Metagenomic insights into the dynamic degradation of brown algal polysaccharides by kelp-associated microbiota

ABSTRACT

Marine bacteria play important roles in the degradation and cycling of algal polysaccharides. However, the dynamics of epiphytic bacterial communities and their roles in algal polysaccharide degradation during kelp decay are still unclear. Here, we performed metagenomic analyses to investigate the identities and predicted metabolic abilities of epiphytic bacterial communities during the early and late decay stages of the kelp Saccharina japonica. During kelp decay, the dominant epiphytic bacterial communities shifted from Gammaproteobacteria to Verrucomicrobia and Bacteroidetes. In the early decay stage of S. japonica, epiphytic bacteria primarily targeted kelp-derived labile alginate for degradation, among which the gammaproteobacterial Vibrionaceae (particularly Vibrio) and Psychromonadaceae (particularly Psychromonas), abundant in alginate lyases belonging to the polysaccharide lyase (PL) families PL6, PL7, and PL17, were key alginate degraders. More complex fucoidan was preferred to be degraded in the late decay stage of S. japonica by epiphytic bacteria, predominantly from Verrucomicrobia (particularly Lentimonas), Pirellulaceae of Planctomycetes (particularly Rhodopirellula), Pontiellaceae of Kiritimatiellota, and Flavobacteriaceae of Bacteroidetes, which depended on using glycoside hydrolases (GHs) from the GH29, GH95, and GH141 families and sulfatases from the S1_15, S1_16, S1_17, and S1_25 families to depolymerize fucoidan. The pathways for algal polysaccharide degradation in dominant epiphytic bacterial groups were reconstructed based on analyses of metagenome-assembled genomes. This study sheds light on the roles of different epiphytic bacteria in the degradation of brown algal polysaccharides.

IMPORTANCE

Kelps are important primary producers in coastal marine ecosystems. Polysaccharides, as major components of brown algal biomass, constitute a large fraction of organic carbon in the ocean. However, knowledge of the identities and pathways of epiphytic bacteria involved in the degradation process of brown algal polysaccharides during kelp decay is still elusive. Here, based on metagenomic analyses, the succession of epiphytic bacterial communities and their metabolic potential were investigated during the early and late decay stages of Saccharina japonica. Our study revealed a transition in algal polysaccharide-degrading bacteria during kelp decay, shifting from alginate-degrading Gammaproteobacteria to fucoidan-degrading Verrucomicrobia, Planctomycetes, Kiritimatiellota, and Bacteroidetes. A model for the dynamic degradation of algal cell wall polysaccharides, a complex organic carbon, by epiphytic microbiota during kelp decay was proposed. This study deepens our understanding of the role of epiphytic bacteria in marine algal carbon cycling as well as pathogen control in algal culture.

INTRODUCTION

Brown macroalgae are important primary producers in coastal marine ecosystems, contributing about 1 Gt of biomass to global primary production annually (1). Among brown macroalgae, the order Laminariales can form kelp forests, one of the most productive ecosystems within the oceans (2). Kelp forests export significant amounts of their photosynthetically derived carbon as dissolved and particulate organic carbon to the surrounding seawater and deep sea through algal exudation or cell death (3). Heterotrophic marine bacteria are responsible for metabolizing and remineralizing alga-derived organic carbon, thus playing key roles in algal carbon cycling (4, 5).

Polysaccharides are major components of macroalgal biomass and constitute a large fraction of organic carbon in the ocean (6). In brown algae, alginate, fucoidan, and laminarin are dominant polysaccharides, with their relative abundances varying across species and seasons (7). Alginate is usually the most abundant polysaccharide in the cell walls of brown algae, constituting up to 40% of their dry weight (8). It is a linear polysaccharide composed of two uronic acid monomers, β-D-mannuronate and its C5-epimer α-L-guluronate, through 1,4-linked glycosidic bonds. Metagenomic and genomic analyses suggest that marine alginate-degrading bacteria are mainly affiliated with Bacteroidetes and Gammaproteobacteria (9–14). Microbial degradation of alginate is initiated by depolymerization of this polysaccharide catalyzed by alginate lyases (15). In the metagenomes of cold coastal sediments, the majority of detected bacterial alginate lyase genes belong to the polysaccharide lyase (PL) families including PL6, PL7, and PL17 (10, 16).

Fucoidan is also a structural polysaccharide in the cell walls of brown algae, constituting 3%–28% of their dry weight (17). It is a highly heterogeneous sulfated polysaccharide with complex branching. In addition to fucose, fucoidan comprises other monosaccharides such as galactose, glucose, mannose, xylose, and uronic acids (18). Fucoidan is grouped into two main classes based on sugar backbone: homofucan consisting of either α-1,3 or alternating α-1,3/1,4-linked L-fucose with sulfation on O-2, O-3, or O-4 and heterofucan composed of mannose, galactose, or glucuronic acid with branches of sulfated fucose (19, 20). Due to its complex chemical structure, fucoidan is more slowly degraded by marine bacteria than laminarin and other labile algal polysaccharides (21, 22). Until now, only a limited number of marine bacterial isolates from Bacteroidetes, Gammaproteobacteria, and Planctomycetes–Verrucomicrobia–Chlamydiae (PVC) are reported to have fucoidan-degrading abilities (23–25). Marine Verrucomicrobium “Lentimonas” sp. CC4, a highly specialized fucoidan degrader, uses hundreds of enzymes to break down fucoidan, primarily exo-acting enzymes from glycoside hydrolase (GH) families, GH29, GH95, and GH141 (26). Based on metagenomic analyses of microbial communities in intertidal sediments as well as during the algal bloom, most enzymes involved in marine fucoidan degradation belong to families GH29 and GH95, which are mainly distributed in Bacteroidetes, Gammaproteobacteria, and the PVC group (10, 21).

Laminarin is the main intracellular storage polysaccharide in brown algae, accounting for up to 35% of the algal dry weight (27). Its structure, relatively simpler compared to alginate and fucoidan, consists of glucose monomers with a β-1,3-linked backbone (20–30 residues) and occasional β-1,6-linked branches. Laminarin is also present in diatoms (28). During diatom-dominated blooms, Bacteroidetes and Gammaproteobacteria are considered as pivotal laminarin degraders, which encode glycoside hydrolases frequently from families GH3, GH16, GH17, and GH30 for laminarin degradation (21, 29, 30).

Both alga-associated and free-living bacteria play crucial roles in the degradation and cycling of algal polysaccharides. Compared to extensive studies on the degradation of algal polysaccharides by bacterioplankton (9, 11, 30–33), there are far fewer reports on the role of epibiota in algal polysaccharide cycling. Microbial communities associated with Macrocystis and other brown algae exhibit distinct profiles from those in the surrounding seawater (34, 35). Moreover, epibiota diversity and composition can be influenced by host species identity, blade tissue age, and condition of kelps, showing seasonal succession (36, 37). Although several metagenomic studies on the diversity and alginate-degrading potential of marine bacteria associated with kelps including Macrocystis pyrifera and Undaria pinnatifida have been reported (34, 38), the successional dynamics of epiphytic microbial communities during kelp decay and their metabolic capacity for degrading fucoidan, laminarin, and other algal polysaccharides remain unknown.

Saccharina japonica (formerly Laminaria japonica), a typical representative species of kelps, is a perennial alga for its wild population, and the life span of its sporophyte is biennial (39). It has also been cultivated in Asia on annual or biennial life spans (40). The natural degradation process of kelps is intricate and influenced by various factors. Here, to simplify the decay process of S. japonica and focus on main kelp-associated microbes, we employed relatively simple laboratory decay mimics. Metagenomic analyses were performed to investigate the succession of epiphytic microbial communities and their metabolic potential to degrade brown algal polysaccharides including alginate, fucoidan, and laminarin during the decay of S. japonica. Based on metagenome-assembled genomes (MAGs), the pathways for algal polysaccharide degradation in dominant epiphytic bacterial groups were reconstructed. Finally, a model for the dynamic degradation of algal polysaccharides by epiphytic microbiota during kelp decay was proposed.

RESULTS

Changes in epiphytic bacterial community during kelp decay

The kelp Saccharina japonica was collected from an intertidal zone of the Bohai Sea, China, in September 2019. To understand the dynamics of epiphytic bacterial communities during kelp decay, shotgun metagenomic sequencing libraries were constructed using environmental DNA extracted from bacteria swabbed from the surfaces of S. japonica in the early and late decay stages. A total of 52.12 Gb clean data were obtained from the two samples, covering 99.8% of the total raw data (Data Set S1). After the removal of reads from the kelp host, the residual clean reads were assembled into contigs ranging from 0.39 to 0.42 Gb with an average N50 of 2,447 bp (Data Set S1). Taxonomic profiling based on metagenomic data revealed that Gammaproteobacteria dominated the epiphytic bacterial communities, accounting for 77.5% and 57.4% of the total in the early and late decay stages, respectively (Fig. 1). In the early decay stage, members of families Moraxellaceae (31.4%) and Vibrionaceae (27.2%) from Gammaproteobacteria were the most abundant, decreasing to 18.4% and 1.0%, respectively, in the late decay stage. A considerable amount of Enterobacteriaceae from Gammaproteobacteria (3.4%), Clostridia (Bacillota, 3.7%), and Fusobacteriia (Fusobacteriota, 4.1%) were also detected in the early decay stage, whereas their relative abundances in the late decay stage were lower than 0.5%. In the late decay stage, Oceanospirillaceae (19.4%) as well as Moraxellaceae (18.4%) from Gammaproteobacteria were predominant. The relative abundances of Alphaproteobacteria, Flavobacteriia (Bacteroidetes), and Opitutae (Verrucomicrobia) increased significantly from ≤2.0% in the early decay stage to 6.7%–12.7% in the late decay stage. In addition, members of Halomonadaceae (particularly the genus Cobetia) from Gammaproteobacteria consistently occurred during kelp decay, accounting for approximately 4.0% of the community composition (Fig. 1).

Fig 1.

The relative abundance of bacterial taxa associated with S. japonica in the early (a) and late (b) decay stages. “Other bacteria” represents bacterial groups with relative abundance lower than 1% at the class level.

At the genus level, Psychrobacter (Moraxellaceae of Gammaproteobacteria) was the most abundant, accounting for 18.2%–31.0%. Other genera including Vibrio (Vibrionaceae of Gammaproteobacteria, 17.6%), Photobacterium (Vibrionaceae of Gammaproteobacteria, 6.6%), and Fusobacterium (Fusobacteriaceae of Fusobacteriia, 3.4%) dominated in the early decay stage, whereas Oceanospirillum (Oceanospirillaceae of Gammaproteobacteria, 12.4%), Celeribacter (Roseobacteraceae of Alphaproteobacteria, 4.7%), Marinobacter (Marinobacteraceae of Gammaproteobacteria, 4.5%), and Marinomonas (Oceanospirillaceae of Gammaproteobacteria, 3.5%) were enriched in the late decay stage.

The observed changes in epiphytic bacterial communities during algal decay (Fig. 1) may be related to the decaying conditions of S. japonica, perhaps owing to the varying abilities of different epiphytic bacterial groups in the utilization of brown algal polysaccharides to some degree.

Changes in alginate-degrading bacteria and their alginate lyase genes

In brown algae, the most abundant polysaccharide is usually alginate, and bacteria-secreting alginate lyases are considered pivotal in its degradation. Currently, alginate lyases are classified into 14 families of polysaccharide lyases including PL5, PL6, PL7, PL8, PL14, PL15, PL17, PL18, PL31, PL32, PL34, PL36, PL39, and PL41 according to the CAZy database (41). Metagenomic analysis showed that alginate-degrading epiphytic bacteria and their alginate lyase genes varied during kelp decay (Fig. 2; Fig. S1). In the early decay stage of S. japonica, a total of 113 alginate lyase genes were identified in epiphytic bacteria, mainly belonging to Gammaproteobacteria, followed by Bacteroidetes (Fig. 2a and c). Of these genes, 38.9% encoded potential extracellular enzymes (Fig. 2a). The majority (80.0%) of the predicted extracellular alginate lyases belonged to the PL7 family, mainly assigned to Vibrionaceae (particularly Vibrio) and Psychromonadaceae (particularly Psychromonas) of Gammaproteobacteria (Fig. 2b through d; Data Set S2). Additionally, a small portion (14.4%) of the predicted extracellular alginate lyases were the PL6 and PL17 enzymes, primarily found in Bacteroidetes (e.g., Flavobacteriaceae) (Fig. 2b through d). Characterized PL6 and PL7 alginate lyases exhibit diverse substrate specificities and modes of action (15, 42), while PL17 enzymes typically function as exolytic oligo-alginate lyases that can degrade alginate oligosaccharides into unsaturated monosaccharides (15, 43). The above bacterial groups carrying extracellular alginate lyase genes are thought to be potential alginate degraders. Moreover, the majority of potential intracellular alginate lyases were grouped into the PL17, PL15, and PL7 families, which were mainly distributed in Vibrionaceae (particularly Vibrio), Psychromonadaceae (particularly Psychromonas), and Flavobacteriaceae (Fig. 2; Fig. S1 and Data Set S2), underscoring their key roles in the degradation of alginate derived from the early decaying kelp.

Fig 2.

Shift in alginate-degrading epiphytic bacteria and their alginate lyase genes during kelp decay. (a) Changes in the relative abundance of predicted extracellular and intracellular alginate lyase genes. Relative abundance and number (in parentheses) of related alginate lyase genes were shown for each sample. (b) Changes in the families of predicted extracellular and intracellular alginate lyase genes. (c) Changes in the bacterial groups containing predicted extracellular or intracellular alginate lyase genes at the class level. In (b) and (c), “X1-Extra” and “X1-Intra” represent predicted extracellular or intracellular alginate lyase genes from early decaying kelp-associated bacteria, respectively, and “X2-Extra” and “X2-Intra” represent predicted extracellular or intracellular alginate lyase genes from late decaying kelp-associated bacteria, respectively. (d) Heatmap showing detailed changes in the relative abundance (log₂) of predicted extracellular alginate lyase genes and their related bacterial groups. The colors in the heatmap indicate relative gene abundance, ranging from high (black) to low (white) abundance. Relative abundance and number (in parentheses) of alginate lyase genes from each family were shown on top of the heatmap.

Compared to the early decay stage, the detected bacterial alginate lyase genes in the late decay stage were similar in number but significantly decreased in abundance (Fig. 2a; Fig. S2). In the late decay stage, among the epiphytic bacteria carrying alginate lyase genes, members of Gammaproteobacteria remarkably decreased, accompanied by a significant increase of Bacteroidetes, Verrucomicrobia, and Planctomycetes (Fig. 2c). Most alginate lyases including potential extracellular and intracellular enzymes belonged to the PL6, PL7, and PL17 families, which were predominantly distributed in Gammaproteobacteria (such as Psychromonadaceae), Flavobacteriaceae of Bacteroidetes, and Verrucomicrobia (particularly Puniceicoccaceae) (Fig. 2d; Fig. S1), suggesting the important roles of these bacterial groups in the degradation of late decaying kelp-derived alginate. At the Patagonian continental shelf, the added alginate can be almost completely degraded by Alteromonadaceae and other planktonic bacteria within 7 days (9). The lower abundance of alginate-degrading bacteria associated with the late decaying kelp (Fig. S2) suggested that most kelp-derived alginate might be degraded by epiphytic bacteria in a relatively early decay stage of S. japonica.

With the action of alginate lyases, alginate is degraded into oligomers extracellularly, which would be further absorbed and utilized by bacterial cells. Therefore, we also investigated changes in outer membrane transporter genes involved in alginate utilization and their related bacterial groups during kelp decay (Fig. S3). Alginate oligomers are transported into the periplasm via a predicted SusC transporter encoded by the polysaccharide utilization locus (PUL) system of most alginate-degrading bacteria (13, 44–46) and potentially via the porin KdgM/N in Vibrio strains (14, 47). Like the observed decline in alginate lyase genes of epiphytic bacteria (Fig. S2), the relative abundance of their outer membrane transporter genes involved in alginate utilization decreased in the late stage compared to the early stage (Fig. S3), further supporting that alginate was preferred to be degraded by epiphytic bacteria in a relatively early decay stage of S. japonica. In the early decay stage, outer membrane transporter genes involved in alginate utilization were mainly distributed in Vibrionaceae (particularly Vibrio), followed by Flavobacteriaceae and Bacteroidaceae (Fig. S3 and Data Set S2).

Changes in fucoidan-degrading bacteria and their fucoidan-degrading genes

In addition to alginate, fucoidan is also a structural polysaccharide in the cell walls of brown algae. The fucoidan fraction extracted from S. japonica, cultured in Qingdao, China, is a homofucan containing 75% of 1,3-linked fucose residues and 25% of 1,4-linked fucose residues with sulfation on O-4 and/or O-2 in the backbone (48). Fucoidanases that target fucosidic bonds in fucoidan are grouped into seven families including GH29, GH95, GH107, GH139, GH141, GH151, and GH168 in the CAZy database (41). During kelp decay, epiphytic bacteria tended to employ GH29, GH95, and GH141 enzymes to degrade fucoidan (Fig. 3). Currently, characterized fucoidan-active glycoside hydrolases from GH29, GH95, and GH141 families are all exo-acting enzymes (49, 50). Therefore, we hypothesize that fucoidan-degrading epiphytic bacteria may debranch and utilize kelp-derived fucoidan mainly via exo-enzymes, like the highly specialized fucoidan degrader, “Lentimonas” sp. CC4 (26).

Fig 3.

Shift in fucoidan-degrading epiphytic bacteria and their fucoidanase genes during kelp decay. (a) Changes in the relative abundance of predicted extracellular and intracellular fucoidanase genes. Relative abundance and number (in parentheses) of related fucoidanase genes were shown for each sample. (b) Changes in the families of predicted extracellular and intracellular fucoidanase genes. (c) Changes in the bacterial groups containing predicted extracellular or intracellular fucoidanase genes at the class level. In (b) and (c), “X1-Extra” and “X1-Intra” represent predicted extracellular or intracellular fucoidanase genes from early decaying kelp-associated bacteria, respectively, and “X2-Extra” and “X2-Intra” represent predicted extracellular or intracellular fucoidanase genes from late decaying kelp-associated bacteria, respectively. (d) Heatmap showing detailed changes in the relative abundance (log₂) of predicted extracellular fucoidanase genes and their related bacterial groups. The colors in the heatmap indicate relative gene abundance, ranging from high (black) to low (white) abundance. Relative abundance and number (in parentheses) of fucoidanase genes from each family were shown on top of the heatmap.

In the early decay stage of S. japonica, the number and abundance of detected fucoidanase genes were relatively low, nearly all distributed in Bacteroidetes (Flavobacteriaceae and Porphyromonadaceae) (Fig. 3; Fig. S4). However, fucoidanase genes in the late decay stage significantly increased in number and abundance, and the abundance ratio of extracellular to intracellular fucoidanase genes further enlarged (Fig. 3a; Fig. S2), suggesting that epiphytic bacteria preferred to utilize the complex fucoidan mainly in the late decay stage. In the late decay stage, the dominant fucoidan-degrading bacterial communities shifted from members of Bacteroidetes to those of Verrucomicrobia, Planctomycetes, and Kiritimatiellota (Fig. 3). Verrucomicrobia (particularly Lentimonas) was the most abundant bacterial group harboring putative fucoidanase genes (Fig. 3d; Fig. S4). Pirellulaceae of Planctomycetes (particularly Rhodopirellula), Pontiellaceae of Kiritimatiellota (particularly Pontiella), and Flavobacteriaceae of Bacteroidetes also contributed to fucoidan degradation (Fig. 3d; Fig. S4 and Data Set S3).

Different from alginate, laminarin, and other brown algal polysaccharides, fucoidan is a highly sulfated polysaccharide. In addition to glycoside hydrolases, the microbial degradation of fucoidan also involves sulfatases. However, the distribution of sulfatases and their bacterial sources in marine environments is largely unknown so far. Therefore, we analyzed changes in sulfatase genes and their related bacterial groups during kelp decay (Fig. 4; Fig. S5). Like fucoidanase genes (Fig. 3a), the number and abundance of sulfatase genes in the late stage were higher than those in the early stage (Fig. 4a; Fig. S2). Moreover, the abundance ratio of extracellular to intracellular sulfatase genes increased as well in the late stage (Fig. 4), providing further evidence for the degradation of fucoidan by epiphytic bacteria mainly in the late decay stage. At present, characterized fucoidan sulfatases are classified into five families including S1_13, S1_15, S1_16, S1_17, and S1_25 according to the SulfAtlas database (51). During kelp decay, epiphytic bacterial groups carrying sulfatase genes shifted from members of Bacteroidetes and Gammaproteobacteria to those of the PVC group (Fig. 4c). In the late decay stage, sulfatases were dominated by potential fucoidan sulfatases belonging to the S1_15, S1_16, S1_17, and S1_25 families, which were mainly distributed in Verrucomicrobia (e.g., Lentimonas, Verrucomicrobiaceae, and unclassified Opitutae), Kiritimatiellota (particularly Pontiella of Pontiellaceae), and Bacteroidetes (particularly Flavobacteriaceae) (Fig. 4d; Fig. S5 and Data Set S3), further suggesting the important roles of these bacterial groups in fucoidan degradation.

Fig 4.

Changes in sulfatase-producing epiphytic bacteria and their sulfatase genes during kelp decay. (a) Changes in the relative abundance of predicted extracellular and intracellular sulfatase genes. Relative abundance and number (in parentheses) of related sulfatase genes were shown for each sample. (b) Changes in the families of predicted extracellular and intracellular sulfatase genes. (c) Changes in the bacterial groups containing predicted extracellular or intracellular sulfatase genes at the class level. In (b) and (c), “X1-Extra” and “X1-Intra” represent predicted extracellular or intracellular sulfatase genes from early decaying kelp-associated bacteria, respectively, and “X2-Extra” and “X2-Intra” represent predicted extracellular or intracellular sulfatase genes from late decaying kelp-associated bacteria, respectively. (d) Heatmap showing detailed changes in the relative abundance (log₂) of predicted extracellular sulfatase genes and their related bacterial groups. The colors in the heatmap indicate relative gene abundance, ranging from high (black) to low (white) abundance. Relative abundance and number (in parentheses) of sulfatase genes from each family were shown on top of the heatmap.

Changes in laminarin-degrading bacteria and their laminarin-degrading genes

Laminarin, a kind of glucan, serves as the main intracellular storage polysaccharide in brown algae. The relative abundance of laminarin in brown algae can vary across species, seasons, and habitats (27). Saccharina japonica samples collected from Kunashir Island (Okhotsk Sea) in August 2012 were found to contain only 0.8% of laminarin (52). During kelp decay, the low abundance of epiphytic bacteria secreting potential laminarin-degrading enzymes (Fig. 5; Fig. S2) may reflect the low laminarin content in sampled S. japonica in this study. In the early decay stage of S. japonica, the most abundant glucanases putatively involved in laminarin degradation belonged to the GH3 family, which were mainly assigned to Bacteroidetes (e.g., Bacteroidaceae and Flavobacteriaceae) (Fig. 5d). In the late decay stage, in addition to Bacteroidetes, Kiritimatiellota (particularly Pontiellaceae) and Verrucomicrobia (particularly Verrucomicrobiaceae) were also potential laminarin degraders, which were abundant in GH3 and GH128 glucanases (Fig. 5d). Characterized GH3 glucanases are usually exo-enzymes that can release glucose from laminarin and other glucans (53), while GH128 enzymes display diversified exo- and endo-activities against β-1,3-glucan (54). Notably, during kelp decay, some epiphytic gammaproteobacterial groups (e.g., Enterovibrio and Chelonobacter) seemed to be incapable of secreting laminarinases/glucanases but possessed potential intracellular GH1, GH3, or other glucanases (Fig. S6 and Data Set S4), suggesting their preferences for utilizing laminarin oligomers over polymeric laminarin. These bacteria might participate in the degradation of decaying kelp-derived laminarin as cheaters by taking up laminarin oligomers produced by other degraders.

Fig 5.

Shift in laminarin-degrading epiphytic bacteria and their laminarin-degrading genes during kelp decay. (a) Changes in the relative abundance of predicted extracellular and intracellular laminarin-degrading genes. Relative abundance and number (in parentheses) of related laminarin-degrading genes were shown for each sample. (b) Changes in the families of predicted extracellular and intracellular laminarin-degrading genes. (c) Changes in the bacterial groups containing predicted extracellular or intracellular laminarin-degrading genes at the class level. In (b) and (c), “X1-Extra” and “X1-Intra” represent predicted extracellular or intracellular laminarin-degrading genes from early decaying kelp-associated bacteria, respectively, and “X2-Extra” and “X2-Intra” represent predicted extracellular or intracellular laminarin-degrading genes from late decaying kelp-associated bacteria, respectively. (d) Heatmap showing detailed changes in the relative abundance (log₂) of predicted extracellular laminarin-degrading genes and their related bacterial groups. The colors in the heatmap indicate relative gene abundance, ranging from high (black) to low (white) abundance. Relative abundance and number (in parentheses) of laminarin-degrading genes from each family were shown on top of the heatmap.

Identification of potential algal polysaccharide-degrading MAGs

To predict degradation pathways for brown algal polysaccharides in epiphytic bacteria, we assembled 32 non-redundant bacterial MAGs with >70% completeness and <6% contamination (Table S1). These MAGs were classified into 11 bacterial classes, showing notable abundance in epiphytic metagenomes (Table S1; Fig. 1). Eighteen of these MAGs, including seven Gammaproteobacteria MAGs, four Flavobacteriia MAGs, two Opitutae MAGs, and a single MAG in each of Spirochaetia, Bacilli, Alphaproteobacteria, Verrucomicrobiae, and Kiritimatiellia, possessed algal polysaccharide-degrading genes (Fig. 6 and 7; Fig. S7), suggesting their capacity to degrade brown algal polysaccharides.

Fig 6.

Heatmap showing the number of homologs of putative brown algal polysaccharide-degrading enzymes in MAGs.

Fig 7.

Key alginate-degrading (a) and fucoidan-degrading (b) genes in epiphytic MAGs. open reading frames (ORFs) were depicted using different colored arrows based on their functional annotations. Putative alginate lyases, GHs, and sulfatases were indicated according to dbCAN and Pfam analyses. MFS, major facility superfamily; TRAP, tripartite ATP-independent periplasmic; ABC, ATP-binding cassette.

Reconstructed alginate utilization pathways in Gammaproteobacteria

Seven Gammaproteobacteria MAGs, including two Pseudoalteromonas MAGs (X1_bin3 and X2_bin5), two Shewanella MAGs (X1_bin53 and X2_bin34), one Psychromonas MAG (X1_bin8), one Cobetia MAG (X2_bin1), and one Marinomonas MAG (X2_bin13), contained alginate lyase genes (Fig. 6 and 7), suggesting their capacity to degrade alginate. All these MAGs but X1_bin8 had only alginate-degrading genes without other algal polysaccharide-degrading genes, suggesting their potential as specialized alginate degraders. Members of genera Pseudoalteromonas, Shewanella, Psychromonas, Cobetia, and Marinomonas have been reported to have alginate-degrading abilities (46, 55). There are three alginate degradation pathways for alginate-utilizing bacteria, the PUL system in most alginate-degrading bacteria, the “scattered” system in gammaproteobacterial Vibrio strains, and the “pit” transport system in alphaproteobacterial strain Sphingomonas sp. A1 and its relatives (14). The presence of the hallmark susC gene (Fig. 7a) suggested that these Gammaproteobacteria MAGs seemed to utilize alginate via the PUL system. These MAGs were predicted to secrete one to four extracellular alginate lyases belonging to PL6, PL7, and/or PL17, which could break down polymeric alginate into oligomers. Once transported into the periplasmic space via a SusC transporter, oligomers were further degraded into unsaturated monosaccharides by periplasmic alginate lyases from PL17 or other families. Unsaturated monosaccharides were subsequently transported into the cytoplasm via inner membrane MFS permease (Fig. 8a). These Gammaproteobacteria MAGs possessed all the enzyme genes, kdgF (56), dehR (44), kdgK (44), and eda (44), necessary for further catabolism of unsaturated monosaccharides. Via these four enzymes, intracellular monosaccharides would be converted to glyceraldehyde triphosphate and pyruvate to enter the glycolysis pathway (Fig. 8a; Data Set S5).

Fig 8.

Predicted degradation pathways for brown algal polysaccharides in representative epiphytic MAGs. (a) X1_bin53 (Shewanellaceae). (b) X1_bin16 (Enterobacteriaceae). (c) X2_bin3 (Puniceicoccaceae). (d) X2_bin21 (Pontiellaceae). (e) X1_bin9 (Flavobacteriaceae). (f) X2_bin33 (Flavobacteriaceae). Enzymes involved in alginate, fucoidan, and laminarin utilization are colored in red, green, and blue, respectively. Symbols for transporters involved in alginate, fucoidan, and laminarin utilization are shown in red, green, and blue, respectively. Solid arrows denote enzymatic reactions, and dotted arrows denote transport. The unknown transporters are indicated by a question mark (?). The cellular location of enzymes/transporters is predicted according to PSORTb v3.0 (57) and CELLO v.2.5 (58) combined with SignalP 5.0 (59), and enzymes without predictable cellular locations are not shown. DEH, 4-deoxy-L-erythro-5-hexoseulose uronic acid; KDG, 2-keto-3-deoxy-D-gluconate; KDPG, 2-keto-3-deoxy-6-phosphogluconate.

Reconstructed fucoidan utilization pathways in the PVC group

All four MAGs from the PVC group, including X2_bin3, X2_bin9, X2_bin16, and X2_bin21, were reconstructed from the late decaying kelp-associated bacterial metagenome. These MAGs contained 3–50 homologs of exo-acting fucosidases from families GH29, GH95, and/or GH141, 0–5 homologs of endo-acting GH168 fucoidanases, and 23–41 homologs of fucoidan sulfatases from families S1_15, S1_16, S1_17, and S1_25 (Fig. 6), suggesting their roles in fucoidan degradation. X2_bin3, X2_bin9, and X2_bin21 could secrete a series of fucoidanases and fucoidan sulfatases to degrade fucoidan into partially desulfated fucooligosaccharides (Fig. 8c and d), while X2_bin16 seemed to target fucooligosaccharide degradation, potentially due to the lack of extracellular fucoidanases (Fig. S8a). Some Bacteroidetes were reported to have PULs targeting fucoidan, in which fucoidan-degrading genes were clustered with some outer membrane susC genes (45). The encoded SusC proteins in these PULs are presumed to facilitate the import of fucooligosaccharides and fucose into these Bacteroidetes. However, all MAGs from the PVC group lacked homologs of such SusC proteins, and the outer membrane transporters responsible for importing extracellular fucooligosaccharides in the PVC group are still unknown so far. Once inside the periplasm, the oligosaccharides were further hydrolyzed into fucose by periplasmic fucoidanases and fucoidan sulfatases. Fucose was then delivered into the cytoplasm via L-fucose/H⁺ symporter FucP. Currently, bacterial fucose catabolism is classified into two pathways, the phosphorylative pathway involving kinases and the non-phosphorylative pathway without kinases (60). The two Puniceicoccaceae MAGs, X2_bin3 and X2_bin9, were suggested to adopt the phosphorylative bacterial microcompartment (BMC) pathway to metabolize fucose (Fig. 8c), akin to other fucoidan-degrading bacteria from the PVC group such as “Lentimonas” sp. CC4 (26) and Planctomyces limnophilus (61). With the actions of mutarotase FucU, isomerase FucI, kinase FucK, and aldolase FucA, intracellular fucose was converted into toxic lactaldehyde intermediate, which was finally fermented into 1,2-propanediol and lactate in a BMC (Fig. 8c; Data Set S6). Like fucose-utilizing Escherichia strains (62), the Pontiellaceae MAG (X2_bin21) appeared to employ the phosphorylative pathway without a BMC (Fig. 8d). Within this pathway, fucose underwent sequential conversions into lactaldehyde by enzymes FucU, FucI, FucK, and FucA, and lactaldehyde was then aerobically oxidized to lactate by dehydrogenase AldA or anaerobically reduced to 1,2-propanediol by reductase FucO (Fig. 8d; Data Set S6). Except for the outer membrane transporter gene, X2_bin21 possessed an inner membrane toaABC transporter gene and all enzyme genes required for alginate utilization (Fig. 8d; Data Set S5). This suggested the capability of X2_bin21 to degrade alginate as well, which has never been reported for Kiritimatiellota. Although the mutarotase FucU/FucM was not found, the Akkermansiaceae MAG (X2_bin16) appeared to use the non-phosphorylative fucose pathway (Route III), akin to Xanthomonas campestris (63). In addition to FucU/FucM, this pathway also involved other enzymes including dehydrogenase FucD (EC 1.1.1.122), hydrolase FucB, dehydratase FucC, dehydrogenase FucF, and hydrolase FucG to covert fucose into lactate and pyruvate (Fig. S8a and Data Set S6).

Reconstructed polysaccharide utilization pathways in Bacteroidetes

Members of Bacteroidetes are well known for their pivotal role in polysaccharide degradation facilitated by diverse PULs (45, 64). Genomic analysis showed that four of five Flavobacteriaceae MAGs, including X1_bin9, X1_bin54, X2_bin25, and X2_bin33, had PULs targeting alginate, laminarin, and fucoidan (Fig. 7; Fig. S7), suggesting their strong capacity to degrade brown algal polysaccharides. Like other Bacteroidetes (44, 64), these Flavobacteriaceae MAGs possessed complete pathways to utilize alginate through the PUL system (Fig. 8e and f; Data Set S5). They also had all transporters and enzymes required for metabolizing laminarin (Fig. 8e and f; Data Set S7). These MAGs could secrete several putative GH16 β-1,3-glucanases and other glucanases to cleave extracellular laminarin into oligosaccharides, which would be transported into the periplasm via SusC and hydrolyzed into glucose by potential periplasmic GH3 and/or other glucanases. Once delivered into the cytoplasm via MFS, glucose was further catabolized through the glycolysis pathway (Fig. 8e and f). Although the number and families of potential fucoidanases and fucoidan sulfatases involved in fucoidan degradation were different among these MAGs, the core pathway for fucoidan utilization including the extracellular breakdown of fucoidan into fucooligosaccharides, the periplasmic hydrolysis of fucooligosaccharides into fucose, and the transport of fucooligosaccharides/fucose remained consistent (Fig. 8e and f; Data Set S6). However, the fucose catabolic pathways were different among these MAGs. X1_bin9, X1_bin54, and X2_bin25 tended to use the phosphorylative pathway without a BMC to catabolize fucose (Fig. 8e), while X2_bin33 seemed to use the non-phosphorylative fucose pathway (Route III) (Fig. 8f).

Reconstructed polysaccharide utilization pathways in other bacteria

Recently, bacteria from the class Erysipelotrichaceae of Bacillota were found to participate in the degradation of fucoidan (65). However, the role of Erysipelotrichaceae in alginate degradation remains unknown. The presence of a homolog to the inner membrane ABC transporter for alginate, three potential intracellular alginate lyases, and homologs of monosaccharide-catabolizing enzymes (KdgF, DehR, KdgK, and Eda) suggested that the Erysipelotrichaceae MAG (X1_bin7) might contribute to alginate degradation (Fig. S8b and Data Set S5). Moreover, genomic analysis showed that the alphaproteobacterial Parasphingorhabdus MAG (X2_bin11) tended to use the PUL system for alginate metabolism (Fig. S8c and Data Set S5), which is different from the “pit” transport system in its relative, Sphingomonas sp. A1 (66).

Until now, reports on algal-degrading Spirochaetes are sparse, leaving the roles of Spirochaetes in the degradation of algal polysaccharides largely unknown. The late decaying kelp-associated Spirochaetaceae MAG (X2_bin45) had all transporters (SusC and MFS) and enzymes required for metabolizing alginate, possibly adopting the PUL system to utilize alginate (Fig. S8d and Data Set S5). In addition to a potential intracellular fucoidanase, X2_bin45 possessed all transporters (a SusC transporter, an ABC transporter, and a TRAP transporter) and enzymes necessary for fucose utilization (Fig. 7b; Data Set S6), suggesting its preference for utilizing fucose/fucooligosaccharides over polymeric fucoidan. X2_bin45 was predicted to use the phosphorylative pathway without a BMC to catabolize fucose (Fig. S8d). Moreover, X2_bin45 possessed a complete pathway to utilize laminarin (Fig. S8d and Data Set S7).

In addition, genomic analysis suggested that the Enterobacteriaceae MAG (X1_bin16) was incapable of degrading fucoidan or other brown algal polysaccharides but seemed to catabolize fucose. Except for the outer membrane transporter gene, X1_bin16 contained an inner membrane ABC transporter gene and all the necessary enzyme genes for the phosphorylative fucose pathway without a BMC (Fig. 8b; Data Set S6).

A proposed model for the degradation of brown algal polysaccharides by epiphytic microbiota during kelp decay

Based on metagenomic analyses, a model for the dynamic degradation of algal polysaccharides by epiphytic microbiota during the early and late stages of kelp degradation is proposed (Fig. 9). During kelp decay, algal polysaccharide-degrading bacteria associated with S. japonica shift from Gammaproteobacteria to the PVC group and Bacteroidetes. Kelp-derived labile alginate is degraded in a relatively early decay stage by epiphytic bacteria mainly belonging to the gammaproteobacterial Vibrionaceae (particularly Vibrio) and Psychromonadaceae (particularly Psychromonas), which are abundant in PL6, PL7, and PL17 alginate lyases (Fig. 9). Members of early decaying kelp-associated Vibrio, Psychromonas, Shewanella, and other genera potentially secrete more than one type of alginate lyases to degrade alginate (Fig. 9; Data Set S2). Furthermore, some of these gammaproteobacterial groups possess the alginate utilization pathways (Fig. 8; Data Set S5), supporting the role of Gammaproteobacteria in alginate degradation. In the late decay stage, more complex fucoidan becomes the preferred substrate for degradation predominantly by epiphytic Verrucomicrobia (particularly Lentimonas), Pirellulaceae of Planctomycetes (particularly Rhodopirellula), Pontiellaceae of Kiritimatiellota, and Flavobacteriaceae of Bacteroidetes, which are abundant in glycoside hydrolases from the GH29, GH95, and GH141 families and sulfatases from the S1_15, S1_16, S1_17, and S1_25 families (Fig. 9). Many genera of the PVC group (e.g., Lentimonas, Pontiella, and Rhodopirellula) and Bacteroidetes (e.g., Zobellia, Formosa, and Wenyingzhuangia) associated with the late decaying kelp can produce diverse glycoside hydrolases and sulfatases involved in fucoidan degradation (Fig. 9; Data Set S3). Moreover, all the four MAGs from the PVC group, reconstructed from the late decay stage, possess the fucoidan utilization pathways (Fig. 8; Data Set S6), further suggesting the role of the PVC group in fucoidan degradation, especially in the late decay stage of S. japonica. In addition, four of five Bacteroidetes MAGs, reconstructed from early and late decay stages, harbor the pathways to utilize alginate, fucoidan, and laminarin (Fig. 8; Data Sets S5 to S7), suggesting their strong capacity to degrade brown algal polysaccharides.

Fig 9.

A proposed model for the degradation of brown algal polysaccharides by epiphytic microbiota during kelp decay. Symbols representing alginate-degrading bacterial groups associated with early decaying Saccharina japonica are scaled to the relative abundance (in parentheses) of corresponding bacterial groups harboring predicted extracellular alginate lyases. Sizes of symbols for fucoidan-degrading bacterial groups associated with late decaying S. japonica are proportional to the relative abundance (colored in black in parentheses) of corresponding bacterial groups containing predicted extracellular fucoidanases. For fucoidan-degrading bacterial groups, the relative abundance (colored in green in parentheses) of individual bacterial groups harboring predicted extracellular sulfatases is also shown. The distribution of extracellular algal polysaccharide-degrading enzymes in early and late decay stages is illustrated in Euler diagrams, wherein sizes of the enzyme symbols are scaled based on the relative abundances of corresponding enzyme genes in each sample and overlapping regions represent enzymes shared by certain epiphytic bacteria.

DISCUSSION

Kelps are important carbon sink in coastal marine ecosystems, providing substantial carbon sources mainly in the form of polysaccharides for the growth of marine bacteria (3). Kelp polysaccharides are complex mixtures of alginate, fucoidan, laminarin, and other brown algal polysaccharides. Epiphytic bacteria play important roles in the degradation and cycling of algal polysaccharides. Although bacterial groups involved in the degradation of alginate, fucoidan, or laminarin have been reported separately (6, 10, 13, 14, 21, 29), the dynamics of epiphytic bacterial communities and their roles in the degradation of mixed kelp polysaccharides during kelp decay remain unclear. Whether marine bacteria degrade these algal polysaccharides simultaneously or show polysaccharide preferences during kelp decay is also unknown by far. Here, based on metagenomic analyses, the succession of epiphytic bacterial communities and their metabolic potential for polysaccharide degradation were revealed during the early and late decay stages of the kelp S. japonica.

In this study, analyses of algal polysaccharide-degrading genes during the decay of S. japonica suggested that alginate tended to be degraded in the early decay stage predominantly by epiphytic gammaproteobacterial Vibrionaceae (particularly Vibrio) and Psychromonadaceae (particularly Psychromonas) (Fig. 2; Fig. S1). These bacterial groups were abundant in PL6, PL7, and PL17 alginate lyases, in some of which the alginate utilization pathways were found (Fig. 2, 7, and 8). Vibrionaceae and other gammaproteobacterial groups also dominated the early decaying kelp-associated bacterial communities (Fig. 1). Previous studies have shown that Gammaproteobacteria and Bacteroidetes are pivotal alginate degraders in seawater and marine sediments (9, 10, 16). Alginate-supplemented seawater microcosms showed that alginate induced the growth of the gammaproteobacterial Alteromonadaceae (Alteromonas) at the Patagonian continental shelf (9). In polar sediments, most of the alginate-degrading bacteria belonged to Gammaproteobacteria (Alteromonas and Pseudoalteromonas) and Bacteroidetes, which contained alginate lyase genes mainly from the PL6, PL7, and PL17 families (16). In temperate and subantarctic intertidal sediments amended with kelp detritus, the most abundant alginate lyase genes were distributed in the PL6, PL7, and PL17 families, primarily originating from Bacteroidetes, followed by Gammaproteobacteria (Psychromonas and Marinomonas) (10). Our results demonstrate that, despite similarities in the dominant families of alginate lyases between kelp-associated and seawater/sedimentary bacteria, the specific taxonomic groups of alginate-degrading Gammaproteobacteria associated with S. japonica differ from those found in seawater and sediments. Members of Vibrionaceae (particularly Vibrio) were found to be major contributors to the degradation of S. japonica-derived alginate. Notably, some pathogenic Vibrios have been reported to cause algal diseases (67), perhaps owing to their potential to degrade alginate or other algal polysaccharides to some degree.

During the decay of S. japonica, there was a shift in the dominant epiphytic bacterial communities from Gammaproteobacteria to the PVC group and Bacteroidetes (Fig. 1). Metagenomic analyses suggested that kelp-derived fucoidan was preferred to be degraded in the late decay stage by epiphytic bacteria, predominantly from the PVC group, followed by Bacteroidetes (Fig. 3 and 4). These bacteria were abundant in glycoside hydrolases from the GH29, GH95, and GH141 families and sulfatases from the S1_15, S1_16, S1_17, and S1_25 families, some of which possessed the fucoidan utilization pathways (Fig. 7 and 8). Metagenomic analysis of Undaria pinnatifida-amended sediments showed that the majority of detected fucoidanase genes belonged to families GH29 and GH95, primarily found in Bacteroidetes, followed by Planctomycetes (10). Different from those in U. pinnatifida-amended sediments (10), members of the PVC group, especially Verrucomicrobia, were main fucoidan degraders during the decay of S. japonica.

Until now, the dynamics in the degradation of kelp polysaccharides by marine bacteria during kelp decay remain unknown. When grown with kelp powder, the epiphytic Firmicutes Bacillus weihaiensis Alg07 was found to prioritize the degradation of alginate from cell walls before switching to the utilization of intracellular laminarin (68). Compared to alginate and laminarin, fucoidan is the most complex and resistant to microbial degradation (21). Metaproteomic analyses of TonB-dependent transporters during a diatom-dominated bloom showed that glucose-based labile algal molecules including laminarin and starch were used throughout the bloom, whereas complex cell wall polysaccharides containing fucose, mannose, and xylose were mostly utilized in later bloom stages (69). Extensive studies on diatom-dominated blooms in the North Sea also suggest that algal glycans can shape the community composition of bacterioplanktons during phytoplankton blooms (30, 70). Our study reveals a transition in algal polysaccharide-degrading bacteria during the decay of S. japonica, shifting from alginate-degrading Gammaproteobacteria to fucoidan-degrading Verrucomicrobia, Planctomycetes, Kiritimatiellota, and Bacteroidetes (Fig. 9). The observed prioritization of alginate over fucoidan and laminarin by epiphytic bacteria during the decay of S. japonica may be attributed to (i) the accessibility of polysaccharides in kelps to epiphytic bacteria, where alginate and fucoidan are cell wall polysaccharides, while laminarin is an intracellular polysaccharide; (ii) the different chemical structures of kelp polysaccharides, of which fucoidan is most complicated; (iii) the initial bacterial community on the kelp surface, in which alginate-degrading bacteria may be abundant; and/or (iv) the different growth strategies adopted by alginate-degrading and fucoidan-degrading bacterial groups (fast-growing γ-strategists or low-growing K-strategists?), which needs further study.

In summary, the succession of epiphytic bacterial communities and their roles in the degradation of brown algal polysaccharides during the decay of S. japonica are revealed. This study paves the way toward a better understanding of the role of epiphytic bacteria in the degradation of algal cell wall polysaccharides, a complex form of organic carbon in marine environments.

MATERIALS AND METHODS

Sampling and DNA extraction

Saccharina japonica blades in the early decay stage were collected from an intertidal zone of the Bohai Sea adjacent to Yantai, China (38.3°N, 120.8°E) in September 2019. All kelp blades were stored in a sterile sampling bag. Upon arrival at the laboratory, one metagenome sample (denoted as X1) from the blade surfaces was collected immediately. To investigate changes in epibiotic communities during kelp decay, another metagenome sample (designated as X2) was obtained from surfaces of obviously rotted S. japonica blades that had been stored at 10°C for 3 months. For each sample, blades were first rinsed with sterile seawater for 10 s to remove transient (non-host-associated) bacteria and then mixed with sterile seawater by vortex shaking for 5 min to collect kelp-associated bacteria through centrifugation. Surfaces of residual blades were further swabbed with a sterile swab to fully census other microbes adhering to the surface. Bacterial DNA was extracted using the E.Z.N.A. stool DNA Kit (Omega Bio-tek, Norcross, GA, U.S.) following the manufacturer’s instructions.

Metagenomic sequencing and assembly

Metagenomic shotgun sequencing was performed on the Illumina NovaSeq 6000 platform with pair-end 150 bp mode at Wuhan Onemore-tech Co., Ltd., China. Raw reads obtained from metagenome sequencing were filtered to remove adapter sequences and low-quality sequences using fastp version 0.20.0 (71). After filtering, clean reads were mapped to the genome of S. japonica (72) using BWA version 0.7.17 (73) to identify and remove reads originating from the kelp host. The average Q30 of each sample was above 91%, indicating that the trimmed data set was of high quality. High-quality reads were assembled into contigs using MEGAHIT version 1.1.2 (74) with kmer values ranging from 47 to 97. Contigs longer than 300 bp were retained for further analysis.

Gene prediction and taxonomic assignment

Gene prediction and translation were conducted using Prodigal v2.6.3 (75) with the metagenomic mode. Genes were clustered to remove redundant sequences using CD-HIT (76) with 90% sequence identity and 90% coverage. To estimate gene abundance, reads were mapped to the individual gene sequences using SOAPaligner version 2.21 (77) with 95% identity. Reads per kilobase per million values were calculated to normalize the relative abundance of individual genes in each sample. Genes were annotated using DIAMOND BLASTP (78) with an e-value cutoff of 10⁻⁵ against NCBI-nr, KEGG (Kyoto Encyclopedia of Genes and Genomes) (79), and Clusters of Orthologous Genes (80) databases for functional analysis, retaining only the best hits. Taxonomy assignments were performed according to the BLAST results from the NCBI-nr database.

