Random chemical mutagenesis of a specific psbDI region coding for a lumenal loop of the D2 protein of photosystem II in Synechocystis sp. PCC 6803

To identify amino acid residues of the D2 protein that are critical for functional photosystem II (PS II), sodium bisulfite was utilized for in vitro random mutagenesis of the psbDI gene from Synechocystis sp. PCC 6803. Sodium bisulfite reacts specifically with cytosine in single-stranded regions of DNA and does not attack double-stranded DNA. Using a hybrid plasmid that was single-stranded in the region to be mutagenized and that was double-stranded elsewhere, mutations were targeted to a specific psbDI region coding for the lumenal A-B loop of the D2 protein. Several mutants were isolated with a total of 15 different amino acid changes in the loop. The majority of these mutations did not result in a loss of photoautotrophic growth or in significantly altered PS II function. However, mutation of Glu-69 to Lys, Ser-79 to Phe, and Ser-88 to Phe were found to influence photosystem II activity; the importance of the latter two residues for proper PS II function was unexpected. Cells carrying the double mutation S79F/S88F in D2 did not grow photoautotrophically and had no functionally active PS II centers. The single mutant S79F was also incapable of photoautrophic growth, but displayed reasonably stable oxygen evolution, while PS II function in the single mutant S88F appeared to be close to normal. Because of the more pronounced phenotype of the S79F/S88F strain as compared to the single mutants, both Ser residues appear to affect stable assembly and function of the PS II complex. The mechanism by which the S79F mutant loses photoautotrophic growth remains to be established. However, these results show the potential of targeted random mutagenesis to identify functionally important residues in selected regions of proteins.     

Molecular analysis of psb A mutations responsible for various herbicide resistance phenotypes in Synechocystis 6714

Mutations conferring herbicide resistance in 3 mutant strains of the cyanobacterium Synechocystis 6714 have been characterized by gene cloning and sequencing. The mutants display very different phenotypes: DCMU-II_A is DCMU-resistant and atrazine-resistant, DCMU-II_B is DCMU-resistant and atrazine-sensitive, and Az-V is DCMU-sensitive, atrazine-resistant and presents particular photoinhibition properties. These mutants were originally obtained either by one-step selection (DCMU-II_A) or by two-step selection (DCMU-II_B and Az-V). psbA copies carrying herbicide resistance have been identified by transformation experiments as psbAI in all cases. Sequences of the psbAI copy of each mutant have been compared to the wild-type sequence. In the single mutant DCMU-II_A, a point mutation at codon 264 (SerAla) results in resistance to both DCMU and atrazine. In the double mutants DCMU-II_B and Az-V, two point mutations were found. DCMU-II_B was derived from DCMU-II_A and had acquired a second mutation at codon 255 (PheLeu) resulting in a slight increase in DCMU resistance and complete abolition of atrazine resistance. Az-V contains two changes at codons 211 (PheSer) and 251 (AlaVal) resulting in high atrazine resistance but only slight DCMU resistance.Abbreviations DCMU: 3-(3,4-dichlorophenyl)-1,1-dimethylurea - PSII: photosystem II     

Targeted site-directed mutagenesis of a heme oxygenase locus by gene replacement in the moss<Emphasis Type="Italic"> Ceratodon purpureus</Emphasis>

The moss Physcomitrella patens is so far the only plant species in which it is possible for nuclear genes to be modified by homologous recombination at a reasonably efficiency. Here we describe the use of homologous recombination for another moss, Ceratodon purpureus. Our approach is based on the repair of the ptr116 mutant allele. In this mutant, codon 31 of the heme oxygenase gene CpHO1 is mutated to a stop codon. Heme oxygenase is necessary for the conversion of heme to biliverdin, the precursor of the phytochrome chromophore. Thus, in ptr116 the phytochrome-mediated responses of phototropism, chlorophyll accumulation and branching are lost. Protoplast transformation with DNA encoding the wild-type protein resulted in a rescue of 0.8% of regenerated protoplasts. In about half of the analyzed lines, formation of CpHO1 concatemers was observed at the CpHO1 locus, whereas in the other half, the mutant CpHO1 gene was replaced by a single DNA copy. This gene repair led to the exchange of single bases, and thus provides the first demonstration of efficient site-directed mutagenesis in a plant nuclear genome. Our studies also revealed an effective mechanism for gene inactivation in Ceratodon. When wild-type protoplasts were transformed with intact or modified CpHO1 genes, approximately 40% of regenerated protoplasts showed the ptr phenotype.     

Cloning and Inactivation of a Gene Essential to Inorganic Carbon Transport of Synechocystis PCC6803

A clone (HP-1) which transforms the high CO₂-requiring mutant (RKb) of Synechocystis PCC6803 defective in inorganic carbon transport to the wild-type (WT) phenotype was isolated from a WT genomic library. The clone contained a 5.4 kilobase-pair DNA insert. Complementation tests with subclones derived from HP-1 allowed the mutation in RKb to be located within 141 base-pair nucleotides. Sequencing of nucleotides in this region revealed an open reading frame encoding a hydrophobic protein consists of 80 amino acids. A defined mutant (M9) constructed by inactivating this putative inorganic carbon transport gene, designated ictA, was unable to transport CO₂ and HCO₃⁻ into the intracellular inorganic carbon pool. Cloning and sequence analysis of the respective RKb gene revealed a base substitution which generates a stop codon in the middle of ictA.     

Wild-type and mutant alleles of the Aspergillus nidulans developmental regulator gene brlA: correlation of variant sites with protein function

The DNA sequences of two wild-type and eleven mutant alleles of the developmental regulator gene brlA from Aspergillus nidulans, which encodes a zinc-finger protein, were characterized. Variant sites were located on rescued plasmids or PCR products based either on their meiotic map position or the use of denaturing gradient gel electrophoresis. Mutations in three null mutants, one of which is partially suppressible, encode premature stop codons. Two environmentally sensitive mutants were characterised by substitution of leucines required for stabilisation of α-helices in each of the two putative zinc-finger domains. A third zinc-finger substitution is predicted to disrupt recognition of a guanine residue in the DNA target. The mutations in four other leaky mutants map C-terminal to the zinc fingers; one minimally leaky mutant has a premature stop codon, which results in the removal of the last 38 residues of the protein product. Received: 16 February 1999 / Accepted: 22 July 1999     

Detection of point mutations in chloroplast genes of Antirrhinum majus L. I. Identification of a point mutation in the psaB gene of a photosystem I plastome mutant

A point mutation in the plastome-encoded psaB gene of the mutant en:alba-1 of Antirrhinum majus L. was identified by an analysis of chloroplast DNA with a modified PCR-SSCP technique. Application of this technique is indicated when a gene or a group of genes is known in which the point mutation is located. Analysis of primary photosynthetic reactions in the yellowish white plastome mutant indicated a dysfunction of photosystem (PS) 1. The peak wavelength of PS I-dependent chlorophyll (Chl) fluorescence emission at 77 K was shifted by 4 nm to 730 nm, as compared to fluorescence from wild-type. There were no redox transients of the reaction center Chl P₇₀₀ upon illumination of leaves with continuous far-red light or with rate-saturating flashes of white light. The PS I reaction center proteins PsaA and PsaB are not detectable by SDS-PAGE in mutant plastids. Hence, plastome encoded PS I genes were regarded as putative sites of mutation. In order to identify plastome mutations we developed a modified SSCP (single-strand conformation polymorphism) procedure using a large PCR fragment which can be cleaved with various restriction enzymes. When DNA from wild-type and en:alba-1 was submitted to SSCP analysis, a single stranded Hinf I fragment of a PCR product of the psaB gene showed differences in electrophoretic mobility. Sequence analysis revealed that the observed SSCP was caused by a single base substitution at codon 136 (TAT TAG) of the psaB gene. The point mutation produces a new stop codon that leads to a truncated PsaB protein. The results presented indicate that the mutation prevents the assembly of a functional PS I complex. The applicability to other plastome mutants of the new method for detection of point mutations is discussed.     

Cloning, nucleotide sequence and mutational analysis of the gene encoding the Photosystem II manganese-stabilizing polypeptide of Synechocystis 6803

Summary Affinity purified, polyclonal antibodies raised against the Photosystem II 33 kDa manganese-stabilizing polypeptide of the spinach oxygen-evolving complex were used to isolate the gene encoding the homologous protein from Synechocystis 6803. Comparison of the amino acid sequence deduced from the Synechocystis psb1 nucleotide sequence with recently published sequences of spinach and pea confirms the homology indicated by antigenic crossreactivity and shows that the cyanobacterial and higher plant sequences are 43% identical and 63% conserved. Regions of identity, varying in length from 1 to 10 consecutive residues, are distributed throughout the protein. The 28 residues at the amino terminus of the psb1 gene product, characteristic of prokaryotic signal peptides, show homology with the carboxyl-terminal third of the transit sequences of pea and spinach and are most likely needed for the transport of the manganese-stabilizing protein across the thylakoid membrane to its destination of the lumen. Synechocystis mutants which contain a kanamycin resistance gene cassette inserted into the coding region for the 32 kDa polypeptide were constructed. These mutants contain no detectable 32 kDa polypeptide, do not evolve oxygen, and are incapable of photoautotrophic growth.     

Studies on the bacteriophage MS2: XLI. Nature of the azure mutation

Azure (or reverse amber) mutants grow normally on wild-type Escherichia coli but not on host strains harbouring a strong UAG suppressor mutation. Three different bacteriophage MS2 azure mutants obtained by treatment with nitrous acid have been characterized at the nucleotide sequence level. The 3′-end fragment of the ³²P-labelled mutant genomes was isolated by DNA:RNA hybridization and treatment with nuclease S₁, and was analyzed by mini-fingerprinting of the RNA. It is known that the wild-type MS2 polymerase gene ends with a UAG codon, followed seven triplets further by an in-phase UAA triplet. All three azure mutants contained an A → G transition in this UAA second stop codon of the polymerase gene, resulting in a second suppressible UAG (amber) codon. Analysis of revertants demonstrated that the azure mutation can be counteracted either by a true site reversion at the second stop or by the creation of a new stop signal for the polymerase gene, either UAA (ochre) or UGA (opal), before or at the first stop, or beyond the second stop. On the basis of these results, a mechanism for the azure mutation is proposed. Silent mutations (one in the coding region, three in the untranslated 3′-terminal sequence) have also been observed in these phage stocks.     

Deletion and insertion mutations in early region 1a of type 5 adenovirus that produce cold-sensitive or defective phenotypes for transformation

On the basis of earlier findings showing that H5hr1 (hr1) is cold sensitive for transformation, a series of mutants were constructed so that they contained deletions or insertions in different sites of early region 1a (E1a) to ascertain: (i) whether the cold-sensitive phenotype of hr1 was the result of the identified single-base pair deletion of nucleotide 1,055 or due to a missense mutation at another site and (ii) what region and how much of the E1a 51-kilodalton protein is actually required to produce cell transformation. A mutant, H5dl101 (dl101), was constructed to contain a 5-base pair deletion of nucleotides 1,008 to 1,012, which produced a frameshift and a subsequent stop codon at nucleotide 1,241. This mutant, which should encode a truncated 33-kilodalton protein in place of the wild-type 51-kilodalton protein, had a cold-sensitive phenotype for transformation essentially identical to hr1. Consonant with this finding, a mutant (H5in106) engineered to contain a 16-base pair insertion initiated after nucleotide 1,009, with a stop codon beginning at the newly inserted nucleotide 1,013, also had a cold-sensitive phenotype like hr1 and dl101. It is striking, however, that a mutant (H5dl105) with a 69-base pair deletion beginning at nucleotide 1,003, and having a stop codon at nucleotide 1,544, was totally defective for transformation at any temperature. Transfection studies with plasmids containing the E1a or E1a and E1b genes of sub309, hr1, and dl101 further revealed that the cold-sensitive transformation phenotype observed could be exhibited in the absence of viral E1b gene expression.     

Genetic structure and domains of DNA polymerase III of Bacillus subtilis

Summary We have determined the nucleotide sequence of the polC gene of Bacillus subtilis which codes for DNA polymerase III. Our recent analysis has revealed that the gene comprises 4311 nucleotides, from the start to the stop codon, 306 nucleotides more than we reported earlier. The plasmid reported by us and by N.C. Brown's laboratory contained a sequence at the end of the gene which is not related to the polC region of B. subtilis. We have isolated the rest of the gene, the sequence of which is presented in this paper. The new stop codon is followed by a hyphenated palindromic sequence of 13 nucleotides. The C-terminus' of the coding region contains the novel mutation, dnaF, which results in a defect in the initiation of replication due to a change in the codon TCC to TTC (serine to phenylalanine). The hypermutator mutation mut-1 is due to two point mutations in the 3 to 5 exonuclease domain, the proof reading function. The codon changes are GGA to GAA (glycine to glutamic acid) and AGC to AAC (serine to asparagine). The elongation defective mutation, polC26, affecting the catalytic site that adds nucleotides to the growing chain, is due to a change in the codon GTC to GAC (valine to aspartic acid). It is separated from the mutation reported earlier, azp-12, by 306 nucleotides. Knowing the locations of the mutational sites allowed us to deduce the domains of the gene and the enzyme it encodes, and permitted us to present a precise map of the gene at the molecular level.Abbreviations HPUra p-hydroxyphenyl azouracil - nt nucleotide - PCR polymerase chain reaction     

Sequence conservation among the glucose transporter from the cyanobacterium Synechocystis sp. PCC 6803 and mammalian glucose transporters

Synechocystis sp. PCC 6803 is capable of facultative photoheterotrophy with glucose as the sole carbon source. Eight mutants that were unable to take up glucose were transformed with plasmids from pooled gene banks of wild-type Synechocystis DNA prepared in an Escherichia coli vector that does not replicate in Synechocystis. One mutant (EG216) could be complemented with all gene banks to restore ability for photoheterotrophic growth. One of the gene banks was fractionated into single clones and plasmid DNA from each clone used to complement EG216. This yielded a 1.5 kb DNA fragment that was sequenced. It contained one complete open reading frame (gtr) whose putative gene product displayed high sequence conservation with the xylose transporter of E. coli and the mammalian glucose transporters. Further, the isolated gtr gene interrupted in vitro by a kanamycin resistance cassette could be used to construct mutants from wild-type Synechocystis sp. PCC 6803 that lacked a functional glucose transporter, thus confirming the identity of the gtr gene with the glucose transporter gene. This is the first prokaryotic glucose transporter known to share a sequence relationship with mammalian glucose transporters and the first sugar transporter from a cyanobacterium characterized at the sequence level.     

Alcohol dehydrogenase AdhA plays a role in ethanol tolerance in model cyanobacterium Synechocystis sp. PCC 6803

The protein AdhA from the cyanobacterium Synechocystis sp. PCC 6803 (hereafter Synechocystis) has been previously reported to show alcohol dehydrogenase activity towards ethanol and both NAD and NADP. This protein is currently being used in genetically modified strains of Synechocystis capable of synthesizing ethanol showing the highest ethanol productivities. In the present work, mutant strains of Synechocystis lacking AdhA have been constructed and tested for tolerance to ethanol. The lack of AdhA in the wild-type strain reduces survival to externally added ethanol at lethal concentration of 4% (v/v). On the other hand, the lack of AdhA in an ethanologenic strain diminishes tolerance of cells to internally produced ethanol. It is also shown that light-activated heterotrophic growth (LAHG) of the wild-type strain is impaired in the mutant strain lacking AdhA (∆adhA strain). Photoautotrophic, mixotrophic, and photoheterotrophic growth are not affected in the mutant strain. Based on phenotypic characterization of ∆adhA mutants, the possible physiological function of AdhA in Synechocystis is discussed.

     

Detection of point mutations in chloroplast genes of Antirrhinum majus L. I. Identification of a point mutation in the psaB gene of a photosystem I plastome mutant

A point mutation in the plastome-encoded psaB gene of the mutant en:alba-1 of Antirrhinum majus L. was identified by an analysis of chloroplast DNA with a modified PCR-SSCP technique. Application of this technique is indicated when a gene or a group of genes is known in which the point mutation is located. Analysis of primary photosynthetic reactions in the yellowish white plastome mutant indicated a dysfunction of photosystem (PS) 1. The peak wavelength of PS I-dependent chlorophyll (Chl) fluorescence emission at 77 K was shifted by 4 nm to 730 nm, as compared to fluorescence from wild-type. There were no redox transients of the reaction center Chl P₇₀₀ upon illumination of leaves with continuous far-red light or with rate-saturating flashes of white light. The PS I reaction center proteins PsaA and PsaB are not detectable by SDS-PAGE in mutant plastids. Hence, plastome encoded PS I genes were regarded as putative sites of mutation. In order to identify plastome mutations we developed a modified SSCP (single-strand conformation polymorphism) procedure using a large PCR fragment which can be cleaved with various restriction enzymes. When DNA from wild-type and en:alba-1 was submitted to SSCP analysis, a single stranded Hinf I fragment of a PCR product of the psaB gene showed differences in electrophoretic mobility. Sequence analysis revealed that the observed SSCP was caused by a single base substitution at codon 136 (TAT → TAG) of the psaB gene. The point mutation produces a new stop codon that leads to a truncated PsaB protein. The results presented indicate that the mutation prevents the assembly of a functional PS I complex. The applicability to other plastome mutants of the new method for detection of point mutations is discussed.     

Arginine and proline genes ofAspergillus niger

Summary Aspergillus niger mutants defective in arginine or proline biosynthesis have been isolated and 12 genetic loci were identified. Mutation was induced by low doses UV, and mutants were isolated after filtration enrichment. The mutants were classified according to their phenotype in growth tests and were further characterized in complementation tests. The arginine auxotrophic mutants represent nine complementation groups. Three additional complementation groups were found for mutants that could grow on proline (two of them on arginine too). Linkage group analysis was done in somatic diploids obtained from a mutant and a master strain with genetic markers on six chromosomes. Thearg genes belong to six different linkage groups and thepro genes to two. Onearg-mutant could be complemented by transformation with theA. nidulans arg B ⁺ gene, and thisA. niger gene thus appeared to be homologous to theA. nidulans arg B. We isolated anA. niger strain with theargB gene tightly linked with thenicA1 marker. This strain is very suitable as acceptor for transformation with anargB-plasmid, because transformants with inserts on the homologous site can be recognized and analyzed genetically using thenicA1 marker gene.     

Nonsense suppression in Schizosaccharomyces pombe: the S. pombe Sup3-e tRNASerUGA gene is active in S. cerevisiae

Summary The gene encoding the efficient UGA suppressor sup3-e of Schizosaccharomyces pombe was isolated by in vivo transformation of Saccharomyces cerevisiae UGA mutants with S. pombe sup3-e DNA. DNA from a clone bank of EcoRI fragments from a S. pombe sup3-e strain in the hybrid yeast vector YRp17 was used to transform the S. cerevisiae multiple auxotroph his4-260 leu2-2 trp1-1 to prototrophy. Transformants were isolated at a low frequency; they lost the ability to grow in minimal medium after passaging in non-selective media. This suggested the presence of the suppressor gene on the non-integrative plasmid. Plasmid DNA, isolated from the transformed S. cerevisiae cells and subsequently amplified in E. coli, transformed S. cerevisiae his4-260 leu2-2 trp1-1 to prototrophy. In this way a 2.4 kb S. pombe DNA fragment carrying the sup3-e gene was isolated. Sequence analysis revealed the presence of two tRNA coding regions separated by a spacer of only seven nucleotides. The sup3-e tRNA _Ser ^UGA tRNA gene is followed by a sequence coding for the initiator tRNA^Met. The transformation results demonstrate that the cloned S. pombe UGA suppressor is active in S. cerevisiae UGA mutant strains.     

The catalase gene differentiates between some strains of <Emphasis Type="Italic">Staphylococcus aureus</Emphasis> ssp. <Emphasis Type="Italic">anaerobius</Emphasis>

Staphylococcus aureus ssp anaerobius strain S10 was isolated from an outbreak of sheep abscess disease. Sequence of the catalase gene of this strain showed 99 % identity to the catalase gene (katB) sequence of the reference strain (S. aureus ssp. anaerobius strain MVF213) with mismatching of three base pairs. An important substitution located 1036 nucleotides upstream of the initiation codon from “C” in katB to “T” in the catalase gene of strain S10 originated a stop codon. The deduced protein (345 amino acids) is 105 amino acids shorter than that of katB. Partial sequence of the catalase gene of other 8 local isolates in addition to another reference strain (DSM 20714/ATCC 35844) revealed the same mutations in all local (African) strains, whereas the sequence of the reference (European) strain was typical to that of katB. Sequence of the catalase gene of S. aureus ssp. anaerobius strain S10 was deposited in GenBank under accession no. EU281993.     

