ι-Carrageenan catabolism is initiated by key sulfatases in the marine bacterium Pseudoalteromonas haloplanktis LL1

ABSTRACT

Marine bacteria contribute substantially to cycle macroalgae polysaccharides in marine environments. Carrageenans are the primary cell wall polysaccharides of red macroalgae. The carrageenan catabolism mechanism and pathways are still largely unclear. Pseudoalteromonas is a representative bacterial genus that can utilize carrageenan. We previously isolated the strain Pseudoalteromonas haloplanktis LL1 that could grow on ι-carrageenan but produce no ι-carrageenase. Here, through a combination of bioinformatic, biochemical, and genetic analyses, we determined that P. haloplanktis LL1 processed a desulfurization-depolymerization sequential pathway for ι-carrageenan utilization, which was initiated by key sulfatases PhSulf1 and PhSulf2. PhSulf2 acted as an endo/exo-G4S (4-O-sulfation-β-D-galactopyranose) sulfatase, while PhSulf1 was identified as a novel endo-DA2S sulfatase that could function extracellularly. Because of the unique activity of PhSulf1 toward ι-carrageenan rather than oligosaccharides, P. haloplanktis LL1 was considered to have a distinct ι-carrageenan catabolic pathway compared to other known ι-carrageenan-degrading bacteria, which mainly employ multifunctional G4S sulfatases and exo-DA2S (2-O-sulfation-3,6-anhydro-α-D-galactopyranose) sulfatase for sulfate removal. Furthermore, we detected widespread occurrence of PhSulf1-encoding gene homologs in the global ocean, indicating the prevalence of such endo-acting DA2S sulfatases as well as the related ι-carrageenan catabolism pathway. This research provides valuable insights into the enzymatic processes involved in carrageenan catabolism within marine ecological systems.

IMPORTANCE

Carrageenan is a type of linear sulfated polysaccharide that plays a significant role in forming cell walls of marine algae and is found extensively distributed throughout the world’s oceans. To the best of our current knowledge, the ι-carrageenan catabolism in marine bacteria either follows the depolymerization-desulfurization sequential process initiated by ι-carrageenase or starts from the desulfurization step catalyzed by exo-acting sulfatases. In this study, we found that the marine bacterium Pseudoalteromonas haloplanktis LL1 processes a distinct pathway for ι-carrageenan catabolism employing a specific endo-acting DA2S-sulfatase PhSulf1 and a multifunctional G4S sulfatase PhSulf2. The unique PhSulf1 homologs appear to be widely present on a global scale, indicating the indispensable contribution of the marine bacteria containing the distinct ι-carrageenan catabolism pathway. Therefore, this study would significantly enrich our understanding of the molecular mechanisms underlying carrageenan utilization, providing valuable insights into the intricate roles of marine bacteria in polysaccharide cycling in marine environments.

INTRODUCTION

Macroalgae is the dominant primary producer in coastal ecosystems and also contributes substantially as carbon donors to deep sea and sediments (1, 2). Their biomass mainly consists of complex and diverse polysaccharides, such as agar, algin, and carrageenan. Understanding the mechanisms of polysaccharide degradation in marine heterotrophic bacteria is essential for deciphering carbon fluxes in marine ecosystems, as these microorganisms play a pivotal role in environmental carbon cycling (3, 4).

Carrageenan is a kind of linear sulfated polysaccharide that contributes as the main cell wall component of carrageenophyte marine red algae (Rhodophyta) (5). Structurally, the repeating disaccharide subunits of carrageenans composed of β-D-galactopyranose (G-units) and 4-linked α-D-galactopyranose (D-units) or 4-linked 3,6-anhydro-α-D-galactopyranose (DA-units), which are alternately linked by β−1,4- and α−1,3-glycosidic linkages (5). According to the position and number of sulfate esters groups (S) and the presence of 3,6-anhydro-α-D-galactose in the repetitive galactose units, carrageenans can be classified into κ-carrageenan (G4S-DA), ι-carrageenan (G4S-DA2S), α-carrageenan (G-DA2S), β-carrageenan (G-DA), and λ-carrageenan (G2S-D2S,6S) (6, 7). Such diversity of carrageenan means that its complete degradation requires a complex of degradation enzymes and regulatory processes in marine heterotrophic bacteria.

