www.nature.com/scientificreports OPEN Insoluble (1 → 3), (1 → 4)-β-D- glucan is a component of cell walls in brown algae (Phaeophyceae) and Received: 28 October 2016 is masked by alginates in tissues Accepted: 24 April 2017 Published: xx xx xxxx 1 2,3 4 4 2,3 Armando A. Salmeán , Delphine Duffieux , Jesper Harholt , Fen Qin , Gurvan Michel , 2,3 1,5 2,3 Mirjam Czjzek , William G. T. Willats & Cécile Hervé Brown algae are photosynthetic multicellular marine organisms. They belong to the phylum of Stramenopiles, which are not closely related to land plants and green algae. Brown algae share common evolutionary features with other photosynthetic and multicellular organisms, including a carbohydrate-rich cell-wall. Brown algal cell walls are composed predominantly of the polyanionic polysaccharides alginates and fucose-containing sulfated polysaccharides. These polymers are prevalent over neutral and crystalline components, which are believed to be mostly, if not exclusively, cellulose. In an attempt to better understand brown algal cell walls, we performed an extensive glycan array analysis of a wide range of brown algal species. Here we provide the first demonstration that mixed-linkage (1 → 3), (1 → 4)-β-d-glucan (MLG) is common in brown algal cell walls. Ultra-Performance Liquid Chromatography analyses indicate that MLG in brown algae solely consists of trisaccharide units of contiguous (1 → 4)-β-linked glucose residues joined by (1 → 3)-β-linkages. This regular conformation may allow long stretches of the molecule to align and to form well-structured microfibrils. At the tissue level, immunofluorescence studies indicate that MLG epitopes in brown algae are unmasked by a pre- treatment with alginate lyases to remove alginates. These findings are further discussed in terms of the origin and evolution of MLG in the Stramenopile lineage. Brown algae are a large and diverse class of marine and photoautotrophic organisms, consisting of more than 250 genera and 1500–2000 species . They form the monophyletic class of Phaeophyceae which is included in the Stramenopile lineage and that forms one of the eight major phylogenetic groups of eukaryotes . Thus, brown algae evolved complex multicellularity independently from animals, fungi, red algae, green algae and 2, 3 green plants . The Stramenopiles also comprise the pseudofungi oomycetes, the photosynthetic unicellular diatoms (Bacillariophyceae) and various unicellular or filamentous algae, including the Eustigmatophyceae and 4, 5 Xantophyceae which are more closely related to brown algae . Brown algae have an important ecological impact as they are the largest biomass producers in coastal regions. Their environment is very different to that of land plants, being submerged in salt water and exposed to a changing environment of current, waves and tides, to which they adjust quickly to comply with bending or torsion forces . As in other multicellular photosynthetic organisms, brown algae evolved a carbohydrate-rich cell-wall sur- rounding their cells. However, due to the aforementioned phylogenetic distances with other lineages, their cell wall composition and characteristic are distinct . The two main polysaccharides are gel-forming and hydroscopic polymers, namely the alginates and the fucose-containing sulfated polysaccharides (FCSPs), the latest including sulfated fucans . The cellulose microb fi rils are believed to be cross-linked by the FCSPs and to have no connection with alginates , which in turn are associated with phenolic compounds to form a network in which the elements Department of Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, 1871, Frederiksberg, Denmark. Sorbonne Universités, UPMC Univ Paris 06, UMR 8227, Integrative Biology of Marine Models, Station Biologique de Roscoff, CS 90074, Roscoff, France. CNRS, UMR 8227, Integrative Biology of Marine Models, Station Biologique de Roscoff, CS 90074, Roscoff, France. Carlsberg Research Laboratory, J.C. Jacobsens Gade 4, 1799, København V, Denmark. Present address: William G.T. Willats, Newcastle University, Newcastle upon Tyne, United Kingdom. Correspondence and requests for materials should be addressed to W.G.T.W. (email: william. email@example.com) or C.H. (email: firstname.lastname@example.org) Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 1 www.nature.com/scientificreports/ cited above are embedded. Additional cell wall components are proteins, arabinogalactan proteins (AGPs) and 8, 9 halide compounds . Cell wall rigidity is likely controlled by the tuning of alginate fine structure and polyphenol 8, 10 cross-linking, while sulfated polysaccharides may play a key role in the adaptation to osmotic stress . In brown algal cell walls the polyanionic polysaccharides are prevalent over neutral and crystalline poly- saccharides. The later only represent a minor fraction of the wall (~0–10%) and are believed to be composed mostly, if not exclusively, of cellulose . More recent studies point out that some brown algae are able to build up (1 → 3)-β-glucan chains in their wall, transiently upon infection in some Ectocarpales and Laminariales , or more ubiquitously in Fucus vesiculosus cell walls . To the best of our knowledge no additional neutral and/or water insoluble polysaccharide has been reported so far in brown algal cell walls. The presence of an unbranched and unsubstituted glucan featuring a mixture of (1 → 3)-β- and (1 → 4)-β-d-linked glucose residues, also known as a mixed linkage glucan (MLG), has been suspected but never clearly determined . The presence of MLG has been however established in the stramenopile xantophyte alga Monodus subterraneus . MLG has a peculiar distribution pattern among eukaryotes, being mostly found in the plant kingdom and 16 17 18, 19 in fungi and with additional occurrences reported in red algae and bacteria . MLG in the green lineage is mostly distributed in the members of the Poaceae, but also in some less commonly found early diverging vascular 20–23 plants and freshwater green algae . The fact that MLG occurs in some of these evolutionarily very distant taxa but not in all, congruent with some genetic evidences, is in favour of a hypothesis based on multiple origins of the polymer, being ‘reinvented’ several times even at the level of the plant kingdom . In plants the ratio between the two linkages seems to be species specific and to influence the physicochemical properties of the polysaccharide and consequently its functional properties in cell walls . In an attempt to better understand brown algal cell walls, particularly in term of their glucan composition, we performed an extensive glycan array analysis of a wide range of brown algal species, included within the most important phaeophycean orders. This led to the discovery that MLG epitopes are ubiquitous to all the phyloge - netically diverse brown algal cell walls tested in this study. We describe here the highly regular structure of MLG in brown algae, and its association with alginates in tissues. We further discuss the origin and evolution of MLG in this eukaryotic lineage. Results Discovery of MLG in cell walls of brown algae. Cell walls from a range of brown algal species of dis- tinct taxonomical groups were sequentially extracted using a dedicated protocol. Three solvents (CaCl , Na CO , 2 2 3 NaOH) were chosen because they are known to preferentially solubilise FCSPs, alginates and water insoluble 10, 13 polysaccharides (such as cellulose), respectively . The presence of MLG in these fractions was challenged by antibody-based glycan array analyses. e Th monoclonal antibody LM7 is directed to pectic homogalacturonan but also binds to alginates , and was used as a positive control. As expected, alginates were abundantly detected in the Na CO fractions but not in the NaOH fractions. Unexpectedly, MLG was readily detectable in all the brown 2 3 algal species under investigation (Fig. 1). The MLG occurrence was restricted to the NaOH-fractions, which con- tain components held firmly in cell walls. In some brown algal species, further analysis of cell wall extracts taken from blade, stipe or holdfast regions indicated that the detected amounts of MLG were similar between tissues (Supplementary Information Fig. S1). In order to rule out any possibility that the antibodies were cross-reacting with other polymers than MLG in this context, and to validate the presence of MLG in brown algal cell walls, we used a specific enzymatic-based epitope deletion assay (Fig. 2). MLG can be hydrolysed specifically by a lichenase, an MLG-specific endohy- drolase whose target site is known to be a (1 → 4)-β-d-glucopyranosyl linkage immediately adjacent to a (1 → 3)-β-d-glucopyranosyl residue . It does not hydrolyse pure (1 → 3)-β-d-glucans such as laminarin (e.g. the main storage polysaccharide in brown algae) nor (1 → 4)-β-d-glucans (cellulose). A dedicated array printed with the NaOH fractions of several brown algal samples was prepared and submitted to lichenase digestion before probing. As a control a parallel analysis was performed on Brachypodium distachyon (Poaceae) which is known to have the highest content of MLG among grasses and on Arabidopsis thaliana (Brassicaceae) which does not contain MLG. As expected MLG was abundantly detected in the B. distachyon extract but was not detected in A. thaliana. MLG was present in the brown algal extracts but detected at lower levels as compared to B. distachyon. In all samples the MLG signal was abolished by the lichenase treatment. An additional control used a laminarinase that preferentially cleaves soluble (1 → 3)-β-d-glucans such as laminarin . MLG-epitopes in brown algae were unae ff cted by this treatment. A reduction of the MLG signal in B . distachyon is likely attributed to the lichenase side-activity of the laminarinase , on a prevailing and more readily accessible MLG substrate in these cell wall extracts as compared to those from brown algae. Altogether these results indicate that the anti-MLG antibody is not cross-reacting with any putative remaining laminarin in this context and that the polysaccharide detected in the brown algal extracts is a (1 → 3), (1 → 4)-β-d-glucan sensitive to a lichenase treatment. MLG in brown algae has a highly regular structure. By cleaving solely the (1 → 4)-β-D-glucopyranosyl linkages immediately adjacent to (1 → 3)-β-D-glucopyranosyl residues, the digestion of MLG by the lichenase yields diagnostic oligosaccharides. For instance, the major oligosaccharides released by lichenase digestion of poalean MLG are the trisaccharide G4G3G (DP3) followed by the tetrasaccharide G4G4G3G (DP4) , where ‘G’ is β-d-glucopyranose, and ‘3’ and ‘4’ indicate (1→ 3) and (1 → 4) linkages. To better characterise the MLG structure in brown algae, AIRs were digested with the lichenase. The oligosaccharides released from the lichenase digestion were fluorescently labelled and analysed with an Ultra-Performance Liquid Chromatography (UPLC) coupled to a fluorescence detector (Fig. 3). No overlap between starch and cellulose derived oligosaccharides DP3 and DP4 (maltotriose, maltotetraose, cellotriose, cellotetraose) was observed (Supplementary Information Fig. S2). Five algal samples corresponding to three different orders (Ectocarpales, Fucales and Laminariales) were assayed. In addition to the B. distachyon and A. thaliana extracts, a commercial laminarin sample was included as Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 2 www.nature.com/scientificreports/ Figure 1. Detection of MLG epitopes in brown algal extracts by glycan array analyses. The tree shows the schematic phylogenetic relationships of the 34 species of brown algae used in the study: they belong to 6 main orders of brown algae from the early diverging clade Dictyotales to the more recently diversified clades of Laminariales, Fucales and Ectocarpales. Cell walls were sequentially extracted and probed with the anti-mixed linkage glucan (MLG) and anti-alginates (LM7) antibodies. The heatmap represents the binding profile obtained from the glycan array analyses. The colour scale in relation to absorbance values is shown. For each antibody the highest mean signal value in the entire data set was set to 100 and all other signals adjusted accordingly. Values <5 were considered as background and discarded. Stars by algal names refer to the corresponding pictures shown on the side. a negative control. Consistent with previous analyses, the most abundant hexose oligomers in B. distachyon MLG were the DP3 (G4G3G) followed by the DP4 (G4G4G3G) . A. thaliana was negative as it does not possess MLG. Some peaks were observed from the laminarin sample but none overlapping with the DP3 and DP4 from MLG. As expected, the lichenase has no activity towards laminarin. This experiment rules out that the eluted oligosac- charides are derived from a lichenase activity and indicates that they were initially present in the laminarin sam- ple. By contrast, all the five algal samples under investigation oer a simi ff lar pattern of elution to the one observed for B. distachyon with however some quantitative distinctions. As anticipated from the glycan array analyses, the Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 3 www.nature.com/scientificreports/ Figure 2. Deletion of MLG epitopes from glycan arrays by enzymatic treatment. Glycan arrays were treated with buffer only or with either a lichenase (10 µg/ml) or a laminarinase (7.4 µg/ml) for 5 min before probing with the anti-mixed linkage glucan (MLG) and anti-alginates (LM7) antibodies. Brachypodium dystachion is included as a positive control, while Arabidopsis thaliana is included as negative control. G4G3G (DP3) Ectocarpus sp. Sargassum muticum Laminaria digitata Laminaria ochroleuca Laminaria hyperborea Arabidopsis thaliana laminarin G4G4G3G (DP4) Brachypodium distachyon 81 10 12 14 618 Retention time (min) Figure 3. UPLC analysis of lichenase oligomers. Ultra-Performance Liquid Chromatography (UPLC) analysis of the oligosaccharide products released by the lichenase digestion of cell wall extracts obtained from 6 brown algae belonging to 3 distinct orders (Ectocarpales, Fucales, Laminariales). Brachypodium dystachion is included as a positive control, while Arabidopsis thaliana and a commercial laminarin sample are included as negative controls. The chromatogram shows the relative abundance of the tri- (G4G3G) and tetra-saccharides (G4G4G3G). Note that the algal MLG has a very regular structure with trisaccharides only, in contrast to B. dystachion which has both tri- and tetrasaccharide units. amounts detected were very low compared to the positive control. More importantly, only the DP3 was detected in reasonable quantities, indicating that the MLG in brown algae has a very regular block structure consisting of cellotriose units linked by single (1 → 3)-β-d-linkages (G4G3G). We noted the presence of some faint signal at the MLG derived DP4 level in some of the samples, and we cannot rule out that some irregularities in the linkage pattern might be present in some individuals, species or life-cycle stages. Immunolocalisation of MLG in brown algal tissues. Based on our understanding that MLG is an insol- uble polymer firmly wall-bound in brown algae we hypothesised that it might contribute to the stiffness of some of the tissues. We explored its location in the stipes of Laminariale species, a stem-like rigid structure that con- nects the flexible body blade to the anchoring holdfast. Cut surfaces of stipes were tissue-printed and the nitro- cellulose sheets were immunolabelled to detect cell-wall epitopes. The positive control LM7 reacted strongly with the tissue prints which indicates the abundant and consistent distribution of alginates within the stipes (Fig. 4). Much of the spatial information regarding distribution is maintained during the printing process and indicated that the LM7 labelling was stronger in the radial rows separating the distinct tissue layers, and in the medulla of L. hyperborea and L. digitata. The presence of MLG within the stipes was also supported by a positive immuno- labelling of the prints. MLG showed a more restricted occurrence as compared to alginates, being detected in all Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 4 Emmision Unit (EU) www.nature.com/scientificreports/ TISSUEPRINTS CROSS alginates negative MLG SECTION (LM7) control co co md Laminaria md hyperborea 4 mm Saccharina latissima 4 mm 2 cm Laminaria digitata 4 mm Figure 4. Detection of MLG in stipes from Laminariales by tissue printing. The tissue prints of stipes from Laminariales were probed by the anti-mixed linkage glucan (MLG) and anti-alginates (LM7) antibodies. The samples being naturally pigmented, negative controls are included to appreciate the signal due to the antibody binding. Corresponding cross sections of the stipes are shown on the side. A representative picture of a whole plant of Saccharina latissima is indicated for size indication. The LM7 epitopes are abundant in all tissues. MLG is of a more restricted occurrence and shows distinct spatial localization depending on the alga. It is mostly detected in the medulla for Laminaria hyperborea, while it is more largely distributed in the meristoderm and the cortex in S. latissima. m, meristoderm; co, cortex; md, medulla. The dot lines broadly indicate the different regions in tissues as observed by light microscopy. meristoderms, but showing distinct spatial localisations depending on species for other tissues. It was mostly detected in the medulla for L. hyperborea, while it was more largely distributed in the cortex for S. latissima and L. digitata. It is interesting to mention that tissue printing usually allows the transfer of soluble material, indicating that the insoluble MLG may have been dragged along with other polysaccharides which it could be associated to, and alginates would be a possible candidate. To study in more details the MLG distribution in these tissues, we cross-sectioned stipes from these Laminariale species and localised MLG by immunouo fl rescence microscopy (Fig.5, Supplementary Information Fig. S3). Our first attempts to localise MLG in these tissues were disappointing as almost no signal was obtained, or only at a very low level for L. hyperborea (Fig. 5a). Knowing that in plants pectic homogalacturonans can 27, 28 mask the detection of hemicellulose polysaccharides , and that MLG is probably associated to alginates in our samples, we performed an unmasking experiment based on the use of recombinant alginate lyases from 29, 30 marine bacteria . The treatment was effective in removing most alginates as shown by the loss of LM7 binding (Fig. 5b). Aer deg ft radation of alginates, MLG was clearly labelled in all cell walls and particularly in the cortex of S. latissima where it appears to be more easily labeled (Fig. 5b). The signal was abolished by an additional lichenase treatment, indicating that the antibody was not cross-reacting with other polymers in this context but truly detecting MLG. Observations at higher magnifications indicated that MLG had a restricted distribution to the most inner and outer part of the cell wall (Fig. 5c). Discussion Cell walls of brown algae have been extensively studied in term of composition, notably regarding their gel-forming polysaccharides alginates and sulfated fucans/FCSPs . Cellulose is the only crystalline component which has been reported in the walls from brown algae so far and it only occurs at low levels . In eukaryotes MLG has been largely studied in the green lineage where it has been adopted by different plant and algal species during 20–23 16, 31, 32 the course of evolution , and in fungi where it is known as lichenan . The discovery of MLG in brown algae was therefore unexpected. Our survey revealed that MLG epitopes are not only present but ubiquitous to all brown algal cell walls tested in this study. Most studies on brown algal cell walls have been focusing on the eco- nomically relevant polymers and the water insoluble residues have therefore been only poorly described. It could be argued that glucose residues measured within these fractions were assumed to derive from cellulose. Some rare and old studies already depicted the presence of a mixture of (1→ 3)-β- and (1 → 4)-β-linked glucose residues 33, 34 in alkali-soluble materials of Fucus vesiculosus but the chemical structure of the originated polymer was not established. More recent studies point out that some brown algae are able to build up (1→ 3)-β-d-glucan chains in their wall, transiently upon infection in some Ectocarpales and Laminariales , or more ubiquitously in Fucus vesiculosus cell walls . The presence of the two linkage types on the same glucan chain was already suspected but not conclusively determined . This is the first report which uses a combination of chemical and enzymatic Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 5 www.nature.com/scientificreports/ Figure 5. Unmasking MLG in stipes from Laminariales by immunouo fl rescence imaging. (a) Overview of MLG detection in cross-sections of stipes from Laminaria hyperborea and Saccharina latissima. The diagram illustrates the distinct tissues under study. The micrographs show the indirect immunouo fl rescence detection of MLG epitopes for some of the tissues. The Calcouo fl r White (calc.) stains all cell walls in sections. MLG is not or only weakly detected. The strongest binding is obtained in the cortical cell walls from Laminaria hyperborea. (b) Indirect immunouo fl rescence detection in cortical cell walls from Saccharina latissima . MLG epitopes are strongly detected in all cell walls aer a p ft re-treatment with alginate lyases. Equivalent sections are labelled with LM7 and indicates the loss of alginate detection aer t ft he enzymatic treatment. Equivalent sections further treated with a lichenase indicate a decrease in MLG detection. (c) Same sections observed at higher magnification indicate that the MLG epitopes are strongly detected aer t ft he enzymatic treatment in the most inner and outer parts of the cell walls, both in cortical and medulla cells. m, meristoderm; co, cortex; md, medulla. All scale bars = 50 μm. analyses and that demonstrates the presence of the two adjacent linkage types as a form of an insoluble MLG in brown algal cell walls. Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 6 www.nature.com/scientificreports/ MLG of various eukaryotic and bacterial origins have been structurally resolved. The use of a specific (1 → 3), (1 → 4)-β-d-glucan endohydrolase, known as a lichenase, has been particularly important in this matter. The lichenase hydrolyses a (1 → 4)-β-d-glucopyranosyl linkage immediately adjacent to a (1 → 3)-β-d-glucopyranosyl residue. Digestion of MLG either yields a trisaccharide (G4G3G) followed by a tetrasaccharide (G4G4G3G) in 16, 25 32 cereal grains , an asymmetry which is even more pronounced in fungi . The reverse is observed in the early 21, 23, 35 diverging land plants Equisetum spp. with a predominance of tetrasaccharides over di- and tri-saccharides , and a combination of these two profiles can be observed in green algal species . Additional structural variations include the consecutive linkage alternation in the acetylated MLG from the bacteria Sinorhizobium meliloti and the occurrence of a sulfated MLG in the red alga Kappaphycus alvarezii which contained only a minor (~8% mol) fraction of (1 → 3)-β-linkages . Brown algal MLG is rather unique in structure in containing almost exclusively the trisaccharide (G4G3G) as a repeating unit. This regular conformation may allow long stretches of the mol- ecule to align and therefore to form well-structured micofibrils insoluble in water: an analysis that t fi s with the observation of MLG being only extracted by high alkali in our brown algal samples. MLG in brown algae is a component of a cell wall architecture that drastically differs from what is known in other lineages such as land plants and fungi. Instead of co-exisiting with high levels of either glucuronoarabinox- ylans (e.g. Poales), pectins and cellulose (e.g. Equisetum spp.) or (1 → 3), (1 → 6)-β-d-glucans (e.g. Ascomycota), MLG occurs in brown algae with high levels of alginates and sulfated fucans/FCSPs, together with low levels of cellulose. In all brown algae, MLG was left in the residue aer s ft equential extraction with CaCl and Na CO , being 2 2 3 only extracted with high alkali, suggesting that it is firmly wall-bond and co-extracted with cellulose microb fi rils. Our findings also seem to indicate that MLG may be associated with another cell wall polymer. In cell wall immu- nochemistry studies, we hypothesized that it might be alginates, since the binding of the anti-MLG antibody to untreated transverse sections of stipes of Laminariales indicated a weak recognition of highly restricted locations and binding was shown to be increased considerably by the enzymatic removal of alginates. The detection of both MLG and alginates in the tissue prints also indicated that MLG might have been dragged along with other polysaccharides. On the other hand, however, the glycan array analysis indicates a contradictory result, as no MLG was detected in the fractions based on the use of Na CO , which is effective in releasing most (but not all) 2 3 alginates. All together, the presented results here do not yet allow any conclusion on the potential crosslinking of MLG structures with other polysaccharides. The conformational regularity of the (1 → 3), (1 → 4)-β-linkages will define the MLG tendencies to self-aggregate . This points to a putative major structural role for MLG in brown algal cell walls. This regular glu- can block structure also reflects a precise biosynthetic mechanism and/or an important functional requirement in brown algal cell walls. MLG in cereals differ in the peculiar distribution of the linkages along the chain, which has a profound influence on the rheological properties of the polysaccharide, thus forming a gel-like matrix in cell walls . In the grasses, MLG accumulates transiently in young and growing tissues therefore it has traditionally been associated with the regulation of cell wall expansion . In Equisetum spp. it is believed to serve a different purpose and has been suggested to be involved in silica deposition . Brown algal cell walls are not known to contain silica and our findings show that MLG has no distinction of occurrence between the tissue-types but is consistently present at low levels. These results suggest that MLG may perform a distinct role in brown algae compared to land plant organisms, probably in strengthening the cell wall. The non-conserved structural conformation of MLG between lineages, as its uneven distribution pattern among eukaryotes, is indicative of an independent evolution of MLG synthesis in multiple species. Candidate genes encoding cellulose synthase-like (Csl) proteins from the glycosyltransferase GT2 family, have been iden- tified that might participate in MLG synthesis in land plants. The synthesis of MLG in the Poaceae is effective by the evolution of a single clade of genes, namely CslF or CslH genes . Recent progress has been made in unrav- elling the mechanism of MLG synthesis in plants, with the identification of specific amino acid changes in CslF6 38, 39 isoforms that can influence the proportions of the different repeating units . How a single protein can generate 16, 25 the different block structures with two linkage-types is yet still a matter of debate . None of these genes has been found in any other MLG-producing plants and green algae and it has therefore been postulated that other genes are responsible for producing MLG in these organisms . Also consistent with this, no CslF or CslH homo- 40, 41 logues were found in the genome sequence of the brown algal model Ectocarpus siliculosus , and the nine GT2 genes identified in E . siliculosus form two distinct GT2 clades . Interestingly some of these genes are more closely related to BcsA genes from bacteria than Csl genes from plants , while the bacterial BscsA gene has recently been shown to be involved in MLG synthesis in Sinorhizobium meliloti . The E. siliculosus GT2 genes which are more closely related to the BcsA genes would therefore be serious candidates for the production of MLG in brown algae. Major changes in cell wall composition oen acco ft mpanied key evolutionary events, such as territorialisation 7, 43 of charophytes and their descendants, namely land plants . In the stramenopile brown algae the recruitment of new cell wall materials such as polyphenols and alginates, is believed to have shaped the building of complex mul- 8, 41 ticellularity . The recruitment of insoluble and tightly assembled MLG-chains would have helped in providing structural rigidity for the development of semi up-right forms. However, to fully understand cell wall evolution in the lineage, a comprehensive investigation of cell wall across all stramenopile groups would be needed. So far our understanding of cell wall structures is limited to the major groups (i.e. brown algae, oomycetes), with only a fragmentary view for other groups that evolved in-between. Alginates are known to be produced by a restricted number of stramenopile taxas basal to brown algae (i.e. Schizocladiophyceae, Xantophyceae) but in a cell wall architecture primarily made of cellulose in the case of the xantophyte Monodus subterraneus . Interestingly, MLG has also been shown to be present in this latter organism, with 15% of (1 → 3)-β-linkages, but it is not known if the block structure resembles the brown algal repeat-unit . In the earlier group of the Eustigmatophyceae, MLG production has not been reported but some GT2 genes in Nannochloropsis gaditana are related to those of the brown alga E. siliculosus and to the bacterial BscsA genes . Further studies including less-well studied species would be therefore valuable to address whether or not the biosynthetic mechanisms of MLG biosynthesis are Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 7 www.nature.com/scientificreports/ deeply conserved within the stramenopiles, or rather involve non-related sets of glycosyl transferases. These find- ings would also help in defining the main cell wall modifications that arose during the course of evolution in the Stramenopile lineage. Methods Collection of algal samples. All samples of algae, except the Ectocarpus siliculosus sample, were col- lected from natural environments. These samples were collected from Roscoff (France, GPS coordinates: 48.72, −3.98), unless otherwise stated: Dictyopteris membranacea, Dictyota dichotoma (Dictyotales); Colpomenia per- egrina, Pylaiella littoralis, Ectocarpus sp. (Ectocarpales); Fucus vesiculosus from Aarhus (Denmark, 56.16, 0.20), F. spiralis from Aarhus, F. serratus from Aarhus, F. distichus from Aarhus, Ascophyllum nodosum from Grenaa (Denmark, 56.42, 10.88), Himanthalia elongata from Plougerneau (France, 48.62, −4.56), Pelvetia canaliculata from Plougerneau, Durvillaea antartica from Chiloé-Dalcahue (Chile, 42.60, −73.95), Durvillaea potatorum from Girvan (UK, 55.24, −4.86), Sargassum wightii from Chennai (India, 13.00, 80.32), S. tenerrimum from Chennai, S. muticum, S. longifolium from Chennai, Cystoseira nodicaulis, Bifurcaria bifurcata, (Fucales); Stypocaulon sco- parium (Sphacelariales); Saccorhiza polyschides from Galway (Ireland, 53.00, −9.55) (Tilopteridales); Macrocystis pyrifera from Chiloé-Dalcahue, Laminaria ochroleuca from Plougerneau, L. nigripes from Godthåb/Nuuk (Greenland, 64.18, 51.72), L. hyperborea from Plougerneau, L. digitata from Plougerneau, Saccharina japonica from Chiba (Japan, 36.61, 140.12), S. latissima from Godthåb/Nuuk, Lessonia trabeculata from Chiloé-Dalcahue, Le. Nigrescens from Chiloé-Dalcahue, Eisenia bicyclis from Chiba, Ecklonia maxima from Cape Town (South Africa, −34.31, 18.46), Undaria pinnatifida , Agarum clathratum from Godthåb/Nuuk, Alaria esculenta from Godthåb/Nuuk (Laminariales). Macroalgae were cleaned of epiphytes and washed in tap water. No permissions were required for these locations/activities. The collection of algal samples did not involve endangered or pro- tected species. Ectocarpus siliculosus (Ec32 strain), was cultured in Station Biologique de Roscoff (France) as described previously . Plant material. Brachypodium distachyon and Arabidopsis thaliana were cultivated in greenhouse conditions at the University of Copenhagen. Seedlings of Brachypodium distachyon (L.) P. Beauv. line 21 (Bd21) were grown at average temperatures of 22/16 °C (day/night) at ambient light levels. The mature plants were kindly provided by Vanja Tanackovic Fangel. Seedlings of Arabidopsis thaliana Columbia line (Lehle Seeds, USA) were grown in hydroponics at average temperatures of 24/17 °C (day/night) at ambient light levels. Rosettes and roots were col- lected from 12-week old plants. The collected plant materials were freeze-dried before extractions. Cell wall extractions and microarray profiling. The cell-wall extracts were prepared as previously described . In short, after preparation of the insoluble alcohol-residues (AIRs), the cell-wall polymers were sequentially extracted using a dedicated protocol for brown algae based on the subsequent use of 2% CaCl, 3% Na CO3 and 4 M NaOH. Two intermediate treatments with 0.2 M HCl and 50 mM CDTA respectively, were used to ensure a correct enrichment of fractions (Supplementary Information Fig. S1). Samples were centrifuged between each extraction (1500 g, 15 min) and the supernatants collected. Samples of Brachypodium dystachion and Arabidopsis thaliana were treated following the same protocol and used as a positive and a negative control, respectively. Extracts were then printed on nitrocellulose membranes using a microarrayer robot (piezoelec- tric Sprint Arrayjet, Roslin, UK) in four dilutions and four technical replicates. The microarrays were probed and developed as described previously using the following antibodies: anti-MLG from Biosupplies (BS-400-3) and LM7 from Plant Probes (anti-homogalacturonan cross-reacting and detecting alginates ). These probed arrays were scanned and signals were quantified using the Array Pro-Analyzer 6.3 software (Media Cybernetics, Rockville, MD, USA) to be ultimately converted into heatmaps, in which mean spot signals of the 16-spot sub-arrays are correlated to colour intensity. For each algal sample the data are based on three repeats of the whole process. In some cases the microarrays were treated with enzymes prior to antibody labelling (epitope deletion). Degradation of soluble (1 → 3)-β-d-glucans was performed using the laminarinase LamA from Zobellia galactanivorans at a concentration of 7.4 µ g/ml. Degradation of MLG was performed using a (1 → 3), (1 → 4)-β-d-glucanase known as a lichenase from Bacillus sp. (Megazyme) at a concentration of 10 µ g/mL. The two enzymes were applied independently to the microarrays in 100mM T ris pH6.5, 200 mM NaCl for 5 min at room temperature. Arrays not treated with the enzymes were incubated for an equivalent time with the same buffer. All arrays were subsequently treated for 45 min with 10 µ g/ml proteinase K (Sigma-Aldrich) in PBS before probing to remove any enzyme still attached to the cell wall extracts. The arrays were washed again with PBS before probing. Chromatographic separation. Six brown algae including one Ectocarpale (Ectocarpus sp.), one Fucale (Sargassum muticum) and three Laminariales (L. digitata, L. ochroleuca, L. hyperborea) were harvested in Rosco, ff cleaned of epiphytes and dried. Aer p ft reparation of the AIRs the cell wall components were extracted with 4 M NaOH. For each algal species, cell wall were extracted from a bulked mixture of two to four individuals to account for natural biological variations. The supernatants were neutralized with acetic acid until pH 6.5 and a (1 → 3), (1 → 4)-β-d-glucanase treatment was applied for 2 hours using the lichenase from Bacillus sp. (Megazyme) at a concentration of 3.3 U/mg of sample in 10 mM phosphate buffer pH6.5. The resulting oligomers were fluo- rescently labelled with 2-aminobenzamide and analysed with a Waters Acquity Ultra-Performance Liquid Chromatography (UPLC) System equipped with an Acquity glycan column and a fluorescence FLR detector (excitation wavelength of 350 nm and emission wavelength of 420 nm), using an Acquity UPLC BEH glycan 1.7 μm, 2.1 × 150 mm column with a VanGuard BEH glycan 1.7 μm, 2.1 × 5 mm pre-column at room temper- ature. Samples of Brachypodium dystachion and Arabidopsis thaliana were treated following the same protocol Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 8 www.nature.com/scientificreports/ and with the same enzyme and used as one positive and one negative control, respectively. The Brachypodium sample is presented compressing the scale by eight times for the purpose of visualization. An additional negative control using ethanol-washed commercial laminarin (Sigma); that naturally contains shorter, purely β-1,3-linked gluco-oligosaccharides) was treated with the same lichenase for 2 hours at a concentration of 3.3 U/mg of sample in 10 mM phosphate buffer pH6.5. For each sample and for each algal species, at least four technical replicates of the chromatograms were produced. In order to eliminate possible co-elution between the MLG trisaccha- ride, cellotriose or maltotriose, appropriate controls consisting of MLG (Glucagel) (DKSH, Italy), maltotriose, maltotetraose, cellotriose, cellotetraose (all Sigma, USA) were treated as described above. No co-elution could be observed between the MLG trisaccharide and any of the standards (Supplementary Information Fig. S2). Tissue printing. Stipes of Laminariales (S. latissima, L. hyperborea, L. digitata) were cross-sectioned and pressed firmly onto a nitrocellulose sheet. Biological replicates of the sections for each stipe were produced from two seasonally distinct individuals (collected in April and October). The most representative section was then chosen to be displayed. The prints were allowed to dry for 1 h and then blocked with 3% (w/v) milk powder in PBS (PBS/MP) for 1 h. The antibodies targeting MLG (BS-400-3, Biosupplies) and alginates (LM7, Plant Probes) were added at 10-fold and 1000-fold dilutions respectively, in PBS/MP and incubated for 1 h. A er wa ft shing, the secondary antibodies linked to alkaline phosphatase (AP) were added at 1000-fold and 5000-fold dilutions in PBS/MP (e.g. anti-mouse for anti-MLG and anti-rat for LM7, respectively). Ae ft r a further 1 h the nitrocellulose sheets were washed in water and the binding detected by addition of AP substrate (450µ M of 5-Bromo-4-chloro- 3-indolyl phosphate disodium, 400µ M Nitrotetrazolium Blue chloride). Negative controls were treated in the same manner as described, except that no MLG or alginate specific antibodies were added at any point of the experiment. The reaction was stopped by washing extensively in water. Immunomicroscopy. Stipes of Laminariales (S. latissima, L. hyperborea, L. digitata) were sectioned at 60 µ m thickness using a Leitz 1320 freezing microtome (Ernst Leitz Wetzlar GmbH, Wetzlar, Germany) and immedi- ately fixed overnight in 4% paraformaldehyde in 50% diluted seawater. For the immunolabelling procedure, the samples were washed twice in PBS and incubated in PBS/MP for 1h. Sa mples were then washed in PBS and the antibodies targeting MLG (BS-400-3, Biosupplies) and alginates (LM7, Plant Probes) were added at 10-fold and 1000-fold dilutions, respectively, in PBS/MP and incubated for 1h . Following washing with PBS, the secondary antibodies linked to fluorescein isothiocyanate (FITC) were added at 100-fold dilutions in PBS/MP in dark- ness (e.g. anti-mouse for anti-MLG and anti-rat for LM7; Sigma). Aer 1 ft h, the samples were washed with PBS, incubated with Calcouo fl r White (Sigma) for 5 min in darkness and mounted aer ft washing in a glycerol-based anti-fade solution (Citifluor AF31; Agar Scientific). All slides were observed on an Olympus BX60 microscope equipped with epifluorescence irradiation. Images were captured with an Exi Aqua camera (Qimaging) and the Volocity software (Perkin Elmer). In some cases the cell walls of sectioned algal materials were treated with enzymes prior to antibody labelling. e micr Th oscopic slides were coated with 1% poly-L-lysine (Sigma) and the washed sections were allowed to dry on the slides. Removal of alginates was performed using a mixture of 5 µ g/mL of the alginate lyase AlyA1 from 30 29 Zobellia galactinovorans and 5 µ g/mL of an alginate lyase from Pseudomonas alginovora , in 25 mM Tris pH7.5, 200 mM NaCl for 1 h at room temperature. Degradation of MLG was performed using the lichenase from Bacillus sp. (Megazyme) at a concentration of 10 µ g/mL in 100 mM Tris pH6.5, 200 mM NaCl for 1 h at room temperature. Sections not treated with the enzymes were incubated for an equivalent time with the corresponding buffers. References 1. Silberfeld, T. et al. A multi-locus time-calibrated phylogeny of the brown algae (Heterokonta, Ochrophyta, Phaeophyceae): Investigating the evolutionary nature of the “brown algal crown radiation”. Mol Phylogenet Evol 56, 659–674, doi:10.1016/j. ympev.2010.04.020 (2010). 2. Baldauf, S. L. The Deep Roots of Eukaryotes. Science 300, 1703–1706, doi:10.1126/science.1085544 (2003). 3. Reyes-Prieto, A., Weber, A. P. & Bhattacharya, D. The origin and establishment of the plastid in algae and plants. Annu Rev Genet 41, 147–168, doi:10.1146/annurev.genet.41.110306.130134 (2007). 4. Lin, Y.-C. et al. Distribution Patterns and Phylogeny of Marine Stramenopiles in the North Pacific Ocean. Applied and Environmental Microbiology 78, 3387–3399, doi:10.1128/AEM.06952-11 (2012). 5. Yamagishi, T., Müller, D. G. & Kawai, H. Comparative transcriptome analysis of Discosporangium mesarthrocarpum (Phaeophyceae), Schizocladia ischiensis (Schizocladiophyceae), and Phaeothamnion confervicola (Phaeothamniophyceae), with special reference to cell wall-related genes. Journal of Phycology 50, 543–551, doi:10.1111/jpy.12190 (2014). 6. Harder, D. L., Hurd, C. L. & Speck, T. Comparison of mechanical properties of four large, wave-exposed seaweeds. American Journal of Botany 93, 1426–1432, doi:10.3732/ajb.93.10.1426 (2006). 7. Popper, Z. A. et al. Evolution and diversity of plant cell walls: from algae to flowering plants. Annual Review of Plant Biology 62, 567–590, doi:10.1146/annurev-arplant-042110-103809 (2011). 8. Deniaud-Bouët, E. et al. Chemical and enzymatic fractionation of cell walls from Fucales: insights into the structure of the extracellular matrix of brown algae. Annals of Botany 114, 1203–1216, doi:10.1093/aob/mcu096 (2014). 9. Hervé, C. et al. Arabinogalactan proteins have deep roots in eukaryotes: identification of genes and epitopes in brown algae and their role in Fucus serratus embryo development. New Phytologist 209, 1428–1441, doi:10.1111/nph.13786 (2016). 10. Torode, T. A. et al. Monoclonal antibodies directed to fucoidan preparations from brown algae. PLoS One10 , e0118366, doi:10.1371/ journal.pone.0118366 (2015). 11. Kloareg, B. & Quatrano, R. S. Structure of the cell walls of marine algae and ecophysiological functions of the matrix polysaccharides. Oceanogr. Mar. Biol. Annu. Rev 26, 259–315 (1988). 12. Tsirigoti, A., Beakes, G., Hervé, C., Gachon, C. M. & Katsaros, C. Attachment, penetration and early host defense mechanisms during the infection of filamentous brown algae by Eurychasma dicksonii . Protoplasma 252, 845–856, doi:10.1007/s00709-014-0721- 1 (2014). 13. Raimundo, S. C. et al. Immunolocalization of cell wall carbohydrate epitopes in seaweeds: presence of land plant epitopes in Fucus vesiculosus L. (Phaeophyceae). Planta 243, 337–354, doi:10.1007/s00425-015-2412-3 (2015). 14. Popper, Z. A. & Tuohy, M. G. Beyond the green: understanding the evolutionary puzzle of plant and algal cell walls. Plant Physiol 153, 373–383, doi:10.1104/pp.110.158055 (2010). Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 9 www.nature.com/scientificreports/ 15. Ford, C. W. & Percival, E. Polysaccharides synthesised by Monodus subterraneus. Part II. e Th cell-wall glucan. Journal of the Chemical Society 3014–3016, doi:10.1039/jr9650003014 (1965). 16. Burton, R. A. & Fincher, G. B. (1,3;1,4)-beta-D-glucans in cell walls of the poaceae, lower plants, and fungi: a tale of two linkages. Mol Plant 2, 873–882, doi:10.1093/mp/ssp063 (2009). 17. Lechat, H., Amat, M., Mazoyer, J., Buléon, A. & Lahaye, M. Structure and distribution of glucomannan and sulfated glucan in the cell walls of the red alga Kappaphycus alvaezii (Gigartinales, Rhodophyta). Journal of Phycology 36, 891–902, doi:10.1046/j.1529-8817.2000.00056.x (2000). 18. Lee, J. & Hollingsworth, R. I. Oligosaccharide β-glucans with unusual linkages from Sarcina ventriculi. Carbohydrate Research 304, 133–141, doi:10.1016/S0008-6215(97)00204-8 (1997). 19. Pérez-Mendoza, D. et al. Novel mixed-linkage β-glucan activated by c-di-GMP in Sinorhizobium meliloti. Proceedings of the National Academy of Sciences 112, E757–E765, doi:10.1073/pnas.1421748112 (2015). 20. Eder, M., Tenhaken, R., Driouich, A. & Lütz-Meindl, U. Occurence and characterization of arabinogalactan-like proteins and hemicelluloses in Micrasterias (Streptophyta). Journal of Phycology 44, 1221–1234, doi:10.1111/j.1529-8817.2008.00576.x (2008). 21. Fry, S. C., Nesselrode, B. H. W. A., Miller, J. G. & Mewburn, B. R. Mixed-linkage (1 → 3,1 → 4)-β-d-glucan is a major hemicellulose of Equisetum (horsetail) cell walls. New Phytologist 179, 104–115, doi:10.1111/j.1469-8137.2008.02435.x (2008). 