Searching for algal polysaccharide-degrading genes in metagenomic data

To identify algal polysaccharide-degrading genes in the metagenomic data, all genes encoding carbohydrate-active enzymes (CAZymes) were determined according to dbCAN (81) and Pfam (82) analyses, and only genes featuring start codons ATG, TTG, and GTG were included for subsequent analyses. To identify laminarin-degrading enzymes, predicted CAZymes belonging to the GH1, GH3, GH5, GH9, GH16, and GH30 families were further blasted against characterized glucanases retrieved from the CAZy database (41) with a stringency of 30% sequence identity, 50% coverage, and an e-value cutoff of 10⁻⁵. To identify sulfatase genes in the metagenomic data, HMM (Hidden Markov model) searches were carried out against the SulfAtlas database (51) with an e-value cutoff of 10⁻¹⁰. Among the identified CAZymes and sulfatases, only enzymes predicted to possess Sec/SPI or Tat/SPI signal peptides by SignalP 5.0 (59) were considered as extracellular enzymes. In addition, to identify outer membrane transporters involved in the utilization of alginate, BLASTP analysis was performed against SusC sequences from reported bacteria with PULs targeting alginate (13, 44–46) and KdgM/N porins from alginate-degrading Vibrio strains (14, 47), and hits with e-value <10⁻⁵, sequence identity >30%, and coverage >50% were probed.

Genome binning

Genome reconstruction of microbes from metagenomic data was performed separately for each sample. Contigs obtained from the two metagenomes were binned using Maxbin 2.0 (83) and MetaBAT 2 (84) within the MetaWRAP binning module separately. The default of the minimum length of contigs for bin construction with Maxbin 2.0 and MetaBAT 2 were 1,000 and 1,500 bp, respectively. Refinement of MAGs was performed using the bin_refinement module of MetaWRAP (85). To improve the quality of bins, metagenomic reads were mapped to bins and reassembled with metaSPAdes (86) through the reassemble_bins module of MetaWRAP. Completeness and contamination of MAGs were assessed using CheckM version 1.0.3 (87), with MAGs meeting the criteria of over 70% completeness and lower than 6% contamination selected for further analysis. MAGs were named based on the following scheme: the prefix X1 or X2 corresponds to the metagenome assembly from which they were binned, while unique MAGs within each coassembly were distinguished by numerical identifiers (e.g., X1_bin3). The taxonomy classification of each bin was initially determined by CheckM and further confirmed using GTDB v.1.3.0 (88).

Metabolic reconstruction of MAGs

Genes in MAGs were predicted by Prodigal v2.6.3 (75). All the predicted genes were searched against NCBI-nr, KEGG (79), and CAZy (41) databases to identify genes involved in the transport and catabolism of brown algal poly-/oligo-/mono-saccharides. Predicted transporters/enzymes participating in the utilization of alginate were confirmed by blasting against known transporters/alginate-metabolizing enzymes in bacteria with PULs targeting alginate (13, 44–46), alginate-degrading Vibrio strains (14, 47), and the alphaproteobacterial strain Sphingomonas sp. A1 (14, 66) using an e-value cutoff of 10⁻⁵, a minimum query coverage of 50%, and a minimum sequence identity of 30%. To identify transporters involved in the utilization of fucoidan, BLASTP analysis was performed using a 30% identity threshold, 50% coverage, and an e-value cutoff of 10⁻⁵ against outer membrane SusC sequences from reported Bacteroidetes with PULs targeting fucoidan (45, 89) and inner membrane transporters such as FucP, TRAP, and ABC found in fucose-utilizing bacteria (26, 60). Predicted fucose-metabolizing enzymes were confirmed by blasting against classified enzymes related to reported bacterial fucose catabolic pathways including the phosphorylative (26, 62) and non-phosphorylative (60, 63, 90) pathways using a stringency of 30% identity, 50% coverage, and an e-value cutoff of 10⁻⁵. To uncover transporters involved in the utilization of laminarin, BLASTP analysis was performed against transporters from reported Bacteroidetes with PULs targeting laminarin (29, 31, 45) using a 30% identity threshold, 50% coverage, and an e-value cutoff of 10⁻⁵. The cellular location of identified proteins in MAGs was predicted according to PSORTb v3.0 (57) and CELLO v.2.5 (58) combined with SignalP 5.0 (59).

DATA AVAILABILITY

Metagenomic raw sequence data obtained in this study have been deposited in the Genome Sequence Archive in National Genomics Data Center (91, 92) and China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA: CRA011766), that are publicly accessible at https://ngdc.cncb.ac.cn/gsa. The whole MAG sequence data assembled in this study have been deposited in the Genome Warehouse in National Genomics Data Center (92, 93) and Beijing Institute of Genomics, Chinese Academy of Sciences/China National Center for Bioinformation, under accession numbers GWHDOGK00000000-GWHDOGT00000000, GWHDUBU00000000-GWHDUBZ00000000, GWHDUCA00000000-GWHDUCG00000000 and GWHDUCI00000000-GWHDUCQ00000000 that are publicly accessible at https://ngdc.cncb.ac.cn/gwh.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.02025-23.

Figs. S1 to S8, Table S1. aem.02025-23-s0001.pdf.

Fig. S1 (Heatmap showing detailed changes in the relative abundance [log2] of predicted intracellular alginate lyase genes and their related bacterial groups), Fig. S2 (Relative abundance of brown polysaccharide-degrading bacterial groups associated with S. japonica in the early and late decay stages), Fig. S3 (Shift in outer membrane transporter genes involved in the utilization of alginate and their related bacterial groups during kelp decay), Fig. S4 (Heatmap showing detailed changes in the relative abundance [log2] of predicted intracellular fucoidanase genes and their related bacterial groups), Fig. S5 (Heatmap showing detailed changes in the relative abundance [log2] of predicted intracellular sulfatase genes and their related bacterial groups), Fig. S6 (Heatmap showing detailed changes in the relative abundance [log2] of predicted intracellular laminarin-degrading genes and their related bacterial groups), Fig. S7 (Key laminarin-degrading genes in MAGs), Fig. S8 (Predicted degradation pathways for brown algal polysaccharides in epiphytic MAGs including X2_bin16 [a], X1_bin7 [b], X2_bin11 [c] and X2_bin45 [d]), and Table S1 (Characteristics of epiphytic metagenome-assembled genomes).

Additional experimental details. aem.02025-23-s0002.xlsx.

Data Set S1: Information on metagenomic data and assembly results.

Additional experimental details. aem.02025-23-s0003.xlsx.

Data Set S2: Information on the relative abundance (RPKM) of epiphytic bacterial groups containing predicted alginate lyase genes and outer membrane transporter genes involved in alginate utilization at the genus level.

Additional experimental details. aem.02025-23-s0004.xlsx.

Data Set S3: Information on the relative abundance (RPKM) of epiphytic bacterial gropus containing predicted fucoidanase and sulfatase genes at the genus level.

Additional experimental details. aem.02025-23-s0005.xlsx.

Data Set S4: Information on the relative abundance (RPKM) of epiphytic bacterial groups containing predicted laminarin-degrading genes at the genus level.

Additional experimental details. aem.02025-23-s0006.xlsx.

Data Set S5: Information of genes involved in the alginate utilization pathways in MAGs.

Additional experimental details. aem.02025-23-s0007.xlsx.

Data Set S6: Information of genes involved in the fucoidan utilization pathways in MAGs.

Additional experimental details. aem.02025-23-s0008.xlsx.

Data Set S7: Information of genes involved in the laminarin utilization pathways in MAGs.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.