Numerous evolutionary innovations were required to enable freshwater green algae to colonize terrestrial habitats and thereby initiate the evolution of land plants (embryophytes). These adaptations probably included changes in cell-wall composition and architecture that were to become essential for embryophyte development and radiation. However, it is not known to what extent the polymers that are characteristic of embryophyte cell walls, including pectins, hemicelluloses, glycoproteins and lignin, evolved in response to the demands of the terrestrial environment or whether they pre-existed in their algal ancestors. Here we show that members of the advanced charophycean green algae (CGA), including the Charales, Coleochaetales and Zygnematales, but not basal CGA (Klebsormidiales and Chlorokybales), have cell walls that are comparable in several respects to the primary walls of embryophytes. Moreover, we provide both chemical and immunocytochemical evidence that selected Coleochaete species have cell walls that contain small amounts of lignin or lignin-like polymers derived from radical coupling of hydroxycinnamyl alcohols. Thus, the ability to synthesize many of the components that characterize extant embryophyte walls evolved during divergence within CGA. Our study provides new insight into the evolutionary window during which the structurally complex walls of embryophytes originated, and the significance of the advanced CGA during these events.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The colonization of land by plants was a pivotal event in the history of life. It had profound effects on atmospheric chemistry and climate, and these changes facilitated the expansion of terrestrial ecosystems and the emergence of terrestrial metazoan groups (Kenrick and Crane, 1997; Waters, 2003; Niklas and Kutschera, 2010). The first land plants (embryophytes) are believed to have appeared during the mid-Paleozoic era, between 480 and 450 million years ago, when freshwater green algae first emerged from their aquatic habitats (Graham, 1993; Kenrick and Crane, 1997; Karol et al., 2001; Waters, 2003; McCourt et al., 2004). This transition required various adaptations, including the development of mechanical support for the plant body, the formation of organs to exploit newly available resources, and the ability to resist new biotic and abiotic stresses. Many of these adaptations probably involved, or required, changes in the composition and architecture of the polysaccharide- and/or phenylpropanoid-rich walls that surround the cells of all green plants (Niklas, 2004). Subsequent diverse modifications of these walls underpinned the formation of complex body plans and the development of woody and vascular tissues that have allowed flowering plants to adapt to and survive in diverse terrestrial environments (Carpita and Gibeaut, 1993; Graham et al., 2000; Cosgrove, 2005; Harris, 2005).
Much of our understanding of plant cell-wall composition and architecture has been obtained from studies of angiosperms and gymnosperms. These cell walls are dynamic and complex structures comprising networks of cellulose microfibrils tethered by cross-linking glycans (hemicelluloses) that include xyloglucans (XyGs), xylans, arabinoxylans, mannans and mixed-linkage (1→3) (1→4)-β-d-glucan (MLG) (Bacic et al., 1988; Fry, 2004; Scheller and Ulvskov, 2010). In the primary walls of many plant taxa (gymnosperms, dicots and non-graminaecous monocots), these structural features (cellulose and cross-linking glycans) are embedded in a matrix of structurally diverse pectic polysaccharides that include homogalacturonan (HG), rhamnogalacturonan I (RGI) and rhamnogalacturonan II (RGII), as well as small amounts of proteins and proteoglycans (Ridley et al., 2001). The more rigid secondary cell walls of woody and vascular tissues typically contain more hemicelluloses than pectin and are reinforced by the hydrophobic phenylpropanoid polymer lignin (Boerjan et al., 2003). Plants combine these components in differing amounts and modify their fine structures to build walls that are tailored to functional requirements imposed by cellular context, development or challenges from pathogens and the environment.
The comparatively few analyses of early-diverging embryophyte taxa suggest that the walls of pteridophytes, lycopodiophytes, bryophytes and angiosperms contain broadly similar components (Popper and Tuohy, 2010; Sørensen et al., 2010; Popper et al., 2011). However, in order to fully understand the early evolution of plant cell walls, it is necessary to investigate the walls of charophycean green algae (CGA), which are the closest living algal relatives of the embryophytes and are believed to have given rise to the land plant lineage (Graham, 1993; Karol et al., 2001; Lewis and McCourt, 2004; McCourt et al., 2004; Becker and Marin, 2009). Extant CGA consist of six orders (Charales, Coleochaetales, Zygnematales, Klebsormidiales, Chlorokybales and Mesostigmatales), which include several thousand species with a diversity of morphologies, ranging from the asymmetric unicellular Mesostigma viride to members of the Charales that have complex and erect multicellular bodies (Figure 1) (Lewis and McCourt, 2004). It is likely that the gradual transition to land was driven by evolutionary events within CGA that allowed some of these freshwater algae to occupy progressively drier environments (Graham, 1993; Becker and Marin, 2009).
Here we present a detailed characterization of the cell walls of 10 CGA species using a combination of immunocytochemical and biochemical approaches, together with a technique based on use of glycan microarrays of extracted polysaccharides probed using monoclonal antibodies with specificities for cell-wall components (Moller et al., 2007). Determination of glycosyl linkages by methylation analysis and use of glycan microarrays are complementary approaches when there is a lack of prior knowledge about cell-wall structure, because, as is the case for CGA, monosaccharide linkages cannot necessarily be assigned to specific polysaccharides by methylation analysis. In contrast, although glycan arrays are only semi-quantitative, the epitopes recognized by monoclonal antibodies typically consist of several sugar residues that are diagnostic of specific polymeric cell-wall components, and this information aids the interpretation of methylation analysis data. Moreover, during glycan array analysis, polysaccharides are released sequentially from cell walls using three solvents [1,2-diaminocyclohexane-N,N,N′,N′-tetraacetic acid (CDTA), NaOH and cadoxen] and the presence of epitopes in the various fractions provides some information about cell-wall architecture as well as composition per se.
Results and Discussion
The cell walls of advanced CGA species and embryophytes have several polysaccharides families in common
Glycan array data are presented in Figure 2, and glycosyl linkage and cellulose content data are presented in Table S1 and Figure S1, respectively. Immunofluorescence and immunogold labelling are shown in Figures S2 and S3, respectively. Species tested included both basal and advanced lineages from five of the six CGA orders (Figure 1). In addition to cellulose, we detected HG, mannan, xylan/arabinoxylan, MLG and arabinogalactan proteins (AGPs) (Figures 2,3 and Figures S1–S4, and Table S1). We also obtained evidence for the presence of XyG epitopes in some members of the Charales, Coleochaetales and Zygnematales (Figures 2,3 and Figures S1,S2,S4 and Table S1).
The inability to detect isoprimeverose (α-d-Xylp-(1→6)-Glcp) (a disaccharide that is diagnostic of land plant XyGs) after driselase treatment of Chara corallina cell walls led to the suggestion that XyG is present exclusively in embryophyte cell walls (Popper and Fry, 2003). However, methylation analysis of Spirogyra cell walls revealed the presence of 4,6-Glcp, 1,4-Glcp and terminal xylose, which could be components of XyG as this hemicellulose has a backbone composed of 1,4-Glcp substituted at O6 with xylose (Ikegaya et al., 2008). Glycan array analysis indicated that the XyG epitope recognized by the monoclonal antibody LM15 was present in Netrium digitus, Chara corallina, Coleochaete nitellarum and Cosmarium turpini and to a lesser extent Spirogyra sp., with the highest levels being present in N. digitus (Figure 2). We also detected 4,6-Glcp in all the advanced CGA species, including those in which the XyG epitope recognized by LM15 was not detected (Table S1). It is unlikely that the 4,6-Glcp originated from starch as the walls were treated with α-amylase prior to analyses of glycosyl-linkage composition. It is known that the side chains of XyG can vary considerably across taxa (Peña et al., 2008; Hsieh and Harris, 2009; Hsieh et al., 2009), and that binding of the monoclonal antibody LM15 to XyG is inhibited by the presence of galactose in XyG side chains (Marcus et al., 2008). Thus, it is possible that XyG is present in several CGA species but not is recognized by LM15 because