Molecular Genetics and Genomic Analysis of Scytonemin Biosynthesis in Nostoc punctiforme ATCC 29133

Strains, cultivation, and media.Inocula for all experiments were obtained from stocks of N. punctiforme ATCC 29133 (PCC 73102) that were kept on agar plates under axenic conditions at 25°C. The base growth medium was AA of Allen and Arnon (1). For liquid batch cultures, a fourfold-diluted formulation was used (AA/4), while full strength was used for plates solidified with 1.5% (wt/vol) Noble agar (Difco). The medium was supplemented as needed with 5 mM MOPS (morpholinepropanesulfonic acid) and 2.5 mM NH₄Cl at pH 7.8 (AA/4+N). Visible light was provided by cool-white fluorescent tubes at an intensity of 7 W m⁻²; UV-A irradiation was provided by 20-W black-light fluorescent tubes (General Electric) at an approximate intensity of 10 W m⁻². The spectral output of these UV-A lamps has a maximum at 365 nm and has been previously described (19). Escherichia coli strains were cultured in Luria-Bertani (LB) broth at 37°C (40), either in liquid or on plates solidified with 1.5% (wt/vol) Bacto agar (Difco). The growth of batch E. coli cultures was monitored by the optical density at 600 nm (OD₆₀₀). Media were supplemented with antibiotics, as needed, with final concentrations of 25 μg ml⁻¹ each for kanamycin and neomycin, 1 μg ml⁻¹ for streptomycin, and 60 μg ml⁻¹ for chloramphenicol (47).

Transposon mutagenesis and screening of mutants.Strains and plasmids used in this study are listed in Table 1. Putative mutants incapable of synthesizing scytonemin were sought using transposon mutagenesis in a triparental conjugation system (12). Briefly, this system was composed of the cyanobacterial recipient strain, an E. coli HB101 (7) donor conjugal strain carrying the broad-host-range conjugal plasmid pRK2013 (15), and an E. coli DH5α (20) donor cargo strain carrying the transposon Tn5-1063a on plasmid pRL1063a (47). To prepare N. punctiforme for conjugation, filaments were sonicated using a microtip sonicator to produce short fragments of three to seven cells and allowed to recover on an orbital shaker at 25°C under white light in AA/4+N before mating. E. coli strains were grown to an OD₆₀₀ of 0.6 in LB broth supplemented with kanamycin for the strain carrying pRK2013 and with both neomycin and streptomycin for the strain carrying pRL1063a. The cells were combined and washed twice in plain LB broth before mating. For conjugation, 0.5 ml of the E. coli mixture at an OD₆₀₀ of 9 to 10 and 0.5 ml of N. punctiforme at a concentration corresponding to 75 μg of chlorophyll a ml⁻¹ were combined and spread onto detergent-free membrane filters (Millipore HATF082), which were then placed onto AA/4+N plates supplemented with 0.5% (vol/vol) LB broth. The plates were incubated overnight under low light (2 to 3 W m⁻²), after which the filters were transferred to AA/4+N plates and incubated for two additional days under low light to allow for gene expression. The filters were subsequently transferred to AA/4+N plates supplemented with streptomycin and neomycin to allow for the growth of only mutant colonies with resistance to streptomycin and neomycin as afforded by the transposon. Mutant colonies began to appear after approximately 10 days.

To identify colonies deficient in scytonemin production, mutant colonies were exposed to white light supplemented with UV-A irradiation (approximately 10 W m⁻² each) to elicit the biosynthesis of the sunscreen (17). After 2 weeks, approximately 400 mutant colonies were examined microscopically for changes in extracellular pigmentation. For this examination, a piece of each colony was retrieved using watchmaker's forceps under a dissecting microscope, and a wet mount was prepared for bright-field microscopy. Putative scytonemin-deficient mutants were then examined spectroscopically to confirm the phenotype after extraction in acetone (17). One putative mutant, SCY 59, lacked the peak at 384 nm characteristic of scytonemin; this mutant was the subject of investigation.