22. Harholt, J. et al. The glycosyltransferase repertoire of the spikemoss Selaginella moellendori ffi and a comparative study of its cell wall. PLoS One 7, e35846, doi:10.1371/journal.pone.0035846 (2012). 23. Sørensen, I. et al. Mixed-linkage (1 → 3), (1 → 4)-ß-D-glucan is not unique to the Poales and is an abundant component of Equisetum arvense cell walls. The Plant Journal 54, 510–521, doi:10.1111/j.1365-313X.2008.03453.x (2008). 24. Mikkelsen, M. D. et al. Evidence for land plant cell wall biosynthetic mechanisms in charophyte green algae. Annals of Botany 114, 1217–1236, doi:10.1093/aob/mcu171 (2014). 25. Burton, R. A. & Fincher, G. B. Evolution and Development of Cell Walls in Cereal Grains. Frontiers in Plant Science 5, doi:10.3389/ fpls.2014.00456 (2014). 26. Labourel, A. et al. e Th β-Glucanase ZgLamA from Zobellia galactanivorans evolved a bent active site adapted for efficient degradation of algal laminarin. Journal of Biological Chemistry 289, 2027–2042, doi:10.1074/jbc.M113.538843 (2014). 27. Marcus, S. E. et al. Pectic homogalacturonan masks abundant sets of xyloglucan epitopes in plant cell walls. BMC Plant Biol 8, 60, doi:10.1186/1471-2229-8-60 (2008). 28. Hervé, C., Rogowski, A., Gilbert, H. J. & Knox, J. P. Enzymatic treatments reveal differential capacities for xylan recognition and degradation in primary and secondary plant cell walls. The Plant Journal 58, 413–422, doi:10.1111/j.1365-313X.2009.03785.x (2009). 29. Lundqvist, L. C. E., Jam, M., Barbeyron, T., Czjzek, M. & Sandström, C. Substrate specificity of the recombinant alginate lyase from the marine bacteria Pseudomonas alginovora. Carbohydrate Research 352, 44–50, doi:10.1016/j.carres.2012.02.014 (2012). 30. oThmas, F. et al. Comparative characterization of two marine alginate lyases from Zobellia galactanivorans reveals distinct modes of action and exquisite adaptation to their natural substrate. Journal of Biological Chemistry 288, 23021–23037, doi:10.1074/jbc. M113.467217 (2013). 31. Olafsdottir, E. S. & Ingólfsdottir, K. Polysaccharides from Lichens: Structural characteristics and biological activity. Planta Med 67, 199–208, doi:10.1055/s-2001-12012 (2001). 32. Pettolino, F. et al. Hyphal cell walls from the plant pathogen Rhynchosporium secalis contain (1,3/1,6)-β-d-glucans, galacto- and rhamnomannans, (1,3;1,4)-β-d-glucans and chitin. FEBS Journal 276, 3698–3709, doi:10.1111/j.1742-4658.2009.07086.x (2009). 33. Bourne, E. J., Bruch, P. & Percival, E. The active carbohydrate metabolites of the brown seaweed Fucus vesiculosus . Carbohydrate Research 9, 415–422, doi:10.1016/S0008-6215(00)80026-9 (1969). 34. Novotny, A. M. & Forman, M. The composition and development of cell walls of Fucus embryos. Planta 122, 67–78, doi:10.1007/ BF00385406 (1975). 35. Xue, X. & Fry, S. C. Evolution of mixed-linkage (1 → 3,1 → 4)-β-d-glucan (MLG) and xyloglucan in Equisetum (horsetails) and other monilophytes. Annals of Botany 109, 873–886, doi:10.1093/aob/mcs018 (2012). 36. Sørensen, I. et al. The charophycean green algae provide insights into the early origins of plant cell walls. The Plant Journal 68, 201–211, doi:10.1111/j.1365-313X.2011.04686.x (2011). 37. Gibeaut, D. M., Pauly, M., Bacic, A. & Fincher, G. B. Changes in cell wall polysaccharides in developing barley (Hordeum vulgare) coleoptiles. Planta 221, 729–738, doi:10.1007/s00425-005-1481-0 (2005). 38. Dimitroff, G. et al. (1,3;1,4)-β-glucan biosynthesis by the CSLF6 enzyme: position and flexibility of catalytic residues influence product fine structure. Biochemistry 55, 2054–2061, doi:10.1021/acs.biochem.5b01384 (2016). 39. Jobling, S. A. Membrane pore architecture of the CslF6 protein controls (1-3,1-4)-β-glucan structure. Science Advances 1, e1500069–e1500069, doi:10.1126/sciadv.1500069 (2015). 40. Cock, J. M., Coelho, S. M., Brownlee, C. & Taylor, A. R. The Ectocarpus genome sequence: insights into brown algal biology and the evolutionary diversity of the eukaryotes. New Phytol 188, 1–4, doi:10.1111/j.1469-8137.2010.03454.x (2010). 41. Michel, G., Tonon, T., Scornet, D., Cock, J. M. & Kloareg, B. The cell wall polysaccharide metabolism of the brown alga Ectocarpus siliculosus. Insights into the evolution of extracellular matrix polysaccharides in Eukaryotes. New Phytol 188, 82–97, doi:10.1111/j.1469-8137.2010.03374.x (2010). 42. Scholz, M. J. et al. Ultrastructure and composition of the Nannochloropsis gaditana cell wall. Eukaryot Cell 13, 1450–1464, doi:10.1128/EC.00183-14 (2014). 43. Harholt, J., Moestrup, Ø. & Ulvskov, P. Why plants were terrestrial from the beginning. Trends in Plant Science 21, 96–101, doi:10.1016/j.tplants.2015.11.010 (2016). 44. Moller, I. et al. High-throughput mapping of cell-wall polymers within and between plants using novel microarrays. The Plant Journal 50, 1118–1128, doi:10.1111/j.1365-313X.2007.03114.x (2007). Acknowledgements A.S.S. was supported by the Danish Strategic Research Council under the Sustainable Enzyme Technologies for Future Bioenergy (SET4Future) program. Visits of A.S.S. to Station Biologique de Roscoff (France) were supported by short term grants provided by the European Commission through under the FP7 ‘capacities’ specic fi program ASSEMBLE. C.H. acknowledges funding from the French National Research Agency with regard to the investment expenditure program Idealg (http://www.idealg.ueb.eu/, grant agreement no. ANR-10-BTBR-04). We sincerely thank all the staff from the Marine Service at the Station Biologique de Roscoff for the sampling of local marine species. J.H. and F.Q. were supported by the Carlsberg Foundation. Author Contributions C.H., G.M., M.C., W.G.T.W. and A.A.S. planned and designed the research. D.D., F.Q., and J.H. performed the research. A.A.S. performed the research, analysed data and contributed in the writing of the paper. C.H., W.G.T.W., J.H., G.M. and M.C. contributed to the writing of the paper. Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 10 www.nature.com/scientificreports/ Additional Information Supplementary information accompanies this paper at doi:10.1038/s41598-017-03081-5 Competing Interests: The authors declare that they have no competing interests. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2017 Scientific Repo R ts | 7: 2880 | DOI:10.1038/s41598-017-03081-5 11
Scientific Reports – Springer Journals
Published: Jun 6, 2017
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera