Flexibility-Rigidity Coordination of the Dense Exopolysaccharide Matrix in Terrestrial Cyanobacteria Acclimated to Periodic Desiccation

ABSTRACT

A dense exopolysaccharide (EPS) matrix is crucial for cyanobacterial survival in terrestrial xeric environments, in which cyanobacteria undergo frequent expansion and shrinkage processes during environmental desiccation-rehydration cycles. However, it is unclear how terrestrial cyanobacteria coordinate the structural dynamics of the EPS matrix upon expansion and shrinkage to avoid potential mechanical stress while benefiting from the matrix. In the present study, we sought to answer this question by investigating the gene expression, protein dynamics, enzymatic characteristics, and biological roles of WspA, an abundantly secreted protein, in the representative terrestrial cyanobacterium Nostoc flagelliforme. The results demonstrated that WspA is a novel β-galactosidase that facilitates softening of the EPS matrix by breaking the polysaccharide backbone under substantial moisture or facilitates the thickening and relinkage of the broken matrix during the drying process, and thus these regulations are well correlated with moisture availability or desiccation-rehydration cycles. This coordination of flexibility and rigidity of the cyanobacterial extracellular matrix may contribute to a favorable balance of cell growth and stress resistance in xeric environments.

IMPORTANCE How the exopolysaccharide matrix is dynamically coordinated by exoproteins to cope with frequent expansion and shrinkage processes in terrestrial colonial cyanobacteria remains unclear. Here we elucidated the biochemical identity and biological roles of a dominant exoprotein in these regulation processes. Our study thus gained insight into this regulative mechanism in cyanobacteria to combat periodic desiccation. In addition, the filamentous drought-adapted cyanobacterium Nostoc flagelliforme serves as an ideal model for us to explore this issue in this study.

INTRODUCTION

Cyanobacteria are widespread in both terrestrial and aquatic habitats. The fascinating Nostoc group of cyanobacteria forms colony morphotypes with filamentous, lamellate, spherical, or spheroid shapes (1–5). These include terrestrial (Nostoc flagelliforme and Nostoc commune) and aquatic (Nostoc sphaeroides, Nostoc verrucosum, Nostoc zetterstedtii, and Nostoc pruniforme) species. Much of the success of terrestrial Nostoc species is related to their ability to withstand long-term desiccation and to recover metabolic activity after rewetting (6, 7). Exopolysaccharide (EPS) is believed to play a key role in providing protection for cyanobacteria against environmental stresses (8–10). In arid regions with intense solar radiation and periodic desiccation-rehydration cycles, the density of the EPS matrix appears to be crucial for colonial Nostoc species. A dense EPS matrix accommodates additional extracellular components, such as UV-absorbing pigments and antioxidant proteins, prevents membrane fusions during desiccation and freeze-drying, and, most importantly, facilitates the absorption and retention of moisture (9, 11–14). The formation of colonial morphotypes embedded in dense EPS has thus emerged as an important acclimation strategy for extreme dryness. There may, however, also be drawbacks from the high density of the EPS matrix. Dense EPS may prevent cells from obtaining sufficient light, nutrients, and dissolved inorganic carbon, thus confining growth (5). The mechanical stress that is induced by the wet expansion and desiccation-induced shrinkage of the EPS matrix may pose a real challenge to the cyanobacteria (15–17). Nevertheless, in natural terrestrial cyanobacterial colonies, the swelling and shrinkage do not lead to any structural damage to cells in structural and ultrastructural observations (18, 19). It is unknown how terrestrial cyanobacteria coordinate the structural dynamics of the dense EPS matrix in response to periodic desiccation in xeric environments.

The regulation of cell walls in plants may shed light on this issue in cyanobacteria. The plant cell wall is a complex network of carbohydrate polymers that is a key determinant of the plant response to environmental stresses (20). To coordinate cell growth and stress resistance, cell wall plasticity and rigidity are dynamically modulated by a number of extracellular proteins (20, 21). Additionally, the cell walls of resurrection plants possess higher flexibility than those of plants that are not desiccation tolerant (17). The cyanobacterial EPS matrix also accommodates many exoproteins, including antioxidant proteins (10, 14, 22–24). However, a particularly attractive protein is the acidic water stress protein, WspA, first reported in N. commune (25). WspA is induced after desiccation or UV irradiation of cells, accumulates in cells, and is secreted into the EPS matrix upon rehydration (12, 26). To date, unique sequences coding for WspA have also been found in three other species, N. flagelliforme, N. sphaeroides, and N. verrucosum (27, 28). Although it has been hypothesized that WspA may modulate the structure and function of the dense EPS matrix (12), its biochemical identity and biological roles associated with this modulation as well as relevant expression patterns remain largely unclear. The representative terrestrial cyanobacterium, N. flagelliforme, can withstand long-term desiccation, intensive solar irradiation, and repeated cycles of desiccation-rehydration processes (2). In the present study, we sought to elucidate the mechanism coordinating the dynamics of the dense EPS matrix under periodic desiccation by resolving the function and roles of WspA protein in N. flagelliforme. Since natural colonies of N. flagelliforme cannot be cultivated under laboratory conditions (1), samples from liquid cultures were also employed for interpreting this issue.

RESULTS

Expansion and shrinkage of N. flagelliforme and secretion of WspA.

In native habitats, N. flagelliforme grows on soil surfaces with little vegetation cover, appearing as a filamentous colony form (Fig. 1a). The rehydrated and air-dried filaments are shown in Fig. 1b. After full rehydration, the filament could expand in diameter by 2- to 4-fold. WspA proteins exist in various N. flagelliforme samples as well as in N. commune samples, which is shown in Fig. S1 in the supplemental material. In N. flagelliforme, as in N. commune (26), WspA was most abundant among the released proteins in the rehydration fluid (Fig. 1c) and showed continued release over time (Fig. 1d). This implies that WspA acts as an important exoprotein functioning in the EPS matrix.

FIG 1.

(a and b) Filamentous colonies of N. flagelliforme grown on soil surface (a) and air-dried and fully rehydrated filaments (b). (c) Patterns of proteins that separated by electrophoresis in 12% SDS-PAGE. Fluid-P, proteins released into the rehydration fluid, in which proteins were concentrated for use; TP, total soluble proteins extracted from the dry colonies; M, protein marker. An arrow indicates the WspA protein. Approximately 10 μg of total protein was loaded in each lane. (d) Western blotting of the released WspA in rehydration fluids at various times, from 5 min to 6 h. Equal amounts (0.1 g dry weight) of air-dried N. flagelliforme were rehydrated, and the released total proteins were concentrated for blotting.

Effects of moisture condition on wspA transcription and WspA dynamics.

The absence of moisture is the most important limiting factor in xeric regions. Moisture-related wspA expression and WspA synthesis were first investigated in natural N. flagelliforme (Fig. 2). Slightly induced WspA synthesis (Fig. 2a) and wspA transcription (Fig. 2c) were observed when the rewetted samples were subjected to air drying. No particular difference in WspA induction in the presence or absence of protease inhibitors suggested the relative stability of this protein under this moisture condition. However, when the air-dried samples were immersed in an aqueous solution, there was a continuous decrease in WspA in them (Fig. 2b), and the decrease in transcription levels of wspA was dependent on the duration of immersion (Fig. 2d). In addition, the presence of protease inhibitors appeared to slow the decrease of WspA in the wetted samples, thereby suggesting the existence of protein hydrolysis or instability. With an extended duration in solution, hydrolysis seemed to be dominant (e.g., after 6 h). Thus, the significant decrease in WspA on immersion was attributed to its release from cells, hydrolysis, and transcriptional inhibition.

FIG 2.

WspA dynamics and wspA transcription under dehydration and rehydration processes in natural N. flagelliforme. (a) Changes in WspA level in samples detected by Western blotting during dehydration with or without protease inhibitors. Equal amounts (0.5 g wet weight) of rehydrated samples were subjected to matric water stress. (b) Changes in WspA level in samples detected by Western blotting during rehydration with or without protease inhibitors. Equal amounts (0.1 g dry weight) of air-dried samples were rehydrated in solutions. In each of the four panels shown in panels a and b, approximately 10 μg of total protein was loaded in each lane. (c and d) Relative changes in wspA transcription detected by qRT-PCR during dehydration and rehydration. Data shown are means ± SD (n = 3). *, significant difference at a P of <0.05 (Student's t test).

