Monosaccharide composition of the outer membrane lipopolysaccharide and O-chain from the freshwater cyanobacterium Microcystis aeruginosa NIES-87


  • The work was performed at the Department of Civil and Environmental Engineering, Graduate School of Engineering, Tohoku University.


Manabu Fujii, Department of Civil Engineering, Tokyo Institute of Technology, 2-12-1 Ookayama, Tokyo 152-8552, Japan. E-mail:



Bacterial lipopolysaccharide (LPS) protruding from the outermost layer of the outer membrane is expected to play an important role in cell physiology by interacting with molecules in the extracellular milieu; however, the structural and functional characteristics of these components in cyanobacteria remain largely unknown. We isolated water-soluble fractions of LPS and O-chain from the bloom-forming freshwater cyanobacterium Microcystis aeruginosa NIES-87 and identified their monosaccharide compositions.

Methods and Results

SDS-PAGE followed by silver staining demonstrated that the isolated total LPS was the smooth type with different numbers of repeating sugar units in the O-chain region. GC/MS analysis after acid hydrolysis, reduction and acetylation treatments indicated that the neutral monosaccharide components of the total LPS include glucose, rhamnose, mannose, galactose and xylose (in decreasing order of weight percentage), while only glucose was detected in the purified O-chain fraction. MALDI-TOF MS analysis suggested that the O-chain fraction is composed of repeating glucose and methylated glucose disaccharide units.


Our results indicate that the monosaccharide composition of M. aeruginosa O-chain is relatively simple.

Significance and Impact of the Study

Although further studies are necessary, these findings provide fundamental information for understanding the structural and functional properties of cyanobacterial LPS and O-chain.


Lipopolysaccharide (LPS) is an amphiphilic compound in the outer membrane of Gram-negative bacteria. In general, LPS is constructed from saccharide polymer, lipid and other adjunct elements such as phosphate and divalent cations. A core oligosaccharide region (R-core) covalently bridges a lipid anchored to the outer leaflet of the phospholipid bilayer (lipid A) with a side chain composed of a number of repeating oligosaccharide units (O-chain) (Wilkinson 1996; Nikaido 2003). LPS is thought to be an endotoxin in animal immune systems (Ulevitch and Tobias 1995; Raetz and Whitfield 2002), is a permeation barrier against exogenous lipophilic toxins (Labischinski et al. 1985; Nikaido 2003) and provides membrane structural integrity (Skurnik et al. 1999; Walsh et al. 2000) and cellular adhesion or recognition (Paradis et al. 1994; Shapiro et al. 1997). However, some evidence suggests that LPSs in cyanobacteria are structurally and functionally different from well-studied proteobacteria (Hoiczyk and Hansel 2000; Snyder et al. 2009). For example, recent studies have indicated that cyanobacterial LPS lacks strong endotoxic activity because of lipid A structures that differ from those generally conserved among proteobacteria (Rapala et al. 2002; Stewart et al. 2006). In addition, almost all cyanobacterial LPSs are devoid of 3-deoxy-d-manno-oct-2-ulosonic acid (KDO), heptose and phosphate components (Weckesse et al. 1974; Mikheyskaya et al. 1977; Schmidt et al. 1980a; Hoiczyk and Hansel 2000), which are commonly observed in the R-core region of proteobacterial LPS. Instead, other saccharide units such as glucose are the main components of some cyanobacterial LPS (Snyder et al. 2009). Therefore, the roles of cyanobacterial LPS other than endotoxicity are expected to be more vital for this type of organism to inhabit a range of ecological niches where extracellular conditions vary relatively rapidly.

Compared to proteobacterial LPS, few studies have investigated the structures and chemical compositions of cyanobacterial LPS. Monosaccharide compositions of LPS extracted by the hot phenol–water procedure, which includes both the R-core and O-chain saccharides, have been investigated in the freshwater cyanobacteria Synechococcus sp., Synechocystis sp. (Schmidt et al. 1980a,b), Microcystis aeruginosa (Martin et al. 1989) and the marine cyanobacteria Synechococcus sp. (Snyder et al. 2009). These studies indicated that neutral sugars, including rhamnose, fucose, xylose, mannose, galactose and glucose, are conserved among these cyanobacterial species. Although saccharide compositions of LPS in cyanobacteria are well known, information on the O-chain is sparse. Indeed, to our knowledge, no one has identified the cyanobacterial O-chain. Therefore, its structures and function remain largely unknown.

In this work, we determined the monosaccharide components of the complete LPS and O-chain in the outer membrane of the freshwater cyanobacterium M. aeruginosa. Microcystis is associated with a number of unfavourable environmental and social problems, including water deterioration (Codd 2000), toxin production (Kardinaal et al. 2007; Alexova et al. 2011) and coagulation inhibition in drinking water treatment systems (Takaara et al. 2010). In the light of its biological and engineering significance, it is important to investigate the chemical properties of the O-chain and other components of cyanobacterial LPS to understand the organism's interactions with extracellular environments.

Materials and methods

Cell preparation

An axenic culture of M. aeruginosa NIES-87, a nontoxic cyanobacterium isolated from Lake Kasumigaura, Japan, in 1982, was provided by the National Institute for Environmental Studies, Japan. Batch culture was performed using autoclave-sterilized MA medium at pH 8·6 (Andersen 2005) under light- and temperature-controlled conditions. The light was vertically supplied by fluorescent tubes in a 12-/12-h light/dark cycle. During incubation, illumination and temperature were maintained at 4000 Lux and 30°C, respectively. Acid-cleaned sterile Erlenmeyer glass flasks with a silicone sponge closure (Shin-Etsu Polymer, Tokyo, Japan) were used; they allow air permeation, but prevent bacterial contamination. Cellular growth was monitored by measuring the optical density at 660 nm (OD660) in a 1-cm cuvette with a UV–VIS spectrophotometer (Shimadzu UV-1600; Kyoto, Japan). Stationary-phase cultures (400 ml, OD660 = c. 0·70) obtained after 2-week incubation were used for further treatment, as a high density of cells was required to obtain sufficient quantities of LPS. Over the duration of laboratory incubation in this work, colony formation was not detected by the visual inspection.

LPS extraction

Cells were harvested by centrifugation for 10 min at 4000 g at 20°C. The pellet was washed by resuspension in 0·9% saline and recentrifuged. After decantation, the pellet was lyophilized by a vacuum freeze dryer (DC801; Yamato Scientific, Tokyo, Japan) and stored in the dark at −80°C. The classical hot phenol–water method (Westphal and Jann 1965) with some modifications described by Papageorgiou et al. (2004) was used to extract the LPS fraction from M. aeruginosa. A portion (0·5 g) of the lyophilized pellet was resuspended in 12 ml of autoclave-sterilized ultrapure water (18 MΩcm resistivity at 25°C, Milli-Q Academic A-10; Millipore, Tokyo, Japan) preheated to 68°C, followed by the addition of 1 volume of 90% (v/v) phenol at the same temperature. A homogenous water–phenol mixture was obtained by vigorously shaking the vessel at a high rotation rate for 20 min at 68°C, and then cooling it to 5°C. Centrifugation for 30 min at 4000 g at 5°C yielded three layers, including the soluble, intermediate and insoluble fractions. To avoid mixing of phenol with the sample, only the soluble fraction was collected. To maximize the LPS yield, another 12 ml of 68°C ultrapure water was added to the remaining intermediate and insoluble layers, and the soluble fraction was again collected after shaking, cooling and centrifuging as described above. The two aqueous fractions were mixed and dialysed using a dialysis tube with a molecular cut-off of 3·5 kDa (Spectrum Laboratories Inc., Shiga, Japan) against ultrapure water for 48 h to remove the phenol. The purified hydrophilic organic matter (HIOM) fraction was concentrated with a rotary evaporator, followed by lyophilization with the freeze dryer. RNA in the HIOM fraction was removed by digestion with ribonuclease A and subsequent dialysis, as follows: the HIOM fraction was dissolved in 0·1 mol l−1 Tris–HCl buffer (pH 7·4; Sigma, Tokyo, Japan) at a final concentration of 6·7 g l−1, and ribonuclease A (25 μg ml−1) was added. RNA was digested by incubating the solution for 6 h at 37°C. Complete RNA digestion was electrophoretically confirmed. The treated mixture was dialysed against ultrapure water for 24 h with a 100-kDa molecular cut-off dialysis tube (Spectrum Laboratories Inc.) and then ultracentrifuged for 4 h at 105 000 g at 20°C. A gel-like precipitate was resuspended in ultrapure water, and the purified LPS fraction was obtained after lyophilization.


The isolated LPS was analysed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using REAL GEL PLATE (15% Tris–HCl Gel; Biocraft, Tokyo, Japan), according to the manufacturer's instructions. A 100-μl aliquot of 0·1–1 mg ml−1 LPS dissolved in ultrapure water was added to the gel, followed by electrophoresis for 3 h at 20 mA and 20°C. The fractionated LPS bands were visualized by silver staining (Tsai and Frasch 1982). Immediately after electrophoresis, the LPS in the polyacrylamide gel was fixed in 200 ml of fixing solution (0·83 mol l−1 acetic acid in 40% ethanol). The fixing solution was tightly covered to avoid evaporation, and the gel was immersed overnight at c. 25°C. The fixing solution was then replaced by 200 ml of oxidizing reagent (the fixing solution containing 36 mmol l−1 HIO4), and LPS was oxidized for 5 min. The gel was washed three times by soaking in ultrapure water for 15 min. The gel was placed into 150 ml of staining solution (a mixture of 28 ml of 0·1 mol l−1 NaOH, 2 ml of concentrated ammonia and 5 ml of 1·2 mol l−1 silver nitrate in distilled water), soaked for 10 min and washed four times with ultrapure water. The LPS in the gel was stained by reducing the diamine silver ion in a developer solution (0·26 mmol l−1 citric acid in 0·019% formaldehyde solution) for 2–5 min. After the staining reaction was terminated in 200 ml of 58 mmol l−1 acetic acid, the gel was washed three times with distilled water. In all but the fixation step, the gel and solution were vigorously agitated with an orbital shaker. To avoid significant contamination, the staining solution container was rigorously washed with detergent and nitric acid prior to use.

Isolation of the O-chain fraction

The LPS fraction (10 mg ml−1) was dissolved in 1% acetic acid and mildly hydrolysed at 100°C for 2·5 h (Fedonenko et al. 2008; Wang et al. 2009). The sample was then centrifuged for 20 min at 13 000 g and 4°C. The supernatant was collected, lyophilized and used for further treatment. The precipitates, including the lipid A fraction, were discarded. To fractionate the O-chain, the lyophilized sample dissolved in ultrapure water (at concentration of c. 20 mg ml−1) was treated by gel filtration chromatography (AKTA FPLC) with a Superdex 30 prep grade column (separation range, <10 kDa; GE Healthcare, Tokyo, Japan). A 500-μL sample was introduced to the system, and 0·05 mol l−1 ammonium bicarbonate (pH 8) was used as a buffer with a flow rate of 0·8 ml min−1. The column eluate was monitored using a refractive index detector. As noted below, the O-chain was recovered by collecting the fraction eluted just after the void volume; this was followed by lyophilization.


The purified LPS and O-chain samples were hydrolysed by heating at 105°C for 2 h in 2 mol l−1 trifluoroacetic acid at concentrations of 1–2 mg ml−1 (Deweger et al. 1987). After acid hydrolysis, the monosaccharide sample was dried under a nitrogen atmosphere. A 0·5-mg monosaccharide sample was dissolved in 1 mol l−1 ammonia solution at 5 mg ml−1. Subsequently, 1 ml of 20 μg ml−1 sodium borohydride prepared in absolute dimethyl sulfoxide was added. The solution was heated at 40°C for 90 min to reduce the saccharides. Then, 0·1 ml of acetic acid was added to digest the excess sodium borohydride. To produce acetylated forms, 0·2 ml of 1-methylimidazole and 2 ml of acetic acid anhydride were added, mixed and vigorously shaken for 10 min. After acetylation, 5 ml of ultrapure water was added to hydrolyse the excess acetic acid anhydride. The solution was cooled to ambient temperature. Then, 1 ml of dichloromethane was added and vortexed to extract fully acetylated sugars in the lower layer. The sample was analysed by GC/MS (Agilent 5975C) using a (5%-phenyl)-methylpolysiloxane capillary column (HP-5MS, Agilent) and helium gas as a carrier. The column temperature was initially maintained at 180°C for 2 min, increased to 240°C at a rate of 2°C per min and then maintained for 10 min. Calibration was performed using identically treated standard monosaccharides including rhamnose, mannose, arabinose, xylose, glucose, galactose, N-acetylgalactosamine and N-acetylglucosamine (Wako Pure Chemical Industries, Ltd, Osaka, Japan). All monosaccharide standards used in this work were highly purified analytical grade (e.g. >98%) except for two saccharides (N-acetylgalactosamine and N-acetylglucosamine) with relatively lower purities (90–95%).

In the analysis of GC/MS data, we assumed that integration of signal peak area provides the proportional mass of each sugar tested. The observed peaks are generally attributed to the acetylated sugars that potentially include both the complete and its fragmented forms. Fragmentation of acetylated monosaccharide may occur under some conditions depending on the conditions of sample pretreatment and GC/MS measurement, resulting in relatively broad signal peaks when sugars are substantially fragmented. In addition, a width of signal peak region may be influenced by a resolution of separation by capillary column and any other experimental conditions. In this work, however, relatively sharp signal peaks were obtained in both samples and standards, suggesting that (i) the degree of fragmentation is negligibly small and (ii) the experimental conditions employed were suitable for the determination of monosaccharide species of interest. In the GC/MS data treatment, therefore, we assumed that fragmentation of acetylated monosaccharide is negligible.


The O-chain fraction was dissolved in ultrapure water at 1 mg ml−1. The sample solution was analysed by MALDI-TOF MS (Shimadzu AXIMA-CFR Plus) using DHB matrix and an m/z range of 1–2000. Measurements were taken in the reflectron mode.



SDS-PAGE analysis was performed for the LPS fraction isolated from Microcystis cells. As shown in Fig. 1, a ladder of bands was observed, confirming the presence of repeating O-chain sugar units in the water-soluble LPS extracts.

Figure 1.

SDS-PAGE electrophoretic analysis of lipopolysaccharide (LPS) isolated from Microcystis aeruginosa NIES-87. The left, middle and right lanes represent the purified LPS at 0·1, 0·5 and 1·0 mg l−1, respectively. A 100-μL aliquot of the sample was added to each lane. The ladder-like banding patterns show the repeating units present in the O-specific chain.

Monosaccharide composition of LPS

To examine the speciation of the monosaccharide in the LPS, GC/MS analysis was performed after acid hydrolysis of the isolated LPS, followed by reduction and acetylation treatments. The chromatogram showed five major peaks with different signal intensities (Fig. 2). By comparison with the retention times of the standard monosaccharides, the peaks were identified as rhamnose, xylose, mannose, glucose and galactose. Peak area integration provided the proportional mass of each sugar (Glc: 66%, Rha: 11%, Man: 9·3%, Gal: 8·7% and Xyl: 4·4%, Table 1).

Figure 2.

GC/MS profiles of purified water-soluble lipopolysaccharide (LPS) from Microcystis aeruginosa NIES-87 (upper panel) and standard monosaccharides (lower panel). The purified LPS and mixture of standard monosaccharides were subjected to the GC/MS analysis following the acid hydrolysis, reduction and acetylation treatments.

Table 1. Monosaccharide compositions of cyanobacterial lipopolysaccharide
 Microcystis aeruginosa Synechococcus PCCaSynechocystis PCCaMarine Synechococcus
NIES 87bPCC 7806cPCC 7820cUV 017cUV 006dNRC 1d690763076911660369086311631269106714680368076308WH 8102eCC 9311e
  1. Numbers in boldface represent the most abundant monosaccharide in each strain. Percentage of each monosaccharide relative to total monosaccharide detected was shown.

  2. a

    Schmidt et al. (1980a,b).

  3. b

    This work.

  4. c

    Martin et al.(1989).

  5. d

    Raziuddin et al. (1983).

  6. e

    Snyder et al.(2009).

Rhamnose11·210·27·47·5   37·1 36·2 37·4        1·94·0  
Fucose 11·41·67·5  7·96·710·6 42·1 9·111·3 42·9 8·32·39·415·8 25·7   
Mannose9·323·6 20·6 23·2  6·46·28·322·3 68·7 65·2 23·2 67·2 15·78·6 30·5 15·3 0·9
Galactose8·75·012·411·9  20·119·424·415·79·711·714·19·49·48·613·515·3  
Glucose 66·4 24·3 15·1 30·9 49·8 40·3 9·411·23·99·62·32·811·13·4 70·8 55·3 17·424·5 100 97·4
Glucosamine 5·04·54·616·228·85·55·65·19·15·65·27·67·51·92·35·06·4 1·7

Isolation and characterization of the O-chain fraction

The O-chain fraction was isolated by treating the LPS extract with 1% acetate. Mild hydrolysis followed by centrifugation typically yields O-chain and R-core in the supernatant (Fedonenko et al. 2008; Wang et al. 2009). The gel filtration chromatography profile in Fig. 3 demonstrated three major peaks for the O-chain, R-core and other crude compounds of low molecular weight (presumably, some LPS fragments). The purified O-chain was obtained by recovering the eluates in the first peak. Dry weight measurement of the collected fraction indicated that O-chain accounts for 31% of total LPS and 1·7% of parent cell mass. In contrast to total LPS, GC/MS analysis indicated that the O-chain polysaccharide of the Microcystis strain considered in this study is solely composed of glucose (or its derivatives) (Fig. 4). To obtain further information on the repeating structure, MALDI-TOF MS analysis was undertaken (Fig. 5) and it revealed four major peaks at 360·99, 698·95, 1036·94 and 1374·97 Da with the distance between the adjacent peaks being a uniform 338 Da. The secondary peaks at 536·97, 874·95 and 1213·00 Da were also equidistant.

Figure 3.

Gel chromatographic profile for the isolation of O-chain fraction. For the chromatographic separation, the sample was prepared by the mild hydrolysis of purified lipopolysaccharide with 1% acetic acid. After centrifugation, supernatant was introduced to the chromatographic system. During the chromatographic treatment, a refractive index of column eluate was continuously monitored. The O-chain fraction was retrieved by collecting the fraction eluted just after the void volume.

Figure 4.

GC/MS profile for the purified O-chain from Microcystis aeruginosa NIES-87. The O-chain fraction separated by the gel chromatographic technique was subjected to GC/MS analysis after the acid hydrolysis, reduction and acetylation treatment.

Figure 5.

MALDI-TOF-MS analysis of the O-chain isolated from Microcystis aeruginosa NIES-87. The distance between the peaks corresponds to the molecular weight of the O-chain repeating unit. The intensity was increased by a factor of 10 at molecular weights larger than 800 Da. Because of the measured low spectral intensity, no discernible peaks were identified at a molecular weight of >2000 Da.


SDS-PAGE showed that the LPS isolated in this work belonged to the smooth-type category. The ladder-like patterns in Fig. 1 indicate the presence of different numbers of repeating sugar units in the O-chain; the distance between adjacent bands is most likely to correspond to the molecular weight of the repeating units. While LPS with relatively high molecular mass can appear in the upper region of the electrophoretic image, those with a smaller number or lack of structural repetition (e.g. the rough or semi-rough types) would emerge in or near the bottom of the lane. The electrophoretic analysis indicates that the latter type of LPS was not produced to a significant extent by the water-soluble isolate, although rough LPS in the absence (or smaller amount) of O-chain is typically extracted into the phenol phase because of its hydrophobicity (Wilkinson 1996). The silver staining method employed here is not sensitive enough to detect LPS with acidic sugars, as silver ions may have less affinity to such sugars (Kido et al. 1990). Moreover, LPS with a low number of fatty acids may be lost in the fixing step of the visualization process (Fomsgaard et al. 1990). Therefore, alternative methods have been developed to recover certain types of LPS (Tan and Grewal 2002). However, as shown in Fig. 1, the classical technique was sufficiently sensitive to recover the LPS produced by M. aeruginosa, and the absence of acidic sugars in the LPS was further confirmed by GC/MS.

Monosaccharide analyses by GC/MS indicated that glucose is the most abundant monosaccharide in the LPS of M. aeruginosa NIES-87. The sugar species detected were comparable to those previously reported for freshwater and marine cyanobacterial LPS (Table 1). In a study by Martin et al. (1989), rhamnose, fucose, xylose, mannose, galactose and glucose were identified as the conserved neutral sugars in the hot phenol–water LPS extracts from three strains of M. aeruginosa (PCC7806, PCC7820 and UV017). Glucose was the most abundant monosaccharide in the dry mass of the strains PCC7806 and UV-017 and the second most abundant monosaccharide in PCC7820. Raziuddin et al. (1983) reported that glucose accounts for the majority of monosaccharide of LPS (c. 9–11%) in M. aeruginosa (UV 006 and NRC-1), although other monosaccharides were not analysed. Although strain-specific differences were seen, the monosaccharide compositions of M. aeruginosa reported in previous works were relatively similar to those determined in this work. Studies by Schmidt et al. (1980a,b) have also indicated that the major sugars common to the LPS of 12 strains of Synechococcus and Synechocystis sp. are fucose, mannose, galactose and glucose, while other sugars such as rhamnose are strain specific. Furthermore, a study on the LPS produced by the marine cyanobacteria Synechococcus sp. (Snyder et al. 2009) indicated that glucose is the main saccharide component of the water-soluble LPS (Table 1).

It is interesting that only glucose was identified in the gas chromatogram of O-chain polysaccharide (Fig. 4). In addition, MALDI-TOF MS analysis showed that the distances between major peaks were 338 Da. These results indicate that the single repeating unit of O-chain consists of two molecules of glucose or its derivative. Given that the molecular mass of two repeating glucose units is 324 Da because of dehydration by glycosidic bonds, it is most likely that one of the glucoses in the repeating unit is methylated.

The fact that glucose (and its derivative) is the sole monosaccharide component in the O-chain of M. aeruginosa implies that the functional roles of the O-chain may differ from its role in proteobacteria, wherein the O-chain structural configuration is generally much more complex. Indeed, the simple structure of cyanobacterial LPS and the lack of the important components found in proteobacteria, such as KDO and heptose, may indicate that this type of organism lacks the ability to synthesize more complex LPS because it is among the most ancient organisms on Earth (Snyder et al. 2009). Although the present work was undertaken by using the dispersed cell culture, cyanobacterial cells in natural environment may form colonies that are maintained by extracellular mucilage. In view of this, it would be necessary to investigate not only the LPS of individual cell but also the extracellular mucilage in future studies to provide a comprehensive understanding in interactions of this organism with external environments.

In summary, we investigated the monosaccharide composition of LPS and its O-chain fraction in the freshwater cyanobacterium M. aeruginosa NIES-87. The LPS was extracted from a dried cell pellet by the hot phenol–water method. SDS-PAGE followed by silver staining visually confirmed that the LPS in the water-soluble phase was smooth type, with different numbers of repeating sugar units in the O-chain. GC/MS after acid hydrolysis, reduction and acetylation treatments identified the major monosaccharide components present in the LPS fraction as rhamnose, xylose, mannose, galactose and glucose, with glucose being the most abundant monosaccharide by mass (66%). These five neutral sugars are common to other cyanobacterial LPS. Purified O-chain polysaccharides were recovered from the LPS extract by mild hydrolysis with 1% acetic acid and chromatographic isolation. GC/MS and MALDI-TOF MS indicated that the repeating unit of the O-chain consists of a disaccharide containing glucose and its methylated derivative. Although further research is necessary to understand the structural linkage and function of M. aeruginosa LPS, this study demonstrated the relatively simple composition of O-chain polysaccharide in this organism. To the best of our knowledge, this is the first report describing the monosaccharide composition of the cyanobacterial O-chain.


This work was partially supported by a Grant-in-Aid for Young Scientists (B) from the Japan Society for the Promotion of Science (23760501) and a Grant-in-Aid for Challenging Exploratory Research (21656130).