To utilize carrageenan as the carbon source, marine bacteria produce a series of proteins, including carrageenases and sulfatases, for carrageenan depolymerization, desulfurization, as well as other enzymes for the metabolism of 3,6-anhydro-D-galactose and D-galactose. The endo-acting κ-carrageenases (GH16) and ι-carrageenases (GH82) specifically cleave the β−1,4 glycosidic bonds (8, 9). Various carrageenan sulfatases act on the sulfate groups of oligo-carrageenan to free inorganic sulfuric acid (10–13). The genes encoding these proteins are usually adjacent and co-regulated in a region of Polysaccharide Utilization Loci, termed PUL (14–16). The first characterized carrageenan-specific polysaccharide utilization locus (CarPUL) was from the marine bacterium Zobellia galactanivorans Dsij^T (17). This strain first hydrolyzes κ- and ι-carrageenan into κ- and ι-neocarrageenan oligosaccharides using κ-carrageenase and ι-carrageenase, respectively. The sulfate group from G4S of κ-neocarrageenan oligosaccharides can be removed by G4S-sulfatase S1_7. G4S-sulfatase (S1_19) and DA2S-sulfatase (S1_17) are responsible for the removal of the sulfate groups from G4S and DA2S of ι-neocarrageenan oligosaccharides, respectively. The released β-neocarrageenan oligosaccharides without sulfation were finally conversed into D-galactose and 3,6-anhydro-D-galactose for metabolism by exo-α−3,6-anhydro-D-galactosidase (GH127 and GH129-like) and exo-β-galactosidase (GH2).

Pseudoalteromonas spp. are globally distributed marine-associated strains that can degrade carrageenan through diverse carrageenan-utilization pathways. For example, Podosphaera fuliginea PS47 can grow on both κ- and ι- carrageenan but harbors three κ- carrageenase and no ι-carrageenase encoding genes in its CarPUL (12). In vitro enzymatic analysis suggests that the conversion of ι-carrageenan to β-neocarrabiose is catalyzed by a G4S sulfatase S1_19A, an exo-acting DA2S sulfatase S1_NC, an exo-acting G4S sulfatase S1_19B, a GH16B enzyme, and a GH167 exo-glycosidase (12, 18). Pseudoalteromonas carrageenovora 9^T possesses both κ- and ι-carrageenases but cannot utilize κ- or ι-carrageenan as the sole carbon source due to the absence of exo-α−3,6-anhydro-D-galactosidase (19). Besides Pseudoalteromonas spp. and Z. galactanivorans, the CarPUL in Flavobacterium algicola has also been characterized, which contains a multifunctional G4S sulfatase (OUC-S1_19B) for the removal of G4S or G2S sulfate groups from κ-, ι-, and λ-neocarrageenan oligosaccharides (20). These findings suggest that the CarPULs exhibit high complexity and diversity, especially in different glycoside hydrolase (GH) families and different sulfatase subfamilies. Moreover, the functions of sulfatases from CarPULs are critical for the microbial utilization of various carrageenans (10).

We previously isolated a marine strain LL1 of Pseudoalteromonas that shows no ι-carrageenase activity but grows with κ- or ι-carrageenan as the carbon source (21, 22). The strain was previously considered Pseudoalteromonas porphyrae according to the 16S rRNA gene sequence and was further determined to be Pseudoalteromonas haloplanktis based on the average nucleotide identity analysis of the genome DNA. In this study, we performed an in-depth study on the strain LL1 to better illustrate the carrageenan metabolic versatility. Based on the bioinformatics analysis of the CarPUL genes, the biochemical characterization of the key sulfatases, and the construction of sulfatase-deleted mutants, we concluded that the strain LL1 employs sulfatases PhSulf1 and PhSulf2 to initiate the ι-carrageenan catabolism pathway. The worldwide distribution of PhSULF1 and PhSULF2 genes and transcripts was also investigated. These results provide new insights into the mechanisms of carrageenan utilization, especially for the sulfatase-initiated ι-carrageenan catabolism omitting ι-carrageenase, and will enhance our understanding of the ability of marine bacteria to effectively obtain energy and nutrients from complex polysaccharides, showcasing the versatility in the ecological context.

RESULTS

P. haloplanktis LL1 utilizes both κ-carrageenan and ι-carrageenan

The cell growth of P. haloplanktis LL1 in the medium containing κ-carrageenan, ι-carrageenan, or the monomer galactose as the carbon source was analyzed by determining the pellet proteins of the cultivated cells. P. haloplanktis LL1 exhibited enhanced cell growth in the presence of both types of carrageenan and galactose compared to the condition without carrageenan or galactose supplementation. No significant difference was observed toward different carrageenan types, indicating that P. porphyrae LL1 could efficiently utilize both κ-carrageenan and ι-carrageenan (Fig. 1A). The supernatants derived from three cultures were then used to determine the carrageenase activity against both κ-carrageenan and ι-carrageenan. LL1 was observed to produce elevated levels of κ-carrageenase activity to support its growth on κ-carrageenan, and under this condition, no ι-carrageenase was detected (Fig. 1B). However, slight κ-carrageenase activity and no ι-carrageenase activity were detected with ι-carrageenan as the carbon source.

Fig 1.

The cell growth and carrageenase production of P. haloplanktis LL1 toward κ-carrageenan and ι-carrageenan. P. haloplanktis LL1 was cultivated for 28 hours with κ-carrageenan (κ-Car), ι-carrageenan (ι-Car), or galactose (Gal) as the carbon source, and the medium without the addition of carrageenan and galactose (NC) was used as a control. The cell growth was determined by monitoring the pellet protein (A). The supernatants were used to determine extracellular carrageenase activity against κ-carrageenan and ι-carrageenan (B) and to detect the generation of oligosaccharides by thin-layer chromatography (TLC) analysis (C). ns, no significance.

Consistent with the observation of enzyme activity, thin-layer chromatography (TLC) analysis indicated that κ-carrageenan, rather than ι-carrageenan, was utilized to produce oligosaccharides (Fig. 1C). These results indicate that LL1 produced extracellular κ-carrageenase for κ-carrageenan hydrolysis but utilized ι-carrageenan without producing ι-carrageenase. Considering the similar cell growth of LL1 on κ-carrageenan and ι-carrageenan, the strain should employ a specific way for ι-carrageenan hydrolysis. It is also noteworthy that, among the three supernatants, the one relative to κ-carrageenan utilization exhibited the highest carrageenase activity, which was indicative of substrate-coupled regulation of the κ-carrageenase production (Fig. 1B).

The CarPUL of P. haloplanktis LL1 lacks ι-carrageenase genes but possesses various sulfatase genes

We further sequenced the genome of P. haloplanktis LL1 to investigate the specific catabolism mechanisms of κ/ι-carrageenan. The genome length is 5,042,532 bp consisting of a total of 4,281 protein-coding genes and 90 tRNAs with an average genomic GC content of 40.73% (Table S1). A genome survey revealed that P. haloplanktis LL1 contains a CarPUL, named PhCarPUL, harboring genes encoding glycoside hydrolases, sulfatases, transcriptional regulators, galactose metabolism enzymes, TonB-dependent receptors, and a sugar transporter (Table S2). PhCarPUL exhibits similarity to the recently identified CarPULs from P. fuliginea PS47 (12), P. carrageenovora 9^T (19), Z. galactanivorans (17), and F. algicola (20, 23) (Fig. 2A).

Fig 2.

Bioinformatics analysis of CarPUL gene clusters and sulfatases. (A) Schematic representation and comparison of CarPULs in reported carrageenan-degraders. The locus tags and annotations of the genes in the PhPUL cluster are listed in Table S2. (B) Phylogenetic analysis of PhSulf1, PhSulf2, and PhSulf3 of PhCarPUL using the Neighbor-Joining method with 500 bootstrap replications. The sulfatase subfamilies are highlighted with different colors according to the phylogeny. The sulfatases derived from the strain LL1 are marked with red dots. Light blue triangles indicate the strain with no ι-carrageenanase. Blue and white boxes indicate the presence or absence of corresponding sulfatase activity as exonuclease (Exo-)/endonuclease (Endo-) toward sulfate groups at G4S/ DA2S positions of κ-carrageenan (κ-Car)/ι-carrageenan (ι-Car), respectively.

Three glycoside hydrolase genes were identified in PhCarPUL (Fig. 2A). Specifically, LL1_GM000391 encodes a κ-carrageenase (PhCgk, GH16 subfamily GH16_17) as being characterized in our previous study (22, 24). LL1_GM000393 encodes a putative carrageenase (PhGH16, GH16 subfamily GH16_13) which exhibits 59% and 67% sequence similarity to known furcellaranase Ce387 from Colwellia echini A3^T (acting on heterozygous κ/β-carrageenan) and endo-α/β-carrageenase GH16B from P. fuliginea PS47 (acting on heterozygous κ/β-carrageenan and ι/α- carrageenan), respectively (Fig. S1). LL1_GM000414 encoded a putative β-galactosidase (PhGal, GH167) with a 57% amino acid sequence similarity to GH167 (EU509_08920) in P. fuliginea PS47 (Fig. S1), which is known for releasing β-NC2 (neo-carrabiose) from β-carrageen and/or hybrid carrageenans. Although most known CarPULs contain one to three genes to encode the GH82 family ι-carrageenase, we did not detect such genes in PhCarPUL. Besides the PhCarPUL cluster, we searched the genome of P. haloplanktis LL1 for GH genes from GH16, GH82, GH150, and GH167 families that were potentially relative to carrageenan degradation (12, 17, 20, 23). Only two GH16-encoding genes were identified to encode a β-glucanase (LL1_GM002501) and a β-agarase (LL1_GM003768) and were not considered to participate in carrageenan degradation in LL1. These results were consistent with the failed detection of ι-carrageenase activity for P. haloplanktis LL1 (Fig. 1). Therefore, similar to P. fuliginea PS47 (12, 18), P. haloplanktis LL1 might harbor a sulfatase-dependent pathway to initiate ι-carrageenan catabolism.

Based on phylogenetic and structural analyses (Fig. 2B; Fig. S2), three genes of PhCarPUL, LL1_GM000411, LL1_GM000412, and LL1_GM000396, potentially encode PhSulf1 as the 2S-ι-carrageenan sulfatase from S1_NC subfamily, PhSulf2 as the 4S-ι/κ-carrageenan sulfatase from S1_19A subfamily, and PhSulf3 as the exo-4S-κ-carrageenan sulfatase belonging to S1_19B subfamily, respectively. Unlike Z. galactanivorans Dsij^T and F. algicola, which possessed four and six sulfatases respectively, from S1_7, S1_17, and S1_19 subfamilies, as indicated in Fig. 2A, all of the Pseudoalteromonas CarPULs contained three conserved sulfatases, suggesting a distinctive carrageenan desulfurization mechanism in Pseudoalteromonas compared to other strains.

The hydrolysis of β-NC2 into D-galactose and 3,6-anhydro-D-galactose via the activity of an α−1,3-(3,6-anhydro)-D-galactosidase is a vital step in releasing energy from κ- and ι-carrageenan (12). As indicated in Fig. 2A, PhCarPUL lacks the 3,6-anhydro-D-galactosidase homologs of GH127 and GH129 families that are present in Z. galactanivorans and F. algicola. Nevertheless, LL1_GM000394 in PhCarPUL encodes the protein of 65% amino acid sequence similarity with a candidate α−1,3-(3,6-anhydro)-D-galactosidase (EU509_08875) from P. fuliginea PS47 (Fig. 2A; Table S2) (12), indicating the potential β-NC2 hydrolysis capability of P. haloplanktis LL1.