WspA was not detected in the liquid culture of N. flagelliforme or in those of N. commune and N. sphaeroides, though it was present in environmental samples of these cyanobacteria (Fig. 3a). wspA transcription and WspA dynamics in natural and liquid-cultured N. flagelliforme samples were thus monitored to further evaluate the effects of abiotic stresses (Fig. 3b to f). Natural samples showed an increase in relative transcription after NaCl stress and UV-B irradiation and a similar response to mannitol stress, although this effect disappeared between 1 and 6 h (Fig. 3d). The synthesis of WspA in the sample was induced by UV-B, whereas upon NaCl stress, WspA was clearly reduced (Fig. 3b). We speculate that NaCl might severely disturb the stability of this protein in the EPS matrix and lead to an excessive loss from the sample, since the interaction between WspA and the UV-absorbing pigments (mycosporines and scytonemin) potentially leads to the formation of stable complexes in the matrix (12), and we also observed that salt solution could lead to the obvious shrinkage of the filaments. In liquid cultures, transcripts of wspA largely decreased under all three stresses (Fig. 3e), and WspA was not detected after stress induction (Fig. 3c). However, WspA was gradually induced in liquid cultures, clearly appearing on the day 8, when the liquid cultures were subjected to a moderate drying process (∼96% water content) in the perforated plates (Fig. 3f). Together, these results demonstrate a profound alteration in wspA transcription and WspA dynamics under different moisture conditions.

FIG 3.

Responses to WspA/wspA induction between natural colonies and liquid cultures. (a) No presence of WspA in liquid cultures was detected by Western blotting. Nf and LNf, natural and liquid-cultured N. flagelliforme, respectively; Nc and LNc, natural and liquid-cultured N. commune, respectively; Nsp and LNsp, natural and liquid-cultured N. sphaeroides, respectively. (b) Western blotting of WspA in natural N. flagelliforme subjected to NaCl, mannitol, and UV-B stresses. NaCl, 300 mM; mannitol, 400 mM; UV-B, 0.5 W/m². (c) Western blotting of WspA in liquid-cultured N. flagelliforme subjected to the three stresses. (d and e) Relative changes in wspA transcription detected by qRT-PCR in natural and liquid-cultured N. flagelliforme subjected to the three stresses. Data shown are the means ± SD (n = 3). *, significant difference at a P of <0.05 (Student's t test). (f) Western blotting of WspA in liquid-cultured N. flagelliforme under drying induction in perforated plates as shown on the left. Approximately 10 μg of total protein was loaded in each lane of panels a, b, c, and f.

Enzymatic identity of WspA and relevant features.

To explore the enzymatic identify of WspA, we adopted the Escherichia coli protein expression system to obtain protein for biochemical analysis. We chose the representative wspA1 gene (NCBI accession no. ABA54841) towards this aim. In the E. coli protein expression system, the induced full-length WspA1 protein always formed inclusion bodies despite many attempts. In this study, we separated WspA1 into three regions for in vitro recombinant expression (Fig. 4a). The truncated proteins Wsp_A and Wsp_B were expressed in soluble states and were purified for enzymatic analysis (Fig. 4b). Recombinant E. coli cells harboring Wsp_A were able to hydrolyze X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside), an artificial substrate of β-galactosylhydrolase (Fig. 4c). Purified Wsp_A was also observed to catalyze this reaction but only when the acidic reaction solution was adjusted to be neutral or weakly alkaline (Fig. 4d). Compared with Wsp_B, Wsp_A showed much stronger activity to hydrolyze ONPG (o-nitrophenyl-β-d-galactopyranoside), another artificial substrate (Fig. 4e). Other conditions affecting enzymatic activity included the optimum temperature and acidity of 45°C and pH 8.0, respectively, promoting ions Mg²⁺ and K⁺ and inhibitory ions Ca²⁺ and Zn²⁺ (Fig. S2). At the optimal temperature and acidity, the maximum rate of metabolism (V_max) and K_m value for Wsp_A with ONPG as the substrate were 12.3 μmol/liter · min and 0.5 mmol/liter, respectively, whereas the values for Wsp_B were only 1.9 μmol/liter · min and 1.6 mmol/liter, respectively. No β-glucosylhydrolase activity was detected for either protein. Therefore, WspA has β-galactosylhydrolase activity, and Wsp_A is the central activity region.

FIG 4.

In vitro expression and enzymatic analysis of WspA protein. (a) Truncation of WspA1 for in vitro expression. (b) Induction of Wsp_A and Wsp_B in the E. coli BL21/pET28 expression system and their purification. P, total proteins in the absence of IPTG induction; IP, total proteins in the presence of IPTG induction; W1 and W2, washing fractions; E, elution fractions; D, dialyzed proteins; M, marker protein. (c) Blue-white test of transgenic E. coli cells on separate LB plates supplemented with X-Gal as the substrate. (d) In vitro activity analysis of purified Wsp_A in X-Gal buffer (pH 7.5). (e) Comparative activity analysis in 0.1 M PBS buffer (pH 7.5) using ONPG as the substrate. Reactions were performed at 37°C. Data shown are means ± SD (n = 4). BSA, 50 μg/ml; Wsp_A, 10 μg/ml; Wsp_B, 27 μg/ml. Boiling conditions, 100°C for 15 min.

In general, β-galactosidase also possesses transgalactosylation activity, producing galacto-oligosaccharides (29, 30). This potential capability was evaluated in Wsp_A by using four major monosaccharide components of N. flagelliforme polysaccharides (glucose, galactose, xylose, and mannose) as acceptors (Fig. 5). Compared with the disaccharides, oligosaccharides of higher molecular weight formed in the presence of Wsp_A, thus suggesting the transfer of β-galactosyl residues to these substrates. Taken together, these results indicate that the unique WspA protein is a novel β-galactosidase with glycoside hydrolysis and transgalactosylation activities.

FIG 5.

TLC analysis of transgalactosylation activity of Wsp_A (E) using four monosaccharides as acceptors. Glu, glucose; Gal, galactose; Xyl, xylose; Man, mannose; Lac, lactose; Tre, trehalose. Lac and Tre serve as disaccharide controls. The red outline indicates the oligosaccharide products.

Role of WspA in the softening of EPS matrix under sufficient moisture.

Upon rehydration, the EPS matrix of N. flagelliforme should be softened to adapt to the expansion process. The EPS of N. flagelliforme contains 20 to 30% galactose and β-glycosidic bonds (31, 32). As with the β-galactosidase identified above, the potential hydrolyzing effect of WspA on the EPS matrix was evaluated by using Wsp_A to treat natural N. flagelliforme (Fig. 6). After a 24-h incubation in an aqueous solution, the treated filaments loosened, a little like loose flour, and their trichomes were easily separated or fractured after gentle crushing (Fig. 6b), whereas the trichomes of the untreated filaments maintained their arrangement and integrity (Fig. 6a). The softening by Wsp_A also led to the increased release of proteins into fluid in the medium (Fig. 6c) and could be indexed by the reduced stretching elasticity (Fig. 6d). This long-term treatment did not appear to impair the photosynthetic physiology of cells, in terms of the PSII efficiency parameter F_v/F_m (the ratio of variable to maximum chlorophyll fluorescence) (Fig. 6e). Additionally, no monosaccharides, disaccharides, or oligosaccharides were detected in the treated or untreated solutions (data not shown). These results indicate that WspA can exert a softening role on the EPS matrix under sufficient moisture.

FIG 6.

(a and b) Softening effects of WspA on the EPS matrix. Physiologically recovered samples were incubated in BG11 solutions (pH 7.5) lacking (a) or supplemented with (b) Wsp_A for 24 h and subjected to microscopy. A 40× objective lens was used. Blue bar, 0.2 mm; red bar, 10 μm. (c) Released proteins (fluid protein) in rehydration fluids. Data shown are means ± SD (n = 3). X = 8 μg/ml Wsp_A. FW, fresh weight. (d) Stretching elasticity change of filaments after treatment indicated in panel c (means ± SD; n = 20). **, significant difference at a P of <0.01 (Student's t test). (e) Photosynthetic physiological activity in terms of F_v/F_m of natural samples during Wsp_A treatment (means ± SD; n = 6).

Role of WspA in the thickening of EPS matrix upon drying.

The formation of a dense EPS matrix is critical for cyanobacterial acclimation to xeric environments (2, 11). Here, based on the transgalactosylation activity of WspA, we also used liquid-cultured N. flagelliforme to evaluate its potential role in the thickening process of the EPS matrix upon drying (Fig. 7). The samples were treated with Wsp_A and other control solutions after cultivation on perforated solid plates for 3 days, as described in Materials and Methods. The density of EPS could be indexed by the EPS/chlorophyll a (Chl a) ratio (16). All samples seemed to exhibit an inherent increase in EPS during the drying induction. However, the Wsp_A-treated samples showed a clearly higher increase in the EPS/Chl a ratio during the subsequent 10-day culture than did the other controls (Fig. 7a). Ca²⁺ is inhibitory for the enzymatic activity of WspA, and it exerted an obviously slowed role in the increased EPS/Chl a ratio in the later phase (day 10), compared with the Wsp_A-treated sample. Further, the thickening effects were compared between Wsp_A and Wsp_B in a second experiment (Fig. 7b). Although the enzymatic activity of Wsp_A was much higher than that of Wsp_B, the former showed only a slightly stronger effect than the latter (less than 30%) in the thickening of the EPS. The stability of Wsp_A was less than that of Wsp_B in their preparation and preservation. Thus, the balance between enzyme stability and activity may result in similar effects on the two proteins. Together, these results indicate that WspA might facilitate the thickening of the EPS matrix under a drying process.

FIG 7.

Promotion of WspA in the thickening of the EPS matrix as indicated by the relative change in the EPS/Chl a ratio. (a) Changes in the EPS/Chl a ratio relative to that on day 0 after liquid-cultured samples were subjected to drying in perforated plates. Buffer, 20 mM HEPES (pH 7.0); BSA, Wsp_A, and boiled Wsp_A, 16 μg/ml; Ca²⁺, 5 mM. (b) Comparative effects of Wsp_A and Wsp_B (16 μg/ml) on the EPS/Chl a ratio change under drying induction. Data shown are means ± SD (n = 3). *, significant difference at a P of <0.05 (Student’s t test) relative to each of the other controls.

DISCUSSION

To acclimate to xeric environments, cyanobacteria must use extensive strategies, including molecular, physiological, morphological, and even colonial mechanisms (11, 33–36). The importance of a dense EPS matrix in terrestrial colonial cyanobacteria has been recognized (11, 35, 37). The present study presents further recognition of the mechanisms coordinating the expansion-shrinkage process of the EPS matrix beyond its physicochemical properties. The composition and glycosyl linkage of EPS (17) and multilayered envelopes (15) may be involved in this regulation process in some species. Nevertheless, the elucidation of extracellularly predominant WspA proteins as a β-galactosidase yielded deeper insight into this mechanism. It was previously mentioned that a putative 36-kDa WspA purified from native proteins showed very weak β-galactosidase activity (38). In contrast, a significant breakthrough was the separation of WspA into several “domains” for in vitro expression, which clearly revealed its enzymatic identity. We found that the hydrophobic N terminus easily led to the formation of inclusion bodies and thus prevented the obtaining of soluble proteins. The typical biochemical results (hydrolytic or synthetic) of β-galactosidase are dependent on the acceptors of the β-galactosyl residue, H₂O or sugars (29, 39). The proper concentration of the latter in an aqueous solution is critical for synthesis. It was found that WspA can break down the EPS matrix under moisture-sufficient conditions and enhance the thickening of the matrix under moderate drying, both of which are well correlated with environmental desiccation-rehydration processes or moisture cycles. During these processes, other environmental factors (such as temperature, acidity, and metal ions) possibly exert a complex combined effect through the regulation of enzyme activity. In addition, the moisture cycle-associated transcription patterns of the wspA gene and WspA dynamics were observed by using natural and liquid-cultured N. flagelliforme materials and on the basis of previous research (12, 26). In particular, WspA was detected in the aquatic N. sphaeroides only when it was dried and was similarly induced in liquid culture-acclimated N. flagelliforme only after drying. These results suggest a universal importance for survival during the desiccation period in colonial Nostoc species. WspA is stored in cells, and a suitable level of secretion may guarantee its proper role in the EPS matrix. Otherwise, the excess cleavage of the EPS in natural colonies would lead to decreased resistance to abiotic stress (see Fig. S3 in the supplemental material). Therefore, WspA performs moisture cycle-associated structural regulation of the EPS matrix in cyanobacteria, facilitating their acclimation in xeric environments.

WspA was induced by desiccation and UV-B irradiation in natural N. flagelliforme; for the latter, the induction was still effective under substantial moisture conditions. The significant UV-B-induced synthesis may reflect a need for storage of WspA in cells as a sink and for timely secretion when rehydrated. Cyanobacterial heteropolysaccharides are composed of repeating units of monosaccharides as backbones (40). The intrinsic flexibility of polysaccharides is facilitated by pendant groups (e.g., nosturonic or uronic acids) and various modifications (e.g., methylation, sulfitation, and acetylation) (34, 40). In N. flagelliforme and N. commune, three predominant monosaccharides in the EPS are glucose, galactose, and xylose, at a molar ratio of approximately 2:1:1 (31). The EPS has been suggested to possess a 1,4-linked xylogalactoglucan backbone with d-ribose and nosturonic acid as peripheral groups (34). The main growing season for N. flagelliforme and N. commune is the relatively wet summer. Thus, a substantial softening effect may occur through WspA-facilitated breakage of EPS backbones after rewetting, thus alleviating mechanical tension and allowing for a greater space for cell growth. Long-term immersion in occasional heavy rain would lead to disintegration of the EPS in natural colonies, which might provide an opportunity in xeric regions for the dispersal of broken trichomes. In native habitats, complete air drying of colonies usually requires several hours (2). The precursor or repeating units of polysaccharide are formed intracellularly, and the subsequent polymerization is performed extracellularly (10, 40). The moderate drying of liquid culture-acclimated N. flagelliforme could lead to the achievement of some desiccation resistance (Fig. S4), which was also implied previously (41). No oligosaccharides or polysaccharides could be released into solution from the drying-induced liquid culture when it was rehydrated (data not shown). Therefore, WspA would facilitate the thickening of the EPS by integrating new precursors into the broken polysaccharide as well as the relinkage of the broken polysaccharide (via transglycosylation activity) during the air drying, thus reconstructing the rigidity to combat the mechanical pressure. A flexibility-rigidity coordination model underlying the roles of WspA in the dynamic regulation of the dense EPS matrix is thus proposed for terrestrial cyanobacteria, as illustrated in Fig. 8.

FIG 8.

Proposed model for the WspA-facilitated flexibility and rigidity coordination of the EPS matrix in terrestrial cyanobacteria. The synthesis of WspA is induced upon desiccation and UV irradiation of cells. WspA is partially secreted into the EPS matrix under rehydration, when WspA facilitates softening of the matrix for trichome growth by hydrolyzing the polysaccharide backbone. During drying, WspA facilitates the integration of new oligosaccharides into, or relinkage of, the broken EPS, thus reconstructing the rigidity of the EPS for desiccation resistance. Under prolonged wetting, the extreme softening of the EPS matrix may facilitate the dispersal of broken trichomes.

In plant cell walls, a xyloglucan-cellulose network forms the major tension-bearing structure (21). Extracellular xyloglucan endotransglucosylase/hydrolase (XTH) plays two crucial roles: (i) hydrolyzing the xyloglucan backbone to increase cell wall extensibility and (ii) cell wall strengthening via integration of newly secreted xyloglucans into the cell wall or rejoining of xyloglucan chains (20, 21). As suggested in this study, WspA plays similar roles in the cyanobacterial EPS backbone. Therefore, although cyanobacteria and plants are evolutionarily distant, they share a remarkable ability to facilitate the dynamic coordination of the extracellular protective envelope to cope with periodic desiccation. Although many exoproteins, including XTH, are involved in modulating the structural dynamics of plant cell walls (42, 43), WspA accounts for the majority of exoproteins in the EPS matrix of terrestrial cyanobacteria. The desiccation-rehydration process may take days or weeks in resurrection plants (44, 45), whereas both processes take only minutes or hours in terrestrial cyanobacteria (2, 6). Therefore, the WspA-facilitated roles in terrestrial cyanobacteria may be a specific but highly efficient mechanism for coping with matric water stress. Given the predominance of galactose in cyanobacterial heteropolysaccharides (40) and the wide sequence diversity of β-galactosidases (46), the universal significance of this regulation may be further revealed in future studies.

MATERIALS AND METHODS

Cyanobacterial strains.

N. flagelliforme (Berk. & Curtis) Bornet & Flahault is found in arid or semiarid steppes of the western and northwestern parts of China (2). The primary N. flagelliforme sample used in this study was collected in 2013 from Huhehaote City, Inner Mongolia. Air-dried samples were rehydrated in BG11 solution at 25°C under continuous illumination at 20 to 40 μmol photons m⁻² s⁻¹ for 16 to 24 h to fully recover physiological activity (47). Liquid-cultured N. flagelliforme was previously developed as described in reference 41. N. commune UTEX584 (liquid culture) was obtained from the Freshwater Algae Culture Collection of the Institute of Hydrobiology, Chinese Academy of Sciences. N. sphaeroides was sampled from mountain paddy fields in Hefeng County, Hubei Province, China, and its liquid culture was developed in BG11₀ solution (3).

Rehydration and dehydration treatments of natural N. flagelliforme.

For rehydration, air-dried samples were immersed in BG11 solution with or without the addition of broad-spectrum protease inhibitor mixtures (inhibiting serine proteases, esterases, cysteine proteases, metalloproteinases, and trypsin-like proteases) according to the specifications of the manufacturer (Wuhan Boster Biological Technology, Ltd., China). At various time points (0, 1, 3, 6, 15, 24, and 30 h), samples were collected for protein extraction, Western blotting, and transcriptional analysis. For dehydration, physiologically fully recovered samples were soaked in BG11 solution with or without protease inhibitors for 5 min and subjected to natural drying in a temperature-controlled chamber (25°C, ∼50% relative humidity). The time intervals for sampling were at least 1 h, and water loss (%) was calculated by comparing the weight at each sampling point with the initial weight (100%) of the rewetted sample. At each time interval, samples were ground in liquid nitrogen and subjected to protein extraction with buffer (50 mM Tris-HCl, 20 mM MgCl₂, 20 mM KCl, pH 7.0) at 4°C. Crude proteins were mixed with 5× protein loading buffer (0.3 M Tris-HCl, 0.05% bromine phenol blue, 50% glycerol, pH 6.8) and boiled for 10 min for concentration determination with the Bradford protein assay (48). Approximately 10 μg of total protein was loaded in each lane of a 12% SDS-PAGE gel for electrophoresis and subjected to Western blotting, using a primary rabbit antibody generated against purified Escherichia coli-expressed WspA protein (28). Blot densities of samples of each panel were compared based on visual assessment. Their changing trends were verified by at least two independent replicates. Transcriptional analysis was performed by quantitative reverse transcription-PCR (qRT-PCR) using SYBR Premix Ex Taq (TaKaRa, Dalian, China), as previously described (47). The PCR primers were as follows: RT-wspA-forward (5′-GTTAATAATGTAGATCAAGCC-3′) and RT-wspA-reverse (5′-TCAGTTCCTTGTGTTTGTGC-3′). Each primer was designed against the wspA sequence (NCBI accession no. KF632585) with two base changes, feasible for this environmental sample and the liquid-cultured sample. The transcript level was normalized to the constitutively expressed 16S rRNA gene (47), and the relative transcription levels were calculated.

Abiotic stresses on natural and liquid-cultured N. flagelliforme.

To examine the effects of abiotic stressors on the induction of WspA, fully rehydrated natural samples were immersed in BG11 solutions supplemented with 300 mM NaCl or 400 mM mannitol for 0, 1, 3, and 6 h. The rehydrated samples were also irradiated with 0.5 W/m² UV-B for 0 to 6 h. Liquid-cultured samples were treated similarly; however, the duration of the stress treatments was extended to 24 or 48 h. At various time points, samples were collected for protein extraction, Western blotting, and transcriptional analysis, as described above. In addition, a long-term WspA induction experiment was performed by spreading liquid-cultured samples on two layers of gauze on solid BG11 plates (6.8 cm in diameter) with perforated lids (12 holes of ∼2 mm in diameter) to allow for gradual drying in the chamber mentioned above. Crude proteins were extracted at 0, 8, 13, 18, and 23 days and subjected to Western blotting.

In vitro expression of WspA and purification.

The E. coli BL21(DE3)/pET28a protein expression system (Novagen, USA) was used to express the target protein. The truncated wspA sequences, wsp_B and wsp_A, were amplified by PCR from the pMD18-wspA1 plasmid we previously constructed (28). This full-length wspA1 gene (NCBI accession no. ABA54841) from N. commune was first reported by Wright et al. (12) but was also amplified by PCR in N. flagelliforme. A partial sequence of wspA1 was previously deposited into the NCBI database by us under accession no. KF632585. The primer pairs for wsp_B and wsp_A are, respectively, as follows: wspa36 (5′-catATGGCTCTTTACGGCTATAC-3′) and wspa-w3 (5′-TTATTCATTCACAATTGCAAAG-3′), and wspa128 (5′-catatgGTAGATCAGCCTTTTGCTCC-3′) and wspa-w3. The PCR products were cloned into the NdeI site of the pET28a vector for in vitro expression. The recombinant E. coli cells with an optical density at 600 nm (OD₆₀₀) of 0.6 to 0.8 after culture with shaking at 200 rpm at 37°C were subjected to induction with 0.2 mM IPTG (isopropyl-β-d-thiogalactopyranoside) for 6 h. The cells were broken by ultrasonication for protein extraction, and the crude proteins were purified with His-Bind resin (Novagen, USA). For protein purification, the basal buffer (20 mM Tris-HCl, 500 mM NaCl, pH 8.0) was supplemented with 5 mM (binding buffer), 60 mM (wash buffer), and 1,000 mM (elution buffer) imidazole. Purified Wsp_B and Wsp_A proteins were dialyzed in dialysis buffer (20 mM Tris-HCl, 150 mM NaCl, pH 8.0) prior to enzymatic activity analysis.

β-Galactosidase assay.

The recombinant E. coli cells bearing the pET28a::wsp_A vector were first subjected to a blue-white test after IPTG induction with 5-bromo-4-chloro-3-indolyl β-d-galactoside (X-Gal) as the substrate (49). The β-galactosylhydrolase activity was further assayed in 1 ml of 0.1 M phosphate-buffered saline (PBS; pH 7.5 to 8.0) solutions using 3 mM o-nitrophenyl-β-d-galactopyranoside (ONPG) as the substrate. The reactions were conducted at 37°C or 45°C for 0 to 2 h and stopped by supplementation with 100 μl of 1 M Na₂CO₃ solution. The absorbance of the reaction product, ortho-nitrophenol, was measured at 405 nm. The Michaelis-Menten kinetic parameters (V_max and K_m) (50) were determined at the optimum temperature (45°C) and acidity (pH 8.0). Owing to the significant activity difference, the reaction times for Wsp_A and Wsp_B were set at 1 and 9 h, respectively.

Transgalactosylation activity was examined by a thin-layer chromatography (TLC) assay, as described by Gupta et al. (51), with some modifications. Briefly, transgalactosylation reactions were performed at 45°C for 3 h by incubation of 20 μl of the enzyme (0.1 mg/ml), 20 μl of the acceptor (500 mM), and 60 μl of ONPG (50 mM) in 0.1 M PBS buffer (pH 8.0). The acceptors used were four monosaccharides: galactose, glucose, mannose, and xylose. Two disaccharides, lactose and trehalose, served as controls for the produced oligosaccharides. Products were separated by TLC using a GF254 silica gel plate (branch of Qingdao Haiyang Chemical Co., Ltd., China) with methanol-chloroform (40:60) as the mobile phase. After the chromatography, the plate was air dried at room temperature. A chromatogram was developed with 20% H₂SO₄ and heating at 115°C for 5 min (or until color appeared).

Softening and thickening experiments of the EPS matrix.

To examine the potential biological roles of WspA on the structural regulation of the EPS matrix, moisture-related softening and thickening experiments were conducted. For the softening experiment, air-dried natural N. flagelliforme was rehydrated for 12 h to recover physiological activity and then incubated in 5-ml BG11 solutions (buffered with 20 mM HEPES, pH 7.5) either supplemented with or lacking 20 μg/ml Wsp_A for 24 h. Three replicate experiments were carried out. Both control and Wsp_A-treated samples were prepared for microscopy. Prior to the observation, the slides were gently pressed, which proved to be useful in distinguishing the degree of EPS integrity. The arrangement and fracture of trichomes were observed under a stereomicroscope (Discovery V12; Zeiss Inc.) and then a fluorescence microscope (Axio Scope A1; Zeiss Inc.) with or without excitation of green fluorescent protein (GFP) fluorescence. Under the fluorescence excitation, trichomes present with red color due to the existence of chlorophyll, while the background color of EPS matrix is minimized. In a second experiment, three replicates of rewetted samples (1.0 g) were incubated in 5 ml HEPES-buffered BG11 solutions (pH 7.5) supplemented with various concentrations of Wsp_A (0, 8, 16, and 32 μg/ml), and after a 24-h treatment, solutions were collected and concentrated for total protein quantification. The proteins released into the fluids or solutions (fluid proteins) were determined by subtraction of the added Wsp_A. The flexibility of the N. flagelliforme filaments was tested by fixing both ends of the filaments (4 to 5 cm long) with tweezers and then stretching them until breakage. The relative stretching extent (fold) was calculated by comparing the maximum stretching length and the original length. The potential physiological damage on the samples during the incubation with Wsp_A was indicated by the sensitive PSII efficiency parameter F_v/F_m when samples were incubated for 12 or 24 h, as described previously (47).

Liquid-cultured N. flagelliforme with the loss of integration of the EPS matrix shows induced morphogenesis after periodic drying treatment (41). Thus, liquid culture was employed for the thickening experiment upon drying. The liquid cultures were centrifuged at 6,000 rpm (centrifuge model 5810R; Eppendorf, Germany) for 5 min, and the pellets were evenly spread on BG11 solid medium plates with perforated lids as mentioned above. After 3 days of natural drying in the chamber, samples in each plate were infiltrated with 1 ml of (i) 20 mM HEPES-buffered BG11 solution (pH 7.0) supplemented with 16 μg/ml Wsp_A, (ii) 16 μg/ml boiled Wsp_A, (iii) 16 μg/ml bovine serum albumin (BSA), (iv) 5 mM Ca²⁺, and (v) 16 μg/ml Wsp_A plus 5 mM Ca²⁺. The infiltration was performed again on day 5 after the first infiltration (day 0). Samples on days 0, 5, and 10 were collected for EPS and chlorophyll a (Chl a) content determination. Total EPS and Chl a contents were determined as previously described (31, 52). The EPS/Chl a ratio was used as an indicator of the density of the EPS matrix surrounding the cells (16). The fold changes in the EPS/Chl a ratio were calculated and normalized to the ratio of the sample before treatment (day 0) in this study. In a second experiment, the comparative effect of Wsp_A and Wsp_B in the EPS thickening was similarly determined.