INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The microfossil record suggests that cyanobacteria or cyanobacteria-like prokaryotes were present on the primitive Earth in the Archaean era more than 3.5 billion years ago (28). The exquisite preservation of these microfossils is thought to reflect the intrinsic stability of the extracellular polysaccharide (EPS) and its ability to bind heavy metals as well as resist degradation (13). Extant cyanobacteria dominate the microbial populations of many extreme environments including soda lakes (Spirulina, Cyanospira), the nutrient-poor open ocean (Trichodesmium), thermal springs (Synechococcus and Mastigocladis), and the cold dry polar deserts (Chroococcidiopsis) (35). In these environments the cyanobacteria produce copious amounts of EPSs in the form of sheaths, slimes, and capsules. Very little is known about the diversity, mode of synthesis, structure, or properties of these biopolymers (19). A recent review emphasized the potential role of EPSs in the desiccation tolerance of prokaryotes (23). However, much further research is needed to resolve the specific mechanisms which biopolymers contribute to such a complex process.

The terrestrial cyanobacterium Nostoc commune has a marked capacity for desiccation tolerance and can survive storage at 400 MPa (0% relative humidity) for centuries (23). The cells produce large amounts of an unusual excreted polysaccharide that contributes in at least four ways to the marked stabilization of cells during prolonged storage in the air-dried state, at low or high temperatures. First, the glycan inhibits fusion of membrane vesicles during desiccation and freeze-drying (10) and acts as an immobilization matrix for a range of secreted enzymes which remain fully active after long-term air-dried storage (11, 27, 32). Second, the glycan provides a structural and/or molecular scaffold with rheological properties which can accommodate the rapid biophysical and physiological changes in the community upon rehydration and during recovery from desiccation. The glycan swells from brittle dried crusts to cartilaginous structures within minutes of rehydration. Third, the glycan matrix contains both lipid- and water-soluble UV radiation-absorbing pigments which protect the cell from photodegradation (12). Fourth, although epiphytes colonize the surfaces of Nostoc colonies, there is no penetration of the glycan due in part to a silicon- and calcium-rich pellicle and inherent resistance of the glycan to enzymatic breakdown. Preliminary structural work on one water-soluble UV-absorbing pigment (released from the glycan by acid hydrolysis) indicated the presence of an oligosaccharide (4), raising the possibility that the pigment may be covalently linked to the glycan in the desiccated state.

An understanding of the biochemical and biophysical properties of such biopolymers and the isolation of genes and enzymes required for their synthesis and modification can lead to an understanding of the underlying principles of extremophile stability. Furthermore, one can envision the utilization of such materials for the commercial stabilization of labile agricultural chemicals, food, pharmaceuticals, and/or biomedical materials. As part of an overall project aimed at understanding the functional genomics of extremophile biopolymers and the utilization of these materials for enhanced stability and performance, we determined the predominant structural unit of the glycan produced by desiccation-tolerant N. commune DRH-1.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Growth conditions. Cultures were grown in an air lift fermentor (2 liters) at 25°C in BG11₀ medium (25a). Both the fermentor and growth medium were autoclaved and subsequently inoculated with N. commune DRH-1 (250 ml) taken from a smaller culture. The cells were grown under an incident photon flux density of 1,750 µmol of photons m² s¹ for 2 weeks after which time the culture was harvested by a combination of centrifugation and filtration.

Isolation of the released polysaccharides. The cell-free supernatant fraction (ca. 1.5 liters) was passed through a tangential-flow filtration concentrator (10,000 molecular weight cutoff [MWCO]; Millipore Corp.), which reduced the volume approximately 10-fold. The solution was freeze-dried to provide an amber powder (1 to 2.5 g), which was dissolved in water (250 ml) and precipitated by pouring the mixture into a rapidly stirred solution of ethanol (95%, 750 ml). The insoluble material was recovered by filtration, washed successively with ethanol and acetone, and subsequently dried to provide a straw-colored material (500 to 750 mg). The material was then dissolved in water (200 ml) and passed through a cation-exchange resin (Dowex; H⁺ form) to generate the cation-free polysaccharide, which was obtained by freeze-drying as a white mass with the consistency of cotton (300 to 600 mg). Percent recoveries, based on the mass obtained after the tangential-flow filtration, ranged from 30 to 50%.

Formation and isolation of the oligosaccharides. Oligosaccharide fragments were obtained by partial acid hydrolysis using aqueous 1 M trifluoroacetic acid (TFA). The polysaccharide (300 mg) was dispersed in deionized water (184 ml) by stirring at room temperature (30 min). TFA (16 ml) was added, and the flask was placed in an oil bath (80°C) for 4 h. The mixture was evaporated to dryness under reduced pressure at 40°C and further evaporated three times with isopropyl alcohol (100 ml). The materials were then taken up by deionized water (8 ml) and centrifuged. The supernatant was freeze-dried, redissolved in deionized water (3 ml), and separated by gel filtration on a Bio-Gel P2 column (2.5 by 45 cm) eluted with deionized water (0.4 ml/min, 15 min/fraction). Carbohydrates were detected by spot testing (naphthoresorcinol reagent), and fractions were pooled according to thin-layer chromatography (TLC) on silica gel (1-butanol-formic acid-H₂O; 33:50:17). Selected fractions were further purified a second time on the same column; this time elution was with 0.1 M sodium acetate (NaOAc) (pH 3.7; 0.4 ml/min, 5 min/fraction). Fractions were again pooled based on the TLC results, desalted (Dowex; H⁺ form), and freeze-dried. The recovered oligosaccharides were then analyzed by gas chromatographic (GC) techniques and nuclear magnetic resonance (NMR) spectroscopy.

Methanolysis. Large-scale methanolyses were conducted with stirring in 50-ml screw-cap (Teflon-lined) test tubes at 80°C using methanolic HCl. The reaction solution was prepared by the addition of acetyl chloride (4 ml) to cold anhydrous methanol (MeOH; 21 ml, 40°C). The polysaccharide (100 to 150 mg) was added to this solution, and the mixture was sealed, allowed to warm to room temperature, and placed in an oil bath (80°C) for 16 to 24 h. Processing involved cooling the mixture and subsequently flushing the vial with nitrogen gas (15 min). This was followed by evaporation under reduced pressure (the receiving flask contained a small amount of pyridine to trap the gaseous HCl). Three subsequent additions and evaporations of MeOH provided a syrupy solid which could be purified by silica gel chromatography in acetonitrile-MeOH (90:10 [vol/vol]). The methyl uronates were eluted first, followed by the neutral methyl glycosides. Small-scale hydrolyses were performed in the same fashion to the point of silica gel purification. These syrups were subjected directly to trimethylsilylation (3) and GC analysis (specifics: GC-17A GC [Shimadzu]; Rtx-1 capillary column [30 m by 0.25-mm inside diameter, 0.25-µm thick film; Restek]; He carrier gas; temperature programming: 150°C for 2 min, then to 180°C at 15°C/min, hold for 1 min, then to 240°C at 5°C min; injection [split mode] and detector temperatures were 275°C).

Nosturonic acid structure determination. The methyl uronate fraction (45 mg) from a large-scale methanolysis was subjected to silica gel chromatography, elution with CHCl₃-ethyl acetate (1:1). Collection of the appropriate fractions provided the two anomeric forms. A portion of the mixture was subjected to an overnight NaBH₄ reduction in MeOH (36). The crude reaction product was added directly to a small ion-exchange column (Dowex; H⁺ form), followed by evaporation of the MeOH with a rotary evaporator. Borate was removed from the clear syrup by repeated additions and evaporations of MeOH, providing both anomers of methyl 3-O-[(R)-2-(1-hydroxy)propyl]-D-glucopyranoside (compound 3) in nearly quantitative yield. ¹³C NMR (CD₃OD, 49.0 ppm) results: R-3, 17.81, 55.45, 62.64, 67.25, 71.33, 73.67, 73.88, 80.19, 83.64, 101.06; R-3, 17.92, 57.27, 62.70, 67.93, 71.49, 75.04, 78.01, 80.70, 86.75, 105.29.

Synthesis of methyl 3-O-[2-(1-hydroxy)propyl]-D-glucopyranoside. Diacetone D-glucose (250 mg) was mixed with sodium hydride (150 mg) in dioxane at 65°C (6, 30). 2-Chloropropionic acid (R or S enantiomers; 260 µl) was added after 30 min, and the solution was stirred at 65°C overnight. The reaction mixture was then cooled, and the excess sodium hydride was quenched with water. Extraction with dichloromethane (3×) was followed by acidification of the aqueous phase by the addition of cold 3% HCl (100 ml). The acidified solution was extracted with dichloromethane (3×). This extract was processed to obtain a crude product, which was subjected directly to methanolysis (2 M TFA, 80°C for 16 h). Reaction workup followed by silica gel chromatography (acetonitrile [MeCN]) provided an anomeric mixture which was reduced in a solution containing NaBH₄ and NaOCH₃ in MeOH (36). The reaction product was purified by passage through a strongly acidic cation-exchange column followed by repeated additions and evaporations with MeOH (residual borate removal). The NMR spectra of the R- and S-hydroxypropyl derivatives were compared directly to those of the material produced by reduction of the native uronate. ¹³C NMR (CD₃OD, 49.0 ppm) results: S-3, 17.82, 55.55, 62.54, 67.79, 71.68, 73.22, 73.50, 80.43, 84.07, 101.50; S-3, 17.86, 57.39, 62.63, 67.79, 71.60, 75.00, 77.70, 80.69, 86.89, 105.52.

Nosturonic acid determination in field-isolated materials. Field samples were pulverized with a mortar and pestle under liquid nitrogen. The powder was suspended in MeOH (10 mg/ml) and extracted for at least 24 h with stirring. The solids were isolated by filtration and then subjected to methanolysis (150 mg), using the procedure described above. Processing provided a syrupy solid which was added to a silica gel column (26 g of silica) packed with acetonitrile. Stepwise elution with MeCN (100 ml) followed by MeCN-MeOH (9:1; 300 ml) provided separation of the uronics from the neutral sugars. The crude uronate fraction was subjected to NMR analysis for detection and estimation of nosturonic acid content.

Methylation analysis. Carboxyl reduction of the polysaccharide (50 mg) was conducted as described by Kim and Carpita (15). Sequential methylation of the reduced material (1 mg) using NaOH and CH₃I was performed as described previously (20). The resulting partially methylated alditiol acetates were analyzed by GC-mass spectrometry (MS), using the temperature program described by York et al. (38) (GC-MS specifics: VG 7070E-HF [double focusing, magnetic sector]; scan range, 35 to 400 atomic mass units at 1.5 s/scan; 70-eV electron impact ionization).

Periodate oxidation. Periodate oxidation and Smith degradation were performed as described by Severn and Richards (31). The polysaccharide (100 mg) was predispersed in deionized water (25 ml) and then oxidized with NaIO₄ (final concentration, 0.05 M; 50 ml) in the dark (4°C, 7 days). The reaction was terminated by the addition of ethylene glycol (0.4 ml), and the oxidized material was reduced with NaBH₄ (500 mg; 22°C, 15 h). The excess NaBH₄ was decomposed by adjusting pH to 6.5 with 1 M NaOAc. After extensive dialysis (MWCO = 3,500) against deionized water (3 days), the material was freeze-dried (90 mg) and then hydrolyzed with acetic acid (2%, [vol/vol], 100°C, 2 h). The degraded products were evaporated to dryness at 40°C as described above, dissolved in 0.1 M NaOAc (pH 3.7; 3.5 ml), and centrifuged. The supernatant was applied to a Bio-Gel P2 column (2.5 by 45 cm) eluted with 0.1 M NaOAc (pH 3.7; 0.4 ml/min, 15 min/fraction). The fractions were collected according to the TLC results.

Uronosyl cleavage with lithium and ethylenediamine. Nosturonic acid groups were removed according to the procedure of Mort and Bauer (18), as modified by Lau et al. (16). Polysaccharide (50 mg) was placed in a screw-cap tube along with a stir bar. Anhydrous ethylenediamine (7 ml) was added under argon, and the mixture was stirred at room temperature for 2 h. Hexane-washed lithium metal was added (a 3-mm length, 45 mg/cm), and the mixture turned a deep blue after about 5 min. Additional pieces were added over the course of the next 60 min to maintain a deep blue color. The reaction mixture was then cooled in an ice bath, and water was added slowly (5 ml) to quench the excess lithium. Once addition was complete, the solution was transferred to a round-bottom flask and toluene was added. This solution was evaporated under reduced pressure, and the toluene addition and evaporation procedure was repeated two additional times to provide a syrupy solid. The solid was dissolved in water, and the pH was adjusted from 10.5 to 6.4 by the addition of acetic acid. This solution was passed through a cation-exchange resin (200 ml; Dowex 50WX8-200; H⁺ form), collecting the fractions which were positive to naphthoresorcinol. The combined fractions were freeze-dried to a powder (31 mg) and further purified on a Bio-Gel P2 column with water, collecting the solutions that were eluted prior to the void volume. Freeze-drying provided the "uronosyl-free" polysaccharide (18.7 mg).

Spectroscopy. NMR spectra of the oligosaccharides were recorded on either a Bruker AM360 or a JEOL Eclipse-plus 500-MHz NMR at 27°C using the software supplied by the manufacturer. Samples analyzed in D₂O were freeze-dried from D₂O (99.9%) and subsequently dissolved in D₂O (600 µl, 99.996%) which contained acetone (0.5 µl) as the internal standard (¹H shift, 2.225 ppm; ¹³C shift, 31.07 ppm). Samples analyzed in CD₃OD were referenced to the central solvent peak (¹H shift, 3.30 ppm; ¹³C shift, 49.0 ppm). The ¹³C NMR spectrum of the purified EPS was obtained on a Varian Unity 400 (D₂O, 40°C) using a 10-mm wide-bore probe. Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) spectra were obtained with a Kratos Kompact MALDI-TOF MS using a dihydroxybenzoic acid matrix.

Spectrophotometric assays. Sulfate and phosphate determinations were performed as described previously (1, 33).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Compositional analysis. The five predominant sugars found in the purified polysaccharide by GC analysis of the trimethylsilyl methyl glycosides were ribose, xylose, glucose, galactose, and uronic acid. Smaller amounts of mannose and glucuronic acid were also found. The molar ratio of ribose to xylose to glucose to galactose to uronic acid was determined to be 1:1:2:1:1, and the neutral carbohydrates were all of the D configuration (39). Spectrophotometric assays for sulfate and phosphate were negative. Silica gel chromatography (MeCN-MeOH) of the methanolysis products provided the crude anomeric uronic acid esters, which were separated into their individual anomers by a subsequent silica gel purification using CHCl₃-ethyl acetate. Both anomers contained two carbonyl groups, two methyl esters, and one methyl aglycone. The upfield doublet (3H, 1.4 ppm; J = 8 Hz) which correlated with a quartet (1H, 4.5 ppm) was an isolated spin system indicative of a lactyl moiety. The ¹H and homonuclear correlation (COSY) spectra showed a carbohydrate coupling network of a uronic acid with all vicinal ring protons disposed in a trans fashion. The heteronuclear multiple band correlation (HMBC) spectrum provided a strong correlation between the lactyl methine and carbon-3 of the uronate, proving the structure was that of a 3-O-lactyl uronate.

View larger version (16K): [in this window] [in a new window]   FIG. 1.   Synthesis of the - and -anomers of methyl 3-O-[2-(1-hydroxy)propyl]-D-glucopyranoside (structure 3), starting from both diacetone D-glucose and the purified polysaccharide. Redn, reduction. See text for details.

Characterization of oligosaccharide fragments. Since preliminary experiments aimed at degrading the viscous polysaccharide by the use of enzymes were not successful, acid hydrolysis and periodate oxidation were used to fragment the polymer and subsequently elucidate the structure. Mild acid hydrolysis of the polysaccharide (0.1 M TFA, 70°C, 4 h) and analysis of the methanol-soluble fraction provided only D-ribose, in amounts similar to the concentration determined by the compositional analysis. Stronger acid hydrolyses (1 M TFA, 80°C, 4 h) and subsequent size exclusion chromatography (Bio-Gel P2) provided a disaccharide, a trisaccharide, and oligosaccharides that were characterized by NMR spectroscopy (Tables 1 to 3).

View larger version (22K): [in this window] [in a new window]   FIG. 2.   Structures of the oligosaccharides isolated from the polysaccharide.

View larger version (24K): [in this window] [in a new window]   FIG. 3.   Anomeric and ring proton regions of an absolute-value COSY spectrum (non-PFG) of trisaccharide 5. The important correlations are indicated (see inset for key).

View larger version (46K): [in this window] [in a new window]   FIG. 4.   The -anomeric and ring proton regions of a 500-MHz gradient-enhanced absolute-value COSY spectrum of a mixture of pentasaccharides 7 and 8. The boxed assignments indicate protons with HMBC correlations to anomeric carbons on adjacent rings (see inset for key).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Although lactyl-containing uronic acids were reported previously in cyanobacteria (7), the 3-O-lactyl glucuronic acid described here was reported only once before: in the exopolysaccharide produced by a strain of the bacterium Alteromonas (NMR spectral data not determined) (6). This organism was originally found on the epidermis of a worm (Avinella caudata) growing near a deep sea hydrothermal vent in the eastern Pacific Ocean (6). Since the uronic acid is present in an organism as globally significant as N. commune, we have named this compound nosturonic acid (NosA). Since both Alteromonas and N. commune are found in so-called "extreme environments," nosturonic acid or uronic acids with lactyl moieties may play pivotal roles in the ability of organisms to survive under harsh conditions. Such a functional group can act as a "spacer arm" or "linker," and could aid in adherence of the EPS to inorganic or organic surfaces (biofilms) and/or allow covalent attachment of UV-absorbing pigments or adjacent polysaccharide chains (molecular scaffold).

The field samples that tested positive for the presence of nosturonic acid belong to "form species" N. commune as defined through group I intron analysis (D. Wright, T. Prickett, R. F. Helm, and M. Potts, submitted for publication). Nosturonic acid could not be detected in N. commune UTEX 584 when grown under the same conditions used for N. commune DRH-1. N. commune UTEX 584 is a culture collection strain of unknown origin (and probably incorrectly named) that does not belong in the form species group on the basis of intron analysis (Wright et al., submitted). As such nosturonic acid may be a marker for the EPS of a restricted group of cyanobacteria.

Previous reports on the extracellular polysaccharides of cyanobacteria have suggested that their structures may not be comparable to those of algae, bacteria, or fungi (19). In essence the question of regularity (repeat unit or averaged structure) is considered open, as conflicting evidence has been reported. In our work it appears that the N. commune EPS does contain a predominant repeat unit under the specific conditions used in our experiments. Since mannose and glucuronic acid were found in the original EPS preparations but were not found in any of the oligomers we have investigated, we assume the these carbohydrates were derived from degraded cellular material and/or capsular polysaccharide. The possibility that these groups are also present in the EPS in very small amounts does, however, exist. In this context it should be pointed out that the physical properties of the EPS change according to environmental conditions and that growth for periods greater than the 2-week period used here provides a slightly different polysaccharide (data not shown). This suggests that there is some degree of flexibility in the sequence of, and control over, the polysaccharide assembly process. As a consequence, cyanobacteria may produce a polysaccharide with a specific linkage pattern under one set of environmental conditions but as the environmental cue changes (extreme heat, lack of water) the polysaccharide structure could be modified to insure the viability of the organism. This makes the structural analysis of cyanobacterial polysaccharides quite challenging especially for field materials, as they may contain several polysaccharides, each representing the recent environmental history of the location. Such behavior is not without precedent (5), and, as we have reported here, different amounts of nosturonic acid were measured in the field-grown materials of N. commune from different geographic locations.

The presence of ribose in the N. commune EPS is a novel feature. There are scattered reports of ribose in the extracellular polysaccharides of cyanobacteria (5), but this is the first unambiguous identification. Ribose is well known as a component of the lipo- and capsular polysaccharides from many gram-negative bacteria, where it is found exclusively as a -furanosyl residue (8, 17, 37). Why does a polysaccharide involved in the protection of an organism from an extreme environment have a carbohydrate as labile as ribose? One could theorize that the moiety protects neighboring glycosidic bonds from the more common glycan hydrolases. Under this scenario, the selective removal of the ribose group should leave the polysaccharide more susceptible to enzymatic depolymerization (Z. Huang, T. Prickett, M. Potts, and R. F. Helm, submitted for publication). Another possibility is that, because N. commune is restricted to neutral and/or alkaline environments, the acid-labile nature of ribose is never an important factor. Autoclaving the crude EPS results in a decrease in solution viscosity, and free ribose was detected in the resulting aqueous solution by TLC. This qualitative observation supports the possibility that ribose is partially responsible for the gelatinous consistency of the native material (viscosity modifier).

Colonies of N. commune are a conspicuous feature of nutrient-poor terrestrial soils on all continents from the tropics to the polar regions (24). Desiccated crusts are brittle and friable but have the consistency of cartilage when rehydrated. The massive and rapid swelling of desiccated colonies following rainfall is sufficiently striking that it was even the subject of medieval folklore (25). These rheological properties of the glycan, as well as its resistance to degradation in situ, its ability to prevent membrane fusion upon removal of water, its capacity to immobilize the water stress protein Wsp and UV-absorbing pigments, and its significant contribution to the dry weight of colonies emphasize the pivotal role for this biopolymer in the biology of N. commune, particularly the capacity for desiccation tolerance. The solving of the structure of the glycan provides crucial data for investigating how its synthesis is regulated, how it is modified through environmental stress, and how it interacts with other components of the response to desiccation (secreted UV-absorbing pigments and carbohydrate-modifying enzymes).