|Home | About | Journals | Submit | Contact Us | Français|
The psbAII locus was used as an integration platform to overexpress genes involved in carotenoid biosynthesis in Synechocystis sp. strain PCC 6803 under the control of the strong psbAII promoter. The sequences of the genes encoding the yeast isopentenyl diphosphate isomerase (ipi) and the Synechocystis β-carotene hydroxylase (crtR) and the linked Synechocystis genes coding for phytoene desaturase and phytoene synthase (crtP and crtB, respectively) were introduced into Synechocystis, replacing the psbAII coding sequence. Expression of ipi, crtR, and crtP and crtB led to a large increase in the corresponding transcript levels in the mutant strains, showing that the psbAII promoter can be used to drive transcription and to overexpress various genes in Synechocystis. Overexpression of crtP and crtB led to a 50% increase in the myxoxanthophyll and zeaxanthin contents in the mutant strain, whereas the β-carotene and echinenone contents remained unchanged. Overexpression of crtR induced a 2.5-fold increase in zeaxanthin accumulation in the corresponding overexpressing mutant compared to that in the wild-type strain. In this mutant strain, zeaxanthin becomes the major pigment (more than half the total amount of carotenoid) and the β-carotene and echinenone amounts are reduced by a factor of 2. However, overexpression of ipi did not result in a change in the carotenoid content of the mutant. To further alter the carotenoid content of Synechocystis, the crtO gene, encoding β-carotene ketolase, which converts β-carotene to echinenone, was disrupted in the wild type and in the overexpressing strains so that they no longer produced echinenone. In this way, by a combination of overexpression and deletion of particular genes, the carotenoid content of cyanobacteria can be altered significantly.
Carotenoids are pigments synthesized by photosynthetic and nonphotosynthetic organisms. In photosynthetic membranes, they are essential components of the photosynthetic apparatus and play a protective role against oxidative damage by various mechanisms, including by quenching of chlorophyll triplets, which otherwise could give rise to highly reactive singlet oxygen species (10, 13). Carotenoids may also serve a light-harvesting antenna function: absorbed light energy may be transferred to chlorophyll or may be dissipated in the case of excess radiant energy, thus protecting the photosynthetic apparatus from photooxidation (25). In photosynthetic organisms, carotenoids bind noncovalently but specifically with membrane proteins (2). Their distribution and localization in the photosynthetic membrane vary from one organism to another, but predominantly carotenes rather than xanthophylls (carotene derivatives that contain one or more oxygen atoms incorporated in various functional groups) are associated with the photosynthetic reaction center complexes (13). In plants, xanthophylls are associated mainly with the antenna complexes. Moreover, carotenoids may be found in the cytoplasmic membrane, where they are thought to influence membrane fluidity (3, 12).
Carotenoids, like sterols and gibberellins, are part of a group of compounds named isoprenoids. Despite their structural and functional diversity, isoprenoids are synthesized via a common precursor, isopentenyl diphosphate (IPP). IPP is isomerized to dimethylallyl diphosphate by IPP isomerase. Several condensation reactions convert dimethylallyl diphosphate into geranylgeranyl pyrophosphate. The first specific step in the carotenoid biosynthesis pathway is phytoene synthesis by condensation of two molecules of geranylgeranyl pyrophosphate. Four desaturation steps convert phytoene into lycopene via phytofluene, ζ-carotene, and neurosporene. Cyclization reactions occur on lycopene, giving rise to carotenes, such as β-carotene. Xanthophylls, such as zeaxanthin, are oxygenation products of carotenes (for reviews, see references 3 and 31).
In Synechocystis sp. strain PCC 6803, several genes encoding enzymes involved in carotenoid biosynthesis have been identified (Fig. (Fig.1).1). The crtB gene codes for phytoene synthase (17), crtP codes for phytoene desaturase (18), crtQ codes for ζ-carotene desaturase (6), crtO codes for β-carotene ketolase (8), and crtR codes for β-carotene hydroxylase (19). Only four carotenoids accumulate significantly in Synechocystis; these are myxoxanthophyll [2′-(β-1-rhamnopyranosyloxy)-3′,4′-didehydro-1′,2′-dihydro-β,Ψ-carotene-3,1′-diol], β-carotene (β,β-carotene), echinenone (β,β-caroten-4-one), and zeaxanthin (β,β-carotene-3,3′-diol) (Fig. (Fig.1).1).
Naturally occurring carotenoids are of commercial interest as coloring agents for food, pharmaceuticals, cosmetics, and animal feed. Because of their antioxidant properties (20), some of them have been proposed to act in the prevention of chronic diseases (27). Several of these pigments are currently produced for commercial purposes by total chemical synthesis (4, 24). The recent genetic elucidation of carotenoid biosynthetic pathways in several bacterial organisms may offer new opportunities for genetic engineering of carotenoid production in vivo (see reference 3 for a review). The first reports on synthesis of carotenoids in genetically altered noncarotenogenic bacteria have appeared (30, 32, 36). In an attempt to increase zeaxanthin accumulation in a photoautotrophic prokaryote, Synechocystis sp. strain PCC 6803, a system has now been designed to overexpress genes involved in carotenoid synthesis in this organism. This system employs the psbAII gene, which encodes the highly expressed D1 protein of photosystem II and which has a strong promoter in Synechocystis sp. strain PCC 6803 (23). The Synechocystis genome contains three genes coding for the D1 protein, psbAI, psbAII, and psbAIII, the latter two of which are expressed and by themselves individually can support normal photoautotrophic growth in the absence of the other two psbA genes (22). Therefore, the psbAII locus can be used as an integration platform to overexpress genes in Synechocystis.
To further extend the usefulness of this integration platform, we chose to develop a system that would not lead to the presence of antibiotic resistance cassettes in overexpressing strains. For this purpose, the sacB gene (33), which encodes a levan sucrase, was used as a conditionally negative marker. Expression of the sacB gene is lethal in the presence of sucrose. Introduction of sacB together with an antibiotic resistance gene first allows a positive selection for the mutant genotype by using antibiotic resistance as a screenable phenotype. After segregation of wild-type and mutant genome copies, the antibiotic resistance-sacB cassette can be removed by transformation with a markerless construct, followed by selection for sucrose resistance and screening for a strain that no longer is antibiotic resistant (34).
Synechocystis sp. strain PCC 6803 was cultivated at 30°C in modified BG-11 medium (29) buffered with 10 mM TES-NaOH (pH 8.0), at a photon flux density of 50 μmol of photons · m−2 · s−1 unless otherwise indicated. The BG-11 modification consisted of partial substitution of NaNO3 with an equal concentration of NH4NO3 (the final concentration of ammonia was 4.5 mM). For growth on plates, 1.5% (wt/vol) agar and 0.3% (wt/vol) sodium thiosulfate were added. BG-11 medium was supplemented with 50 μg of kanamycin ml−1 for kanamycin-resistant strains. For initial transformant selection, the DNA-cell mixture was plated on a BG-11 plate (50 ml), and the next day 2.5 mg of kanamycin (dissolved in sterile water) was added to the bottom of the plate. This procedure allows a gentle exposure of transformants to increasing kanamycin concentrations. To remove the kanamycin resistance-sacB cassette from the genomes of mutant strains, mutant cells were transformed with a markerless construct carrying Synechocystis sequences from immediately upstream and downstream of the kanamycin resistance-sacB cassette. This construct may contain a gene that is to be overexpressed. After transformation, the Synechocystis cells were grown in BG-11 medium for 4 days. Transformants were then plated and selected for growth in the presence of 5% (wt/vol) sucrose. Sucrose-resistant colonies were then checked for kanamycin sensitivity.
Plasmids used in this study are listed in Table Table1.1. The genome sequence of Synechocystis sp. PCC 6803 (15) was consulted through CyanoBase (http://www.kazusa.or.jp/cyano/cyano.html) to design the primers (Table (Table2)2) needed to amplify the Synechocystis sequences used to construct the integration platform and the different plasmids. The published sequence of the IPP isomerase gene (ipi) from Saccharomyces cerevisiae (1) was used to design primers and amplify the coding sequence of ipi. Sequences were amplified by PCR with Taq DNA polymerase. The integration platform, pPSBA2, contains regions upstream and downstream of the psbAII gene. It was constructed as follows: a 500-bp PstI-NdeI fragment upstream of and including the psbAII ATG start codon and a 500-bp BamHI-EcoRI fragment downstream of and including the psbAII gene stop codon were cloned by PCR and introduced into plasmid pSL1180 (Pharmacia Biotech), resulting in plasmid pPSBA2. The introduction of an NdeI site (CATATG) at the psbAII start codon allows the cloning of any coding sequence under the control of the psbAII promoter, with its start codon replacing the psbAII ATG. Plasmid pPSBA2KS was created by inserting the kanamycin resistance gene (aphX) from pUC4K (26) and the sacB gene from pRL271 (5, 33) between the HpaI and BamHI sites in the integration platform, pPSBA2 (Fig. (Fig.2).2).
A 1-kb NdeI-BamHI fragment beginning at the translational start site of the ipi gene of S. cerevisiae was amplified by PCR with yeast genomic DNA as a template (the ipi gene does not contain introns and therefore can be used directly for expression in prokaryotic cells). This fragment was cloned between the NdeI and BamHI sites of pPSBA2, leading to pIPI. A 2.5-kb NdeI-BglII fragment beginning at the translational start site of crtP and containing both the crtP and crtB coding sequences was cloned between the NdeI and BamHI sites of pPSBA2, giving rise to pCrtPB. A 0.9-kb fragment containing the coding region of crtR was amplified by PCR, introducing an NdeI site at the start codon, changing it from GTG to ATG. Plasmid pCrtR was constructed by cloning this fragment between the NdeI and BamHI sites in pPSBA2.
A 2.3-kb BamHI-EcoRI fragment containing the crtO gene was introduced between the BclI and EcoRI sites in pSL1180, leading to pCrtO. A 1.37-kb BclI-BclI fragment was removed from pCrtO so that a large part of the crtO coding sequence was deleted; the resulting plasmid was named pΔCrtO. A plasmid with the same deletion but with a kanamycin resistance-sacB construct in its place was named pΔCrtOKS.
Segregation of all mutants was confirmed by PCR. Primers used for PCR are listed in Table Table22.
Total RNA was isolated from Synechocystis sp. strain PCC 6803 as previously described (22). A sample of total RNA (10 μg per lane) was separated by electrophoresis on formaldehyde gels containing 1.2% agarose and transferred to GeneScreen Plus according to the manufacturer's instructions. Probes were prepared by hot PCR with [α-32P]dATP by using the cloned genes as templates. Hybridization was performed at 37°C for 12 h in a buffer containing 30% (vol/vol) formamide, 1% (wt/vol) sodium dodecyl sulfate (SDS), 10% (wt/vol) dextran sulfate, 1 mM Na2EDTA, 30 mM Tris-HCl (pH 7.5), and 3× SSC (1× SSC is 0.15 M NaCl plus 1.5 mM sodium citrate).
Synechocystis sp. strain PCC 6803 cells were harvested from cultures in exponential growth phase (optical density at 730 nm ≈ 0.5, measured with a Shimadzu UV-160A spectrophotometer). Pigments were extracted with 100% methanol and extracts were kept under nitrogen. Carotenoids were separated by high-performance liquid chromatography (HPLC) on a Spherisorb ODS2 4.0- by 250-mm C18 column by using a 15-min gradient of ethyl acetate (0 to 100%) in acetonitrile-water-triethylamine (9:1:0.01, vol/vol/vol) at a flow rate of 1.5 ml/min. Absorption spectra for individual peaks were obtained with a photodiode array detector. Carotenoid species were identified by their absorption spectra and by their typical retention times. The content of each carotenoid was determined by using the following equation: Ccar = Cchl × [(chl × Acar)/(car × Achl)], where Cchl is the chlorophyll concentration in the pigment extract (calculated from the absorbance of the pigment extract at 663 nm and the extinction coefficient of chlorophyll a at the same wavelength: E1% = 820) and chl and car are the specific extinction coefficients of chlorophyll α and the carotenoids, respectively, at 440 nm. The extinction coefficients of the chlorophyll and the carotenoid are the same in methanol, ethanol, and acetonitrile, as the absorbance values and the shape of the absorption spectra in these solvents are the same (data not shown). Therefore, the specific extinction coefficients used in this study were those reported previously with methanol or ethanol (16). They were assumed to remain constant regardless of the ethyl acetate concentration in the HPLC eluent. Achl and Acar are the peak areas on the chromatogram (recorded at 440 nm) of chlorophyll a and the carotenoid species, respectively.
The presence of IPP isomerase was determined by SDS-polyacrylamide gel electrophoresis (PAGE). Soluble fractions were prepared from Synechocystis strains by breaking cells in 25 mM HEPES-NaOH (pH 7.0)– 5 mM MgCl2–15 mM CaCl2–10% (vol/vol) glycerol–0.5% (vol/vol) dimethyl sulfoxide and collecting the supernatant fluid. Samples of these fractions (25 μg per lane) were loaded and polypeptides were separated on an SDS–10% PAGE gel. Proteins were visualized by staining the gel with Coomassie brilliant blue.
DNA from plasmid pPSBA2KS was used to transform Synechocystis sp. strain PCC 6803 (wild type). Transformants were selected in the presence of kanamycin and were sucrose sensitive due to the presence of the sacB gene (28). The complete segregation of the psbAII-KS strain was confirmed by PCR assays with primers upstream and downstream of the psbAII gene (Fig. (Fig.3).3). This strain was then transformed with pIPI, pCrtR, or pCrtPB DNA to remove the kanamycin resistance-sacB cassette and replace it with the coding sequences of ipi (the IPP isomerase gene of S. cerevisiae), crtR (the β-carotene hydroxylase gene in Synechocystis sp. strain PCC 6803), and crtP and crtB (genes involved in earlier carotenoid biosynthesis steps in Synechocystis). Complete segregation of the crtR2 and crtPB2 strains (i.e., strains overexpressing these genes and having a gene copy under the control of the psbAII promoter along with the native one) was confirmed by PCR assays with the same set of primers (Fig. (Fig.3).3). The correct insertion of the ipi gene in the ipiSc strain (i.e., the strain overexpressing the ipi gene of S. cerevisiae under the control of the psbAII promoter) was confirmed by sequencing of a PCR product from the Synechocystis ipiSc strain that covered ipi and the flanking regions (data not shown).
The psbAII-KS strain of Synechocystis already contains a wild-type copy of crtR. Homologous single recombination at this site with pCrtR DNA would result in insertion of plasmid DNA and duplication of crtR. Even though single-recombination events are rare in Synechocystis sp. strain PCC 6803 (35), we chose to probe the crtR locus by PCR. Primers were designed upstream and downstream of the crtR gene to amplify only the wild-type copy of crtR. Indeed, as expected, the normal wild-type crtR copy was present in the genome of the crtR2 strain and no foreign DNA fragment had been inserted at this locus (Fig. (Fig.3).3). Furthermore, PCR assays were performed with a primer designed upstream of the crtR gene combined with one designed downstream of the psbAII gene or with primers designed upstream of the psbAII gene and downstream of the crtR gene. In these cases, no PCR product was amplified (data not shown). This result excludes the possibility of single crossings over at the crtR locus. The presence of wild-type copies of crtP and crtB in the crtPB2 strain was checked in a similar fashion (Fig. (Fig.33).
Disruption of the β-carotene ketolase gene (crtO) has been reported to have no effect on the physiological functions of the resulting echinenoneless strain (8). The specific role of echinenone in Synechocystis remains unknown. Therefore, to simplify the carotenoid content of strains created in this study, the crtO gene was disrupted in the wild type and the crtR2 and crtPB2 strains. DNA from pΔCrtOKS was used to transform these strains. Complete segregation was confirmed by PCR assays with primers upstream and downstream of the crtO gene. The ΔcrtO-KS, crtPB2ΔcrtO-KS, and crtR2ΔcrtO-KS strains, were then transformed with pΔCrtO DNA to remove the kanamycin resistance-sacB construct and to select strains without an antibiotic resistance marker (Fig. (Fig.3A3A and C).
The steady-state levels of ipi, crtPB, and crtR transcripts in overexpressing and control strains were determined by Northern blot analysis with a 400-bp gene internal DNA probe for each of these genes. In the wild-type strain, no transcript was detected with the crtR probe, indicating a very low level of transcript accumulation. However, in the crtR2 strain, a 1-kb crtR transcript accumulated to significant levels (Fig. (Fig.4A).4A). These results show that the psbAII promoter indeed is suitable for overexpression and that the integration of a gene into the psbAII locus leads to overexpression of the inserted gene.
Accumulation of the crtPB transcript in the wild type and the crtPB2 strain was determined by using two different probes that were specific for crtB and crtP. Indeed, with both probes a three- or fourfold-stronger signal was detected in the crtPB2 strain than in the wild-type strain (Fig. (Fig.4B4B and C). With the crtB probe, transcripts of 1.2, 2.2, and 2.6 kb were detected in the crtPB2 strain (Fig. (Fig.4B).4B). The smallest transcript was accumulated at a low level (which was barely reproduced on the photograph) and had a size close to that expected for the crtB transcript. The 2.6-kb band may correspond to a transcript that includes both crtP and crtB. The 2.2-kb transcript may be a processing product of the 2.6-kb transcript. With the crtP probe, 2.2- and 2.6-kb transcripts were detected along with a 1.4-kb transcript, the last being a very weak band presumably corresponding to a crtP transcript (Fig. (Fig.4C).4C). The two gene-specific transcripts found in the overexpressing strain presumably are processing products of the 2.6-kb transcript.
In Synechocystis sp. strain PCC 6803, the gene coding for IPP isomerase has not been identified: in CyanoBase (see Materials and Methods) no sequence with convincing similarity to yeast ipi was found. As expected, no transcript was detected in the wild-type Synechocystis strain with the ipi probe. In the ipiSc strain a 1-kb ipi transcript was easily detected and was accumulated to a significant level (Fig. (Fig.5A).5A). Therefore, as expected, the psbAII promoter can drive overexpression of various genes, including genes from other organisms.
The presence of IPP isomerase in the ipiSc strain was determined by SDS-PAGE. Proteins were visualized by staining the gel with Coomassie brilliant blue (Fig. (Fig.5B).5B). Two proteins with molecular masses close to 40 and 20 kDa were detected only in the ipiSc strain and not in the wild type. The ipiSc-specific band at around 40 kDa has an apparent molecular mass consistent with that determined by SDS-PAGE for IPP isomerase purified from S. cerevisiae (1). The smaller, 20-kDa band (barely visible on the photograph) may represent an IPP isomerase degradation product.
To determine whether the carotenoid content and composition had been affected by introduction of the various mutations, methanol extractions were performed on intact cells and the carotenoid content of the extracts was analyzed by HPLC. The results are shown in Table Table33 (see also Fig. Fig.6).6). Overexpression of crtR in the crtR2 strain led to a 2.5-fold increase in zeaxanthin accumulation and a 2-fold reduction in β-carotene and echinenone content. The myxoxanthophyll content was barely affected by overexpression of crtR. Overexpression of crtP and crtB induced a 60% increase in myxoxanthophyll content and also a significant increase in zeaxanthin content in the crtPB2 strain compared to those in the wild-type strain. Echinenone and β-carotene accumulations were unaffected. In contrast to the results of crtR and crtPB overexpression, introduction of yeast ipi did not lead to any significant change in the carotenoid content.
The carotenoid content in the ΔcrtO, crtR2ΔcrtO, and crtPB2ΔcrtO strains was analyzed as well (Table (Table3).3). Disruption of the crtO gene and an absence of echinenone in a crtO-deficient strain were reported to have no effect on the growth rate, the photosynthetic oxygen evolution, or the carotenoid content (besides the absence of echinenone) (8). As expected, in the three ΔcrtO strains echinenone was absent. In the ΔcrtO strain in a wild-type background, β-carotene and zeaxanthin contents were insignificantly changed compared to those in the wild-type strain, in agreement with previous observations (8). However, in our study, in the ΔcrtO strain myxoxanthophyll accumulated to levels more than twice that of the control. Disruption of crtO in the crtPB2 strain resulted in a slight increase in β-carotene and zeaxanthin content and a more substantial increase (30%) in myxoxanthophyll content. On the other hand, the nonechinenone carotenoid content in the crtR2 ΔcrtO strain was little modified from that in the crtR2 strain.
Growth and carotenoid composition of the different mutant strains were studied after growth at three photon flux densities: 50, 100, and 200 μmol of photons · m−2 · s−1. At these three photon flux densities, the doubling times of the crtPB2 and crtR2 strains were similar to that of the wild-type strain (between 12 and 13 h [data not shown]). Similarly, increasing the photon flux density did not significantly affect carotenoid content in the wild-type strain or in the crtPB2 or crtR2 strain (data not shown).
The psbAII gene is known to have a strong promoter in Synechocystis sp. strain PCC 6803 (23). In this study, we show that this promoter can be used to drive transcription and to overexpress various genes in Synechocystis: introduction of coding sequences of genes involved in carotenoid biosynthesis in lieu of the psbAII coding sequence led to a stable overaccumulation of the corresponding transcripts. The degree of overaccumulation appeared to depend on the gene that was introduced. Overexpression of Synechocystis genes involved in carotenoid biosynthesis appeared to lead in most cases to an increased synthesis of the corresponding enzymes, because the carotenoid content of the crtPB- and crtR-overexpressing strains was modified. This overexpression system thereby provides a way to better understand carotenoid biosynthesis regulation.
Overexpression of the IPP isomerase gene from S. cerevisiae in the ipiSc strain led to the accumulation of the corresponding transcript and of a protein with a molecular mass of about 40 kDa. This molecular mass is consistent with the molecular mass determined by SDS-PAGE for IPP isomerase purified from S. cerevisiae (1). Expression of the yeast IPP isomerase gene was reported to enhance carotenoid biosynthesis in Escherichia coli (14). In this system, increases in carotenoid content were consistent with quantitative increases in IPP isomerase activity. In our study, ipi overexpression had no effect on the carotenoid content of the ipiSc strain. IPP isomerase activity will have to be examined to determine if the exogenous IPP isomerase is functionally active in the mutant strain. The absence of an effect of yeast IPP isomerase on carotenoid composition may indicate that either this enzyme is too different from the Synechocystis IPP isomerase to be functional in this organism (for example, it may need other polypeptides for function in vivo) or isomerization of IPP is not a rate-limiting step in carotenoid biosynthesis in Synechocystis sp. strain PCC 6803.
Overexpression of genes encoding phytoene synthase and phytoene desaturase (crtB and crtP, respectively) in the crtPB2 strain was indicated by an overaccumulation of a 2.6-kb transcript. This transcript is likely to include both the crtP and crtB coding sequences. A transcript of this size was also detected in the wild-type strain, which suggests that the crtP and crtB genes are part of the same operon. The presence of two small (1.2- and 1.4-kb) transcripts specific to crtB and crtP in the crtPB2 strain may indicate either processing of the longer transcript or the possibility of transcription of the two genes independently. Earlier data showing that disruption of crtP did not greatly affect the expression of a reporter gene placed in lieu of crtB were interpreted to indicate that crtB may have its own promoter (17). This result was confirmed recently (9) by detection of crtB and crtP transcripts with sizes smaller than 2.5 kb. However, judging from the sizes on the Northern blots, the majority of the crtB-containing transcripts in our study appear to be dicistronic crtPB transcripts.
Phytoene desaturation has been reported to be a rate-limiting step in carotenoid biosynthesis in Synechococcus sp. strain PCC 7942 (7). However, overexpression of the phytoene desaturase gene (crtP) by itself did not induce an increase in the carotenoid content of the corresponding overexpressing strain (data not shown). On the other hand, overexpression of both crtP and crtB (the latter being the phytoene synthase gene) resulted in a 50% increase in the myxoxanthophyll and zeaxanthin content. As these two carotenoids are terminal biosynthesis products in Synechocystis, this result is consistent with an increased carotenoid biosynthesis capacity. Therefore, we suggest that phytoene synthase is slightly rate limiting in wild-type Synechocystis sp. strain PCC 6803 under the conditions we used.
The myxoxanthophyll biosynthesis pathway remains unknown. However, as indicated in Fig. Fig.1,1, this carotenoid is thought to be synthesized from γ-carotene (2). Indeed, overexpression of crtP and crtB may lead to increased levels of γ-carotene, part of which would be converted to myxoxanthophyll. In the crtPB2 strain, the increased γ-carotene synthesis would be expected to also lead to an increase in β-carotene levels if more of this pigment could be accommodated. The steady-state β-carotene amount remained unchanged, but the level of zeaxanthin, which is formed by hydroxylation of β-carotene, was increased in the crtPB2 strain. Interestingly, the only mutant in which β-carotene levels were significantly increased in comparison to that in the wild type is the crtPB overexpresser strain that lacks crtO: in this strain, it is possible that β-carotene occupied some of the carotenoid binding sites that were left vacant due to the lack of echinenone. These results suggest that the β-carotene level cannot be increased very much but that zeaxanthin or myxoxanthophyll levels are less strictly regulated. Similar observations have been reported with transgenic Synechococcus sp. strain PCC 7942, in which expression of an algal gene encoding β-C-4-oxygenase led to the production of various ketocarotenoids (which normally do not accumulate in this cyanobacterium), with a decrease in zeaxanthin content but no change in β-carotene accumulation (11).
As levels of zeaxanthin and myxoxanthophyll could be increased almost threefold compared to wild-type levels, there may be more unoccupied binding sites available for zeaxanthin and myxoxanthophyll than for β-carotene. In addition, zeaxanthin and myxoxanthophyll may also be present as free molecules in the membrane or cell wall. In the membrane, dipolar carotenoids, like zeaxanthin or myxoxanthophyll, that are not bound to protein are anchored in the bilayer, with their long axis almost perpendicular to the membrane plane, thus essentially spanning the membrane. However, β-carotene is distributed homogeneously within the lipophilic part of the membrane without well-defined orientation. As a result, dipolar carotenoids decrease membrane fluidity whereas β-carotene tends to increase it (12). These different effects of carotenoids on membrane fluidity may explain why zeaxanthin and myxoxanthophyll accumulate in the crtPB2 strain whereas β-carotene does not.
Overexpression of the gene coding for β-carotene hydroxylase in the crtR2 strain led to a large increase in zeaxanthin accumulation in this strain. In logarithmically growing cultures of the wild-type strain, the amount of zeaxanthin represents a quarter of the total carotenoid content and β-carotene is the major carotenoid (a third of the total). In the crtR-overexpressing strain, zeaxanthin is the major carotenoid (more than half of the total amount) and β-carotene becomes a minor component of the cells. This result suggests that β-carotene hydroxylation is a rate-limiting step in zeaxanthin biosynthesis in Synechocystis sp. strain PCC 6803: the conversion of β-carotene into zeaxanthin is regulated by the amount of β-carotene hydroxylase available. This result is consistent with what was previously reported for a noncarotenogenic microorganism: the amount of production of zeaxanthin in E. coli was related to the expression levels of β-carotene hydroxylase (30).
Disruption of the β-carotene ketolase gene (crtO) has been reported not to affect the Synechocystis carotenoid content, except that echinenone, which is produced by β-carotene ketolase activity, is absent in a strain with such a disruption (8). In the same work, the myxoxanthophyll content in the wild type and in the crtO-deficient strain appeared to be very small compared to the content of other carotenoids. In our study a higher myxoxanthophyll content was observed in the ΔcrtO and crtPB2 ΔcrtO strains than in the wild type and the crtPB2 strain. The content of other carotenoids (β-carotene and zeaxanthin) remained virtually unchanged compared to that in the wild-type background. In our study, disruption of the crtO gene in the wild type and in the crtPB2 mutant led to an increase in overall carotenoid levels, mainly due to an increase in myxoxanthophyll amounts. The difference observed in myxoxanthophyll amounts between the previous study (8) and our work may be due to the use of different methods to extract and analyze carotenoids.
How this apparent regulation of echinenone versus myxoxanthophyll accumulation works is as yet unknown, but it may be related to the relative accumulation of one of the intermediates in the biosynthesis pathway. In the crtR2 strain, the echinenone content was reduced because overexpression of β-carotene hydroxylase induced a higher conversion of β-carotene into zeaxanthin and left less substrate available for conversion to echinenone. Consistent with the reasoning above, disruption of crtO in the crtR2 strain did not lead to an increase in myxoxanthophyll content.
Carotenoid biosynthesis gene clusters have been isolated and the functions of individual genes have been identified for several carotenogenic bacteria (3). Metabolic engineering, the modification of metabolic networks in living cells to produce desirable chemicals with superior yields and productivity by recombinant DNA techniques, allows the production of large amounts of useful carotenoids in vivo (see reference 21 for a review). In this study, we show that such a metabolic engineering approach can be used with Synechocystis sp. strain PCC 6803 to overproduce desirable carotenoids, such as zeaxanthin. Furthermore, the system developed in this study allows for gene replacement without introduction of antibiotic resistance cassettes in the final overexpressing strains. The absence of cassettes in such strains is a positive feature highlighting the increasing desire of the biotechnology industry to avoid spreading antibiotic resistance cassettes.
We are grateful to G. Ajlani for kindly providing plasmid pRL271. We also thank the members of the Vermaas laboratory, particularly Jason W. Cooley and Crispin A. Howitt, for useful discussions.
Funding for materials and supplies in the Vermaas laboratory was provided by a grant from the National Science Foundation (MCB-9728400).