The gene expression in PhCarPUL is regulated in response to carrageenan substrates

To investigate the potential substrate-coupled regulation, the LL1 strain was cultivated on the ST media supplemented with κ-carrageenan, ι-carrageenan, or galactose as the carbon source. The transcription of the key genes within the PhCarPUL gene cluster was then analyzed through quantitative reverse transcription PCR (RT-qPCR) (Fig. 3). Generally, all selected genes showed expression with carrageenan as the carbon source but were rarely transcribed when galactose was used as the carbon source, indicating the expression of the PhCarPUL cluster was induced with the presence of carrageenan. The genes encoding the transcriptional factor PhAraC, the endo-α/β-carrageenase PhGH16, and the sulfatases PhSulf2 and PhSulf3 were detected to express when the strain was cultivated either κ- or ι-carrageenan (Fig. 3A through D). In contrast, high-level expression of the κ-carrageenase-encoding gene PhCGK was only detected with the presence of κ-carrageenan (Fig. 3E). This was consistent with the essential role of PhCgk in κ-carrageen hydrolysis. Interestingly, the transcription of PhSULF1 was significantly upregulated when ι-carrageenan was supplemented, indicating the critical role of PhSulf1 in ι-carrageenan catabolism (Fig. 3F). Together with the phenomenon that P. haloplanktis LL1 utilized ι-carrageenan effectively but contained no ι-carrageenase (Fig. 1 and 2A), the strain might employ PhSulf1, a putative 2S-ι-carrageenan sulfatase, to initiate the distinct ι-carrageenan degradation.

Fig 3.

Relative expression levels of several key genes in the carrageenan metabolism pathway of P. haloplanktis LL1. The strain was cultivated with κ-carrageenan (κ-Car), ι-carrageenan (ι-Car), or galactose (Gal) as the carbon source. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

PhSulf1 and PhSulf2 function differently in carrageenan desulfurization

P. haloplanktis LL1 contains no ι-carrageenase activity for ι-carrageenan degradation and thus must rely on efficient sulfatases for the DA2S and G4S desulfurization of ι-carrageenan. To address the roles of PhSulf1 and PhSulf2 in the utilization of ι-carrageenan as suggested by phylogenetic and structural analyses (Fig. 2B; Fig. S2), these two proteins were expressed and purified using Brevibacillus choshinensis expression system according to a previously developed method (24, 25). Based on the SDS-PAGE analysis, the indicated molecular weights of PhSulf1 and PhSulf2 were similar to their predicted values of 54.5 kDa and 52.0 kDa, respectively (Fig. 4A). The desulfurization activity toward κ-carrageenan and ι-carrageenan was then analyzed by measuring the change in the sulfate content of the substrates.

Fig 4.

Determination of the function and substrate specificity of PhSulf1 and PhSulf2. (A) The SDS-PAGE analysis of PhSulf1 or PhSulf2 proteins expressed and purified in B. choshinensis. The purified proteins were then used to determine the sulfate group content of κ-carrageenan (B) or ι-carrageenan (C). Three independent replicates were performed for the statistical analysis. **, P < 0.01. ns, no significance. (D) The function of PhSulf2 and PhSulf1 for the desulfurization of κ-(G4S-DA), β-(G-DA), ι-(G4S-DA2S), and α-(G-DA2S) carrageenan.

As indicated in Fig. 4B, the sulfate content of κ-carrageenan decreased from 21.6% to 12.4% after incubation with PhSulf2, while no activity against κ-carrageenan was detected for PhSulf1. Additionally, similar sulfate contents were detected for PhSulf1 treated or PhSulf1 and PhSulf2 co-treated κ-carrageenan. This result supported that PhSulf2, rather than PhSulf1, was an endo-G4S sulfatase toward κ-carrageenan. For ι-carrageenan (Fig. 4C), the sulfate content decreased from 36.5% to 27.3% and 26.5% after incubation with PhSulf1 and PhSulf2, respectively. Moreover, after the co-treatment with PhSulf1 and PhSulf2, the sulfate content of ι-carrageenan further decreased to 21.2%, which was significantly lower than those treated with PhSulf1 or PhSulf2 alone. Taken together, PhSulf1 was determined to be an endo-DA2S sulfatase toward ι-carrageenan and α-carrageenan to produce κ-carrageenan and β-carrageenan, respectively, while PhSulf2 exhibited as an endo-G4S sulfatase toward ι-carrageenan and κ-carrageenan (Fig. 4D). The endo-sulfatase activities of PhSulf1 and PhSulf2 would enable them to initiate the degradation of carrageenan without the pre-hydrolysis by carrageenases.

The exo-sulfatase activities of PhSulf1 and PhSulf2 were further analyzed by TLC using κ- and ι-carrageenan and the corresponding oligosaccharide derivates as the substrates (Fig. S3). As expected, we observed no significant band shift for ι-carrageenan oligosaccharides after incubation with PhSulf2. However, the bands referring to the κ-carrageenan disaccharide showed different positions compared to the standard neo-κ-carrabiose and neo-κ-carratetraose. This indicated that, besides endo-G4S sulfatase activity, PhSulf2 also had exo-G4S sulfatase activity toward κ-carrageenan. The incubation of both κ- and ι-carrageenan oligosaccharides with PhSulf1 resulted in consistent TLC mobility with the corresponding standards (Fig. S3), suggesting that PhSulf1 showed no exo-DA2S-sulfatase activity toward either κ- or ι-carrageenan.

PhSulf1 and PhSulf2 initiate the ι-carrageenan catabolism in P. haloplanktis LL1

To further investigate the in vivo function of PhSulf1 and PhSulf2 in carrageenan catabolism of P. haloplanktis LL1, we constructed the mutant strains ∆SULF1 and ∆SULF2 by deleting PhSULF1 and PhSULF2, respectively (Fig. S4), using a conjugation-based genetic manipulation system for Pseudoalteromonas (26). The relative cell growth of the mutant strains was determined with the growth level of the wild-type strain LL1 as 100% (Fig. 5). Considering the presence of polysaccharides in the medium, the total protein contents of the bacterial cells were used to evaluate the levels of cell growth. ∆SULF1 had little effect on the cell growth of κ-carrageenan relative to the wild-type strain. However, its growth on ι-carrageenan decreased significantly (Fig. 5A). Furthermore, ∆SULF2 almost abolished the cell growth on both κ- and ι-carrageenan (Fig. 5B). These results together with the in vitro sulfatase activity determination of PhSulf1 and PhSulf2 (Fig. 4B and C; Fig. S2) confirmed the essential roles of PhSulf2 and PhSulf1/PhSulf2 for κ- and ι-carrageenan utilization, respectively.

Fig 5.

The relative cell growth of P. haloplanktis LL1 and the sulfatase mutants ∆SULF1 and ∆SULF2. The strains were cultivated with κ-carrageenan (A), ι-carrageenan (B), or galactose (C) as the carbon source. The cell growth of LL1 under each condition was determined as 100% for calculation. The total protein contents of the cells were used to evaluate the levels of cell growth. (D) Changes in the sulfate group content of supplemented carrageenan during cultivation.

We further executed the specific location where PhSulf1 and PhSulf2 played roles in P. haloplanktis LL1 by analyzing the extracellular sulfatase activity of the cell culture every 24 hours (Fig. 5D). The sulfate content of κ-carrageenan remained at the initial level during cultivation, but the sulfate content of ι-carrageenan decