the PSII reaction center D1 protein from light-induced degradation observed in the npq2-1::ARG7 mutant (Jahns et al., 2000). Our results with Chlamydomonas mutants lacking or accumulating zeaxanthin also are in agreement with recent findings in Arabidopsis that zeaxanthin, besides promoting energy dissipation in the LHC, might protect from photooxidative stress by participating directly in quenching ¹O₂* and/or free radicals in the thylakoid membrane, thus preventing the accumulation of lipid peroxides (Havaux and Niyogi, 1999; Havaux et al., 2000).

The antioxidant effect of zeaxanthin that we have shown here for Chlamydomonas is not restricted to photosynthetic membranes (Demmig-Adams and Adams, 2002). A role of membrane carotenoids in quenching ¹O₂* also has been demonstrated in nonphotosynthetic bacteria (Tatsuzawa et al., 2000) and fungi (Schroeder and Johnson, 1995). In animals, zeaxanthin and lutein are thought to have important antioxidant functions (Mares-Perlman et al., 2002). In the primate eye, they are concentrated up to 1000-fold the serum levels in the macula lutea region of the retina, where they may protect polyunsaturated lipid membranes against light-induced damage (Landrum and Bone, 2001). Loss of zeaxanthin and lutein is correlated with the onset of age-related macular degeneration in humans, whereas increased dietary intake of xanthophylls is associated with a decreased risk of macular degeneration (Seddon et al., 1994). Although at much lower concentration than in the macula, zeaxanthin and lutein also are found in the lens epithelium and cortex, and they may help prevent the oxidation of epithelial lipids, an important etiological factor in the development of cataracts (Hammond et al., 2001). Similarities in the fatty acid composition of thylakoid and retinal membranes (an enrichment in polyunsaturated fatty acids) (Giusto et al., 2000) make Chlamydomonas an attractive system in which to study the functions of these antioxidant molecules in vivo.

In summary, we have shown that in the absence of lutein, zeaxanthin accumulation can prevent lipid peroxidation and restore growth in HL to npq1 lor1, a light-sensitive, xanthophyll-deficient strain of Chlamydomonas. The mechanism of protection by zeaxanthin most likely involves a direct quenching of ¹O₂* and/or free radicals in chloroplast membranes. A decrease in the PSII light-harvesting antenna size in zeaxanthin-accumulating suppressors of npq1 lor1 in HL also might contribute to photoprotection by decreasing the absorption of light and the generation of ¹O₂* and by liberating additional zeaxanthin into the lipid phase of the membrane, where it could quench ¹O₂*.

	METHODS

TOP Abstract INTRODUCTION RESULTS DISCUSSION METHODS References

 
Strains and Growth Conditions
The Chlamydomonas reinhardtii strains used in this study are described in Table 4. The wild-type strain CC-125 and the chlorophyll (Chl) b–less cbn1-48 mutant CC-1354 were obtained from the Chlamydomonas Genetics Center (Duke University, Durham, NC). For physiological studies, cells were grown photoautotrophically in 100 mL of minimal (HS) medium (Harris, 1989) at 25°C with shaking in air in sterile beakers. Continuous illumination was provided from the top by cool-white fluorescent lights at a PFD of 50 µmol·m^-2·s^-1 (low light [LL]) or 500 µmol·m^-2·s^-1 (high light [HL]). The light path through the cultures was 3.5 cm. For HPLC determination of pigment content and for suppressor screening, cells were exposed to continuous HL or LL on agar plates containing minimal medium. Strain stocks were maintained at 10 µmol· m^-2·s^-1 on acetate-containing agar medium (Tris-acetate-phosphate) (Harris, 1989). Cell densities were determined using a hemocytometer. Photon flux densities were measured with a quantum meter (Li-Cor, Lincoln, NE).

Isolation of Suppressors of the npq1 lor1 High-Light Sensitivity
The double mutant strains npq1 lor1 arg7-1 mt+ and npq1 lor1 arg7-8 mt- were grown in low light in HS medium to the exponential phase. A volume containing 10⁶ cells was plated on HS agar medium and placed in HL. Suppressor colonies were detected 7 days after plating from a background of bleached cells. Colonies appeared with a frequency of 10^-4.

Pigment Analysis and Quantification
For pigment analysis, cell samples were collected from agar plates at the growth light intensity and immediately frozen in liquid nitrogen. Pigments were extracted by vortexing in 300 µL of acetone at maximum speed for 2 min, and the extracts were filtered through a 2-µm nylon filter. Twenty microliters of each extract was separated by HPLC on a Waters Spherisorb S5 ODS1 4.6- x 250-mm cartridge column (Milford, MA) using a modification of the method developed by García-Plazaola and Becerril (1999). Pigments were eluted at a flow rate of 1.2 mL/min with a linear gradient from 100% solvent A (acetonitrile:methanol:0.1 M Tris-HCl, pH 8.0 [84:2:14]) to 100% solvent B (methanol:ethyl acetate [68:32]) for 15 min, followed by 3 min of solvent B. Pigments were detected at 445 nm with 550 nm as the reference wavelength. The concentration of individual carotenoids was determined using standard curves of purified pigments (purchased from VKI, Hørsholm, Denmark) at known concentrations. Carotenoid content is given as a molar fraction of Chl a, which was determined from the same HPLC run as the carotenoids, using a calibration curve constructed with Chlamydomonas cell extracts containing known amounts of spectrophotometrically determined chlorophyll (Porra et al., 1989). For total cellular chlorophyll content, the concentration of Chl a plus Chl b was determined spectrophotometrically in 90% acetone (Porra et al., 1989) using cells grown in liquid HS medium.

Measurements of Fluorescence and Oxygen Evolution
To measure fluorescence parameters and oxygen evolution, cells were grown in HS medium to the exponential phase. Chlorophyll fluorescence was measured using a pulse-amplitude modulation fluorometer (FMS2; Hansatech, King's Lynn, UK). Cells corresponding to 30 µg of Chl a were deposited onto 2.5-cm diameter, 12-µm pore size nitrocellulose filters by filtration and dark adapted in a moist Petri dish for 15 min before measurement. The cells were exposed to 8 min of illumination with weak far-red light (light-emitting diode source of 735 nm peak wavelength) before determination of the fluorescence parameter F