of particular side chain structures. In addition, the possibility cannot be discounted that differences in cell-wall architecture limit the solubilization of XyG from some CGA walls and thereby prevent its detection in glycan arrays. Extraction of polysaccharides is required for glycan array detection but not for glycosyl-linkage composition analysis. No fucosylated XyG epitope recognized by the monoclonal antibody CCRC-M1 (Puhlmann et al., 1994) was detected in any CGA species. The fact that varying amounts of terminal fucosyl residues were detected in all the later-diverging CGA (Table S1) suggests that terminal fucose is associated with a polymer other than XyG. These findings are consistent with the results of a previous study indicating that fucosylated XyG appeared after the divergence of mosses and hornworts (Peña et al., 2008). The likely presence of a form of XyG in some CGA is supported by our identification of a putative xyloglucan transglucosylase/hydrolase (XTH) gene in C. nitellarum (Table S2), recent reports of XTH activities and their corresponding genes in other CGA species (Van Sandt et al., 2007), and the discovery that several CGA contain endotransglucosylase activities that graft MLG onto XyG oligosaccharides in vitro (Fry et al., 2008a). In addition, several XyG-related genes, including putative XyG glucan synthases, have been identified in Chara globularis (Del Bem and Vincentz, 2010). These findings do not amount to conclusive proof of the presence of XyG in CGA, and indeed such proof may be very hard to acquire using techniques adapted for embryophyte XyG, but the presence of linkages, orthologues of XyG processing and biosynthetic genes and activities and XyG epitopes collectively suggests that a form of XyG evolved before the transition to land.
Until recently, MLG was believed to be present only in the cell walls of Poales species (Popper, 2008). However, the demonstration that MLG occurs in the cell walls of horsetails (Fry et al., 2008b; Sørensen et al., 2008) and Micrasteriasdenticulata (Eder et al., 2008) indicates that mechanisms for MLG biosynthesis evolved at least once before the emergence of the Poales. Glycan array analysis revealed the presence of MLG in C. turpini, M. furcata and P. trabecula (Figure 2). These results were confirmed by the release of glucan oligomers following treatment of the algal walls with lichenase, a MLG-specific endoglucanase (Figure S4). This diagnostic procedure also provides information about the proportions of cellotetraose and cellotriose subunits that comprise the polymer backbone. The MLG in M. furcata and P. trabecula comprises more cellotriose than cellotetraose subunits, as do the MLGs of most Poales species. By contrast, C. turpini MLG comprises more cellotetraose than cellotriose subunits (Figure S4), as does MLG in Equisetum arvense (Sørensen et al., 2008). Block structure is known to profoundly affect the rheological properties of MLGs, and therefore probably also their functional properties (Johansson et al., 2008). In E. arvense, MLG co-exists with high levels of HG (Sørensen et al., 2008), and this also appears to be true for C. turpini but not M. furcata or P. trabecula. Glycan array analysis showed that levels of HG epitopes in C. turpini were higher than in M. furcata and P. trabecula (Figure 2), as was the level of 1,4-GalAp. Although the functional significance of these observations is unknown, it is possible that the block structure of MLGs in both CGA and embryophytes is correlated with their associations with other cell-wall components, and may reflect distinct roles within cell-wall architectures.
Immunofluorescence and immunogold labelling of CGAs using the same monoclonal antibodies as for the glycan arrays revealed that, while all the epitopes detected were wall-localized, they had diverse distribution patterns (Figure 3a–i and Figures S2,S3). For example, XyG in C. corallina, N. digitus and C. nitellarum and HG in C. turpini were present in the walls of all cell types (Figure 3a–d). In contrast, in M. furcata, P. trabecula and C. nitellarum, epitopes associated with HG, MLGs and AGPs were localized to specific structurally distinct regions in CGA, including pores, hairs or the isthmus between dividing cells (Figure 3e–i). In M. furcata, the HG epitope recognized by monoclonal antibody JIM5 was restricted to the older of the two semi-cells (Figure 3e). Very little is known about cell-wall architecture in CGA, or the extent to which some cell-wall components are masked by others, thus, as in embryophytes, an apparently highly restricted cellular location may in some cases be misleading. Nevertheless, it does seem likely that CGA utilize cell-wall components for highly specialized roles that in some cases are related to specific developmental phases or structures.
One notable difference between the cell walls of advanced CGA and embryophytes was in the occurrence of the structurally complex RGI and RGII domains of pectin. None of the sugars (2-O-methyl fucose, 2-O-methyl xylose, apiose or aceric acid) that are characteristic of RGII were detected in any of the CGA species tested (Table S3). Moreover, the glycosyl linkages and epitopes typically associated with the arabinosyl- and galactosyl-containing side chains of RGI were absent or present at low levels (Figure 2 and Table S1). Previous studies of the cell walls of N. translucens (Anderson and King, 1961a,b) and Chara australis (Anderson and King, 1961c) based on fractionation of wall components followed by monosaccharide composition analysis indicated that both species may contain pectins that include galactose arabinose xylose and rhamnose. Whether these data are truly indicative of the presence of RGI is hard to assess, as the presence of glycosyl linkages was not determined. Nevertheless, the low or trace amounts of 1,5-Araf, 1,4-Galp and 2,4-Rhap detected in some species (C. nitellarum, C. turpini, N. digitus, Klebsormidium flaccidum and Chlorokybus atmophyticus) suggest that certain CGA may possess the biosynthetic capacity to produce the rhamnogalacturonan backbone and some of the side chain structures associated with RGI, even though this ability does not appear to be extensively utilized in the context of cell-wall construction. One basal species, C. atmophyticus, contained high levels of 1,4-Galp (8.7 mol%), but the fact that this species only contained trace levels of 2,4-Rhap indicates that this 1,4-Galp is not a component of RGI.
Similarly, our data strongly suggest that RGII as found in embryophytes is essentially absent from the diverse CGA species that we tested. There are nevertheless intriguing clues that some aspects of RGII biosynthesis may have pre-embryophyte origins. Specifically, the two rare 2-keto sugars 2-keto-3-deoxyoctonate (KDO) and 3-deoxy-2-heptulosaric acid (DHA) that are present in RGII also occur in the elaborate scales of some prasinophyte algae (Becker et al., 1994, 1998), and DHA occurs in the basket scales of Mesostigma viride (Domozych et al., 1991, 1992). It is possible that the presence of KDO and DHA in these scales may indicate early evolution of an RGII-like core structure, which could have been further elaborated after land colonization. This is plausible because, in embryophytes, KDO and DHA are attached directly to GalAp in the RGII backbone at the 3-position. Although we did not test for KDO or DHA, it may be significant that we observed low levels of 3,4-linked GalA in C. nitellarum but only sub-detection limit or trace levels of this linkage in other species (Table S1). Further detailed work is required to test whether C. nitellarum and other CGA species do indeed contain polysaccharides that could have formed a rudimentary RGII core structure.
Biochemical evidence for of lignin-like material in the Coleochaetales
The development of lignified cell walls is considered to have been a crucial adaptation for plants to live on land, and lignification is deemed vital for support, water transport and other cellular functions (Weng and Chapple, 2010). Lignin was believed to be restricted to vascular plants, but has recently been reported to occur in the red alga Calliarthron cheilosporioides (Martone et al., 2009). There is evidence for lignin-like compounds in C. obicularis (Delwiche et al., 1989) and Nitella flexilis (Ligrone et al., 2008), but these two studies, although informative, used techniques that were not truly diagnostic for lignin. The question of whether or not CGA contain lignin therefore remains unresolved, although it is generally believed that green algae do not synthesize this polymer (Sarkar et al., 2009). We used thioacidolysis to analyse cell-wall material from C. nitellarum, C. obicularis and C. scutata (Figure 4a and Table S4). The thioacidolysis method is considered diagnostic for lignins, because cleavage of β-aryl ether linkages (resulting from radical coupling reactions of monolignols) produces the characteristic arylglycerol trithioethyl ether monomers (Rolando et al., 1992). This analysis showed that (threo- and erythro-isomers of) guaiacyl and syringyl monomers, in various proportions, could be released from all three Coleochaete species, suggesting that, if not lignins, at least oligomers/polymers produced in part by radical coupling of monolignols (the process that typifies lignification) were present in these samples.
The total levels of the released monomers were very low, up to only 25 μg/g of cell wall compared to typically approximately 85 mg/g from softwoods (Figure 4a and Table S4). Suberins are considered to derive in part from hydroxycinnamyl alcohol components and also appear to release the thioacidolysis monomers (Lapierre et al., 1996) that are diagnostic of β-ether units resulting from radical coupling reactions involving the hydroxycinnamyl alcohols, if not from lignin per se. In the absence of information on genes involved in lignin biosynthesis in CGA and red algae, it is not possible to establish whether lignin has arisen in phylogenetically distant taxa by convergent or conserved evolution, but the ability of advanced CGA to produce lignin-like components is likely to have been important in the gradual transition of plants to land. It is possible that the low levels of lignin-like compounds that we detected could be the result of contamination, possibly from the algal culture medium. However, immunocytochemical analyses using well-established anti-lignin antibodies produced clear labelling of cell walls in all three species; representative labelling of C. scutata is shown in Figure 4b–g). Together, these data suggest that a form of phenylpropanoid metabolism emerged prior to colonization of land.
Aspects of the cellulose network assembly may be conserved in advanced CGA and embryophytes
Glycan array analysis provided evidence that some features of wall assembly are conserved in advanced CGA and embryophytes. For example, in species in which HG epitopes were present, they were extracted mostly by CDTA, implying that, as in embryophytes, HG exists in these CGA cell walls as a calcium cross-linked complex (Figure 2). Similarly, most of the hemicelluloses were solubilized by alkali (Figure 2), again a feature of most embryophyte walls. It is known that dichlorobenzonitrile, an inhibitor of cellulose synthesis in embryophytes, also disrupts this process in P. margaritaceum (Domozych et al., 2009a,b), and the cellulose synthase inhibitor isoxaben is similarly effective (Figure 3j). Both these compounds caused P. margaritaceum cell walls to bulge under turgor pressure, and it is reasonable to assume that this is due to weakening of the cellulosic network that is primarily responsible for resisting turgor-generated tension. We were interested to explore these observations further and test for a possible connection between cortical microtubule assembly and the production and crystallinity of cellulose microfibrils, which are known to be functionally linked in embryophytes (Sugimoto et al., 2003; Endler and Persson, 2011; Fujita et al., 2011; Himmelspach et al., 2011). As shown in Figure 3(k), treatment with oryzalin, which disrupts microtubule assembly, also caused bulging, which is again indicative of cell-wall weakening. Treatment of P. margaritaceum cells with cellulase did not cause bulging (Figure 3l), but when cells were treated with cellulase and then exposed to oryzalin, the result was rapid cell bursting (Figure 3n). One explanation for these observations is that disruption of existing cellulose can be withstood if an intact cortical microtubule network is present to participate in the replenishment or organization of cellulose microfibrils, but if both the capacity to produce cellulose and microtubules is affected, cell-wall integrity is very severely compromised. These findings suggest that, as in Arabidopsis, the structural quality of cellulose microfibrils in P. margaritaceum is linked to cortical microtubules.
The cell walls of Klebsormidium flaccidum and Chlorokybus atmophyticus are markedly different from those of the advanced CGA tested and those of embryophytes
The cell walls of the two early-diverging CGA species studied here, K. flaccidum and C. atmophyticus (Figure 1i–k) lacked most of the epitopes and glycosidic linkages that are typical of embryophyte walls (Figure 2 and Table S1). The walls of both species were relatively poor in cellulose (Figure S1) but contained callose (Figure 2, Table S1 and Figure S2). This 1,3-linked β-d-glucan is typically present in specific embryophyte cell walls at certain developmental stages, but is not a common structural component of most walls. Moreover, the walls of both species contained high levels of glycosyl linkages that are not typical of embryophyte wall polysaccharides, including terminal rhamnose, 1,3-Xylp, terminal arabinose, 2,5-Araf, terminal glucose, 3,6-Glcp and 3,4-GlcAp (Table S1). Thus, although the cell walls of these two basal algal species contain complex polysaccharides, they are not polysaccharides that are common to the walls of embryophytes or advanced CGA species.
Certain cell-wall biosynthetic genes are conserved in C. nitellarum and embryophytes
The presence of a similar set of cell-wall polymers in advanced CGA and embryophytes could have arisen by deeply conserved evolution, convergent evolution, or most likely a combination of these processes. We obtained genetic evidence that at least some aspects of wall biosynthesis are deeply conserved by sampling the transcriptome of the advanced CGA species C. nitellarum. This Coleochaete species is a member of one of the two groups of CGA that have been proposed as immediate ancestors of land plants (Graham, 1984, 1993; McCourt et al., 2004). We identified orthologues of several genes involved in cell-wall processing and biosynthesis in angiosperms (Table S2), including an XTH (CAZy family GH16 glycosyl hydrolase), which is known to contribute to XyG metabolism in land plants, and members of the CSLD sub-family of CSLs (CAZy family GT2 glycosyl transferases), which have been associated with hemicellulose biosynthesis in some cells (Scheller and Ulvskov, 2010). Other contigs aligned with members of the CAZy family of glycosyl transferase (GT)8 GAUT1-related family, a member of which has been shown to play a role in pectin biosynthesis (Table S2) (Caffall and Mohnen, 2009). This result is consistent with our data suggesting relatively high levels of pectin in C. nitellarum cell walls (Figure 2). Together with previous work on the cellulose synthase (CESA) gene family (Roberts and Roberts, 2007), our data provide evidence that the biosynthetic mechanisms for at least some members of all three major cell-wall polymer classes, cellulose, hemicelluloses and pectins, are conserved in certain CGA and embryophytes.
Nevertheless, it also appears highly likely that some polymers that are common to CGA and embryophyte cell walls have arisen by convergent evolution from various genetic backgrounds. For example, in the Poales, the biosynthesis of MLG is thought to be mediated by cellulose synthase-like (CSL) genes of the F and H classes (Burton and Fincher, 2009). The earliest diverging embryophytes so far sequenced, the lycophyte Selaginella moellendorffii and the bryophyte Physcomitrella patens, do not possess these CSL gene classes, and it therefore seems very unlikely that CGA do. Thus MLG is probably produced by an alterative mechanism in CGA. Similarly, it is likely that the ability to produce lignin arose at least twice, once in red algae (Martone et al., 2009) and once in embryophytes, and possibly a third time in CGA.
The transition from freshwater to land is a gentler ecological gradient than transition from a marine habitat to land, and freshwater adaptation probably allowed CGA to colonize moist habitats that were exposed to periods of drying, thereby facilitating full transition to terrestrial niches (Graham, 1993; Becker and Marin, 2009). It has been proposed that the ancestors of CGA were uniquely placed to make a successful transition to land because certain characteristics of their physiology and biochemistry afforded a degree of pre-adaptation to life on land (Graham, 1993; Becker and Marin, 2009). It now seems likely that the ability to produce cell walls with particular polymer compositions was an important aspect of pre-adaptation that underpinned the success of the pioneering land plants, and that this distinctive ability may be one reason why only the streptophyte algae gave rise to the embryophyte lineage.
This study significantly extends our understanding of the cell walls of CGA, and is unique in encompassing CGA taxa from much of their phylogenetic range. This coverage is important because our results provide insight into marked differences in cell-wall composition that are broadly correlated with evolutionary transitions that gave rise to advanced CGA and ultimately to the embryophytes. Nevertheless, it is still far from a complete survey of glycan biosynthetic capacity. We exclusively sampled vegetative thalli of selected species, and did not analyse the walls of other life stages, including zoospores, meiospores, zygospores (in the case of the Zygnematales) or zygotes (in the case of the Charales). Nor did we analyse the polysaccharide-rich extracellular polymeric substances that are produced by many CGA.
The fact that some cell-wall components appear to be present at very low levels, for example RGI, may suggest that certain synthetic capacities, although not widely utilized by CGA, were nonetheless important latent abilities that were subsequently elaborated and exploited in the embryophytes in response to the particular requirements of terrestrial life. Thus, some aspects of cell-wall production were expanded and some were probably lost – but it seems that very few were truly embryophyte innovations. These modifications ultimately gave rise to the diversity of cell-wall architectures that underpinned the development of complex and specialized tissues and organs that enabled embryophytes to exploit diverse habitats. Furthermore, it is likely that co-evolution of suites of enzymes that modify and restructure cell walls ensured that these walls are dynamic structures that can be altered in response to both biotic and abiotic stimuli.
Algal strains and growth conditions
C. atmophyticus (LB2591), K. flaccidum (LB321) and C. nitellarum (LB1261) were obtained from the University of Texas Culture Collection (Austin, TX). P. margaritaceum, P. trabecula, M. furcata, C. turpini, N. digitus, Spirogyra sp. and C. corallina were obtained from the Skidmore College Algal Culture Collection. All isolates except C. corallina grown in the presence of a commercial anti-biotic/anti-mycotic cocktail (Sigma, http://www.sigmaaldrich.com/) according to the manufacturer’s instructions. All cultures except C. corallina were maintained in sterile Woods Hole medium enriched with soil extract (5%) (Carolina Biological, http://www.carolina.com/). Cultures were kept in a Bionette growth chamber (Fisher Scientific, http://www.fishersci.com/) at 18 ± 1°C, with a photoperiod of 14 h light (35 W m−2 of cool white fluorescent light) and 10 h dark. Cells were sub-cultured every 2 weeks. Cells from the log phase (3–6 weeks after sub-culturing) were harvested for this study. C. corallina was grown at room temperature in water supplemented with Woods Hole Medium for 2–3 weeks in a large aquarium. Young thalli were collected and washed thoroughly with deionized water.
Alcohol-insoluble residue (AIR)
Freeze dried algal cells or thalli were suspended in liquid nitrogen and homogenized using a tissue homogenizer (Qiagen MM 200, http://www.qiagen.com/). Five volumes of aqueous 70% ethanol were added, and the suspension samples were kept for 1 h at 4°C while agitating. The suspensions were then centrifuged at 4000 rpm for 10 min, and the supernatant was discarded. This procedure was repeated five times, before a final wash with acetone for 2 min. The alcohol-insoluble residue (AIR) was then air-dried.
Glycosyl residue and glycosyl linkage compositions and cellulose content of the AIR
Portions (approximately 1 mg) of the AIR and red wine RGII (1 mg) were separately treated for 1.5 h at 120°C with 2 m TFA (500 μl). Residual TFA was removed under a flow of air, and the residue was washed using isopropanol (50 μl). The released sugars were converted to their corresponding alditols by treatment for 2 h at room temperature with 1 m ammonium hydroxide (200 μl) containing sodium borodeuteride (10 mg ml−1). The alditols were than per-O-acetylated and analysed by gas chromatography and electron ionization mass spectrometry (GC-EIMS) as described previously (York et al., 1985). Suspensions of freeze-dried algal tissue were treated for 1 h at 40°C with porcine α-amylase (Sigma; 2 U/mg carbohydrate), and then for an additional 30 min at 40°C with 1 U/mg α-amylase. The insoluble residue was kept overnight in aqueous 96% ethanol, pelleted by centrifugation, and washed three times with aqueous 96% ethanol before a final wash with methanol. Carboxyl reduction (Sandford and Conrad, 1966) was performed prior to methylation to facilitate linkage analysis of uronic acids, samples were methylated using solid NaOH and methyl iodide (Ciucanu and Kerek, 1984) and glycosyl-linkage composition analysis was performed as described previously (Kim and Carpita, 1992; Sims and Bacic, 1995). The cellulose content of the AIR (10 mg) was determined as described previously (Updegraff, 1969) using Avicel crystalline cellulose (Sigma) and AIR derived from mature stems of Arabidopsis thaliana (ecotype Columbia-0) as standards.
Thioacidolysis was performed using a streamlined method (Robinson and Mansfield, 2009) developed from previous protocols (Lapierre et al., 1986; Rolando et al., 1992). Thioacidolysis monomers were analysed as their trimethylsilyl ethers by GC-MS and authenticated on the basis of their retention times and mass spectra using authentic standards. The monomers were quantified using response factors determined daughter ions for G (m/z 269G) and S (m/z 299S) relative to the internal standard 4,4′-ethylenebisphenol, and total ion current response factors of 1.11 (G) and 1.26 (S).
Glycan microarrays of cell-wall components
Glycan array analysis was performed as described previously (Moller et al., 2007). Supernatants of extracted cell-wall material were spotted onto nitrocellulose membrane in three replicates and three dilutions, and three independent analyses were carried out. Mean spot signals from the three experiments are presented as a heatmap (created using the online HeatMapper tool available at http://bar.utoronto.ca), with the values normalized to the highest value (set equal to 100). A cut-off of 5% of the highest mean signal value was imposed, and values below this are represented as 0. Table S5 give details of antibody specificities and references.
Inhibition of mircotubule assembly and cellulose synthesis in Penium margaritaceum
P. margaritaceum cultures were treated for 24 h with the microtubule assembly inhibitor oryzalin (0.28 μm), or for 24 h with the cellulose synthesis inhibitors dichlorobenzonitrile (1 μm) or isoxaben (0.5 μm), or for 12 h with cellulase at 1 mg ml−1. Cultures remained viable for up to 1 week under these conditions, and the effects of the inhibitors or enzymes were reversible.
Lichenase digestion and estimation of (1→3) (1→4)-β-d-glucan (MLG) content
Suspensions of AIR (5 mg) in 20 mm sodium phosphate, pH 6.5 (0.5 ml), were kept for 2 h at 50°C with continuous mixing. The suspension was centrifuged at 10 000 g for 5 min, and the supernatant was used as a negative control. The pellet was resuspended in sodium phosphate, pH 6.5 (0.5 ml), and 2 units of lichenase (Megazyme) were added. The mixture was kept for 2 h at 50°C with continuous mixing, after which the supernatant was collected for HPAEC-PAD (pulsed amperometric detector) analysis of the released oligosaccharides using a Dionex (http://www.dionex.com/) BioLC ICS 300 system equipped with a pulsed amperometric detector. Samples and controls (10 μl) were separated on a CarboPac PA1 column (Dionex) equilibrated with 50 mm sodium acetate in 0.2 m sodium hydroxide, and eluted at 1 ml min−1 using a linear gradient of sodium acetate from 50 to 350 mm in 0.2 m NaOH over 15 min. Glucose and malto-oligosaccharides with a degree of polymerization between 2 and 7 (Sigma), and a lichenase digest of barley flour MLG were used as standards. Oligosaccharides were identified based on their retention times compared with the standards.
Support was provided by the New York State Foundation for Science, Technology and Innovation, and National Science Foundation Plant Genome grant DBI-0606595 to J.K.C.R., National Science Foundation grant MCB-0919925 to D.D., US Department of Energy grant DE-FG02-96ER20220 to M.A.O., Department of Energy Office of Science grant DE-AI02-06ER64299 and Department of Energy Great Lakes Bioenergy Research Center grant DE-FC02-07ER64494 to J.R. F.A.P. and A.B. acknowledge funding from the Australian Research Council through grant LP0989478 and the Australian Research Council Centre of Excellence in Plant Cell Walls grant. I.S. acknowledges funding from the Villum Kann Rasmussen foundation. We thank Dr A. Matas (Department of Biology, Cornell University, Ithaca, NY) for assistance with transcriptome data generation and analysis.