Determination of the site of transposon insertion.In order to obtain information on the specific site of transposon insertion in the genome of SCY 59, total genomic DNA was extracted using a commercial kit (Ultra Clean plant DNA isolation kit; MoBio Laboratories Inc.) and digested with EcoRI; fragments were then ligated using T4 DNA ligase (5). The resulting plasmids, some of which contained the transposon (Tn5-1063a) and the genomic regions flanking the insertion, were used to chemically transform E. coli DH5α-MCR (20), and transformants were selected on the basis of streptomycin and neomycin resistance as afforded by the transposon. Three transformant colonies that contained plasmids with the transposon and SCY 59 genomic flanking regions were independently obtained. The plasmids were isolated using a commercial kit (standard mini plasmid prep kit; MoBio Laboratories Inc.) and sequenced using a primer (5′ TACTAGATTCAATGCTATCAATGAG 3′) complementary to the end of the transposon sequence. Only one plasmid (pTD59a) was chosen for use in further experiments. Unknown portions of pTD59a were then sequenced through the use of primer walking. All enzymes used in this study were from Fermentas Life Sciences, and all DNA products were confirmed by electrophoresis using 1% agarose gels.

Verification of single-site transposition and reconstruction of the mutation.Southern analysis was performed to verify that a single transposition event had occurred in the genome of SCY 59. Southern blotting (40) was performed according to the directions of the manufacturer of the Gene Images AlkPhos direct labeling and detection system (Amersham Biosciences). Briefly, a total of 0.5 to 1.0 μg of whole genomic DNA was isolated as described above and digested with EcoRI. DNA was transferred onto a Zeta-Probe GT membrane (Bio-Rad) by using 0.4 M NaCl. The probe was designed to hybridize to a 373-bp region of the transposon and was labeled according to the manufacturer's instructions. The chemifluorescent signal generated was detected using Quantity One one-dimensional analysis software on a Fluor-S MultiImager (Bio-Rad).

To confirm the phenotype of SCY 59, the mutation was reconstructed through the direct insertion of a transposon into the wild-type strain at the same locus of insertion found in SCY 59. In preparation for this reconstruction, the transposase gene on pTD59a was inactivated through a frameshift mutation introduced by filling in a unique NotI site to yield plasmid pTD59b. To create a positive-selection vector, pTD59b was digested with EcoRV and subsequently ligated to the sacB-containing EcoRV fragment of pRL1075 (6) to obtain pTD59c. This final plasmid conferred sensitivity to 5% (wt/vol) sucrose and resistance to chloramphenicol, streptomycin, and neomycin (12). To reconstruct the mutation, pTD59c was inserted into wild-type N. punctiforme cells by the triparental conjugation method described above, except that pTD59c served as the cargo plasmid instead of pRL1063a. A single-recombinant strain, SCY S7301, was selected by its resistance to chloramphenicol and sensitivity to sucrose, while the double-recombinant strain, SCY M7301, was selected by its sensitivity to chloramphenicol and resistance to sucrose due to the loss of the EcoRV fragment of pRL1075 and its resistance to streptomycin and neomycin due to the insertion of the transposon (9). SCY M7301 was then assayed for scytonemin production upon UV-A irradiation by using the lipid-soluble-pigment extraction procedure described above (17).

Comparative mutant and wild-type characterizations.Growth rates of wild-type N. punctiforme and SCY 59 in batch cultures were determined using AA/4 liquid medium placed under white light and white light supplemented with UV-A irradiation with shaking at 25°C. The increase in biomass was monitored by measuring the chlorophyll a concentrations in 2-ml aliquots, from which chlorophyll a was extracted with 90% methanol (31), taken daily in triplicate through exponential phase and into stationary phase of growth.

For the evaluation of the lipid-soluble pigment complement, pigments were extracted from whole cells in triplicate with acetone by using previously published protocols (17) and assayed using high-pressure liquid chromatography (HPLC). Prefiltered, 50-μl concentrated extracts were used on a Waters Spherisorb S10 ODS2 (10-by-250-mm) semiprep analytical column filled with reverse-phase silica gel and run on a variable solvent gradient at a pressure of approximately 38 × 10⁵ Pa. The gradient varied from 15% water in methanol to pure methanol in 6 min, then to pure acetone in 4 min, at which point the pure acetone was sustained for 5 min, and then back to 100% methanol for a total of 35 min at a flow rate of 1.8 ml min⁻¹. Carotenoids, chlorophyll a, and scytonemin were detected continuously using an online photodiode array detector in the 384-to-665-nm range. Pigments were identified by absorption spectra and quantified using previously published extinction coefficients for chlorophyll a, β-carotene, and myxoxanthophyll (27), echinenone (26), and scytonemin (19) with a chlorophyll a standard.

For the phycobilin complement, cultures were taken in triplicate and resuspended in a 20 mM sodium acetate buffer at pH 5.5. Cells were broken open using a French press at 1,000 lb/in², and the crude cell extracts were precipitated with 1% (wt/vol) streptomycin sulfate at 4°C to eliminate membrane fragments containing chlorophyll (45). The supernatant extracts were read from 350 to 750 nm on a spectrophotometer, and the absorbance values corresponding to the phycobiliproteins phycoerythrin (565 nm), phycocyanin (620 nm), and allophycocyanin (650 nm) were measured. Specific contents were established using a series of trichromatic equations (8, 45) and subsequent normalization of the absorbance values to biomasses (dry weights) determined gravimetrically.

For microscopic examination, all cultures were grown in AA/4 liquid medium as described above. All photographs were of wet-mount preparations on agar-covered slides, taken on a Zeiss AxioImager 8.1 Nomarski microscope using an Olympus C-7070 wide-zoom digital camera.

Comparative genetic and phenotypic characterization of N. punctiforme ATCC 29133 and SCY 59.We obtained approximately 400 mutant colonies after transposon mutagenesis. A microscopic observation of these colonies after the exposure of the plates to UV-A irradiation, the typical environmental cue that elicits the synthesis of scytonemin in cyanobacteria (17), revealed a single putative scytoneminless mutant, SCY 59, which had colorless sheaths (Fig. 1). The phenotype was then confirmed with spectroscopic analyses of lipid-soluble extracts from cells grown under UV-A light. SCY 59 was consistently unable to produce scytonemin, even under higher intensities of UV-A irradiation or longer exposure times than those sufficient for the induction of synthesis in the wild type.

The transposon used, Tn5-1063a, contains a plasmid origin of replication and antibiotic resistance markers. These properties allowed us to obtain plasmid clones of the transposon and regions flanking the insertion site from SCY 59 genomic DNA through the excision of these regions by restriction endonuclease cleavage (Materials and Methods). Three independent clones were thus obtained and sequenced using a primer complementary to the end of the Tn5-1063a sequence. Sequences obtained from the three clones were identical, indicating that the insertion site was common to all three and strongly suggesting, in fact, that SCY 59 contained a single insertion site. To confirm that a single-site transposition event had occurred in SCY 59, we also performed a Southern analysis of the mutant's genomic DNA after digestion with EcoRI. By using a probe specific to the transposon (Materials and Methods), a single band (corresponding to the expected fragment size of about 11 kb) was detected, confirming that a single copy of the transposon was present in the genome of SCY 59 (data not shown).

In order to confirm that the phenotypic differences between the wild type and SCY 59 were restricted to scytonemin synthesis, we characterized some physiological and morphological traits of both strains. Growth rates of the wild type and SCY 59 under white light were not significantly different (n = 3 cultures; P < 0.05). The number of doublings per day, represented as μ (24), in exponential phase was 0.994 ± 0.17 for the wild type, while the value for SCY 59 was 0.959 ± 0.06. Even upon exposure to UV-A irradiation, the growth rates of the wild type and SCY 59 (0.972 ± 0.06 and 0.905 ± 0.19 doublings per day, respectively) were not significantly different (n = 3; P < 0.05).

With respect to the lipid-soluble pigment complement, HPLC analyses confirmed that the only major pigment difference between the two strains was the lack of scytonemin in SCY 59 under conditions of induction by exposure to UV-A irradiation. All other pigments, including chlorophyll a and the carotenoids, myxoxanthophyll, echinenone, and β-carotene, were present in the wild type as well as in SCY 59, and their specific contents were not significantly different (n = 3; P < 0.05) under white light or white light supplemented with UV-A irradiation (Table 2; Fig. 2). In water-soluble extracts from the wild type and SCY 59 exposed to white light or white light supplemented with UV-A irradiation, no significant difference (n = 3; P < 0.05) in the phycobilin contents or compositions was found (Table 2). Additionally no obvious differences between the strains in colony morphology, vegetative cell shape or size, filament formation, or the ability to form heterocysts were detected.

• Open in new tab
• Download powerpoint

FIG. 2.

HPLC chromatograms for the N. punctiforme wild-type and mutant SCY 59 lipid-soluble pigment extracts under white light (A) and under white light supplemented with UV-A irradiation (B). The mutant SCY 59 chromatograms in both panels are right shifted and up shifted 3% relative to those of the wild type to facilitate differentiation between the data for the two strains. Pigments are labeled as follows: a, myxoxanthophyll; b, echinenone; c, chlorophyll a; and d, β-carotene.

Reconstruction of the mutation.The evidence presented above strongly suggests that the mutant phenotype of SCY 59 was restricted to an inability to produce scytonemin and that this characteristic coincided with the interruption of a particular region in the genome of N. punctiforme by a single insertion of the transposon. However, it was still a possibility that the two findings were coincident but independent and that the phenotype was the result of a spontaneous mutation elsewhere in the genome. To establish a cause-effect relationship, it was then necessary to reconstruct the mutation in the wild type at the same location of the original mutation found in SCY 59. This reconstruction was accomplished by ligating a DNA fragment encompassing the transposon and its flanking region (pTD59a) (Table 1) to a positive-selection vector to create a plasmid with regions complementary to the original insertion found in the genome of SCY 59. This plasmid was then recombined into the wild type strain by triparental conjugation (Materials and Methods). Double-crossover recombinants were then selected by the antibiotic resistance encoded in the transposon. The double-crossover recombinant strain resulting from this site-directed mutation (SCY M7301) was incapable of synthesizing scytonemin upon UV-A irradiation induction, thus reproducing the phenotype of SCY 59 and confirming the cause-effect relationship between this mutation and the scytoneminless phenotype.

Characterization of the ORF involved in scytonemin biosynthesis and the surrounding genomic region.Primer walking on one of the plasmid clones obtained previously from the SCY 59 transposon insertion region (pTD59a) (Table 1) yielded 2,495 bp of the genomic sequence flanking the insertion. This fragment is 98% identical to a region of the N. punctiforme genome that includes a fragment of ORF NpR1272, the complete ORF NpR1273, and a fragment of ORF NpR1274 as annotated in the N. punctiforme genomic database. The original mutation involved the interruption of NpR1273, which is the fourth in a series of 18 ORFs that are in relatively close proximity (NpR1276 to NpR1259) (Fig. 3) and are transcribed in the same direction. The nearest ORF upstream from NpR1276, NpF1277, is transcribed in the opposite direction, as is the nearest ORF downstream from NpR1259, NpF1258. Therefore, the region from NpR1276 to NpR1259 constitutes a potentially functional genetic unit and was selected for more-detailed analyses.

• Open in new tab
• Download powerpoint

FIG. 3.

Genomic region in N. punctiforme associated with scytonemin biosynthesis. Arrows denote ORFs and indicate transcriptional orientations. The ORFs are labeled according to annotations in the N. punctiforme database (Table 3). NpR1273 was the ORF interrupted by the transposon in SCY 59, and the ORFs involved in the biosynthesis of the aromatic amino acids are represented as solid arrows. The thick line delimits the region that was sequenced in this study. The nearest ORFs outside of this region are indicated by gray arrows, and the hatch marks indicate a break in the distance scale.

Most of the ORFs in the downstream region of this cluster (NpR1269 to NpR1260) are predicted to encode a set of enzymes involved in the shikimic acid and aromatic amino acid biosynthesis pathways (Fig. 3) (2). Copies of most of these ORFs are also found elsewhere in the genome at dispersed loci, as is the case for most other cyanobacterial aromatic amino acid biosynthesis genes. In fact, these dispersed genes are the only copies that exist in most other cyanobacterial lineages (49). Most of the upstream ORFs within this cluster (NpR1276 to NpR1270) encode products annotated as hypothetical proteins and have relatively similar syntenic orthologues in Anabaena sp. strain PCC 7120, a related cyanobacterium. Aside from the products of these orthologues, the putative proteins in the public databases have less than 36% identity to the proteins encoded by these ORFs. The NpR1274, NpR1273, NpR1272, and NpR1271 products all contain NHL repeats (Table 3) that resemble the WD-40 repeat (a unit of about 40 residues with tryptophan and aspartic acid at defined positions), which would classify them as conserved hypothetical proteins in the beta propeller clan (33). However, this structural resemblance does not imply a functional similarity, since the protein family is functionally heterogeneous.

NpR1273, the mutated ORF, encodes a putative protein of 423 amino acid residues with a predicted signal peptide (amino acids 1 to 25) and a putative cleavage site between amino acids 26 and 27 (37). This predicted protein sequence has 71% identity to an uncharacterized protein in Anabaena strain PCC 7120. The next best protein sequence match for the NpR1273 product is another hypothetical protein from Anabaena strain PCC 7120, at 40% identity, but there are no other protein matches in the pubic databases with an identity greater than 30%. It is interesting that NpR1273 and NpR1272 are approximately 67% similar to one another, which may suggest an ancestral duplication event. Characteristics of putative genes in this cluster as determined by genomic analyses are found in Table 3.

Although we do not have the evidence to conclude that these genes constitute an operon, they are in relatively close proximity to one another, with no more than a 479-bp gap between any two adjacent ORFs from NpR1276 to NpR1259. A search for obvious promoter sequences (48) within this region was not successful, although this result is not surprising since promoter consensus sequences in cyanobacteria are generally not well defined (13).