Generation of a mouse mutant by oligonucleotide-mediated gene modification in ES cells

Oligonucleotide-mediated gene targeting is emerging as a powerful tool for the introduction of subtle gene modifications in mouse embryonic stem (ES) cells and the generation of mutant mice. However, its efficacy is strongly suppressed by DNA mismatch repair (MMR). Here we report a simple and rapid procedure for the generation of mouse mutants using transient down regulation of the central MMR protein MSH2 by RNA interference. We demonstrate that under this condition, unmodified single-stranded DNA oligonucleotides can be used to substitute single or several nucleotides. In particular, simultaneous substitution of four adjacent nucleotides was highly efficient, providing the opportunity to substitute virtually any given codon. We have used this method to create a codon substitution (N750F) in the Rb gene of mouse ES cells and show that the oligonucleotide-modified Rb allele can be transmitted through the germ line of mice.     

Identification of a genetic element (psr) which negatively controls expression of Enterococcus hirae penicillin-binding protein 5.

Enterococcus hirae ATCC 9790 produces a penicillin-binding protein (PBP5) of low penicillin affinity which under certain conditions can take over the functions of all the other PBPs. The 7.1-kb EcoRI fragment containing the pbp5 gene of this strain and of two mutants, of which one (E. hirae R40) overproduces PBP5 and the other (E. hirae Rev14) does not produce PBP5, was cloned in pUC18 and sequenced. In the 7.1-kb EcoRI fragment cloned from strain ATCC 9790, an open reading frame (psr) potentially encoding a 19-kDa protein was identified 1 kb upstream of the pbp5 gene. An 87-bp deletion in this element was found in the 7.1-kb EcoRI fragment cloned from strains R40 and Rev14. In addition, several base substitutions were found in the pbp5 genes of strains R40 and Rev14. One of these converted the 42nd codon, TCA, to the stop codon, TAA, in the pbp5 gene of Rev14. Escherichia coli strains were transformed with plasmids carrying the 7.1-kb EcoRI insert or a 2.6-kb HincII insert containing only the pbp5 gene of the three strains. Immunoblotting analysis of proteins expressed by these transformants showed that the 87-bp deletion in psr was associated with the PBP5 overproducer phenotype of strain R40 and the conversion of the TCA codon to the stop codon was associated with the PBP5 nonproducer phenotype of strain Rev14. None of the other nucleotide substitutions had any apparent effect on the level of PBP5 synthesized.     

Identification of phytochrome B amino acid residues mutated in three new phyB mutants of Arabidopsis thaliana

The growth and development of plants is regulated by light viathe action of photoreceptors which are responsive to the red/far-red,blue and UV regions of the spectrum. Phytochrome B (the apoproteinof which is encoded by the PHYB gene) is one of the red/far-redabsorbing photoreceptors active in this process. In this paper,the isolation and characterization of three new EMS-inducedmutations of Arabidopsis which confer phytochrome B deficiencyare described. Complementation analysis showed that these mutations(phyB-101, phyB-102 and phyB-104) were allelic with PHYB. DNAsequence analysis showed that all three mutants contain nucleotidesubstitutions in the PHYB-101 gene sequence. phyB-101 carriesa nucleotide substitution within the second exon of the PHYBgene. This G-to-A substitution is a missense mutation that convertsa glutamate residue at position 812 of the phytochrome B apoproteinto a lysine residue. phyB-102, another missense mutant, carriesa C-to-T substitution which converts a serine residue at position349 of the phytochrome B apoprotein to a phenylalanine residue.phyB-104 carries a premature stop codon as a result of a G-to-Amutation 1190 bp down-stream of the ATG start codon of the PHYBsequence. The missense mutations in phyB-101 and phyB-102 causesignificant alterations in the predicted second ary structureof their respective mutant polypeptides, and identify aminoacid residues playing crucial roles in phytochrome B function,assembly or stability. Key words: Arabidopsis thaliana, phytochromet, phyB mutants, missense mutations     

Characterization of the Fad12 mutant of Synechocystis that is defective in Δ12 acyl-lipid desaturase activity

The Fad12 mutant of Synechocystis sp. PCC 6803 has a defect in the desA gene for Δ12 acyl-lipid desaturase. We identified a change in the nucleotide sequence of the structural gene for the desaturase, in which a leucine codon has been converted to a stop codon. Western blot analysis revealed that the Δ12 acyl-lipid desaturase was localized in both plasma membranes and thylakoid membranes of wild-type cells but was absent from both types of membrane in Fad12 cells. These findings suggest that the desaturation of fatty acids takes place in both types of membrane in Synechocystis sp. PCC 6803. The mutation in the Δ12 desaturase did not affect the lipid composition of thylakoid and plasma membranes, but it changed the fatty acid composition of lipids in similar ways in both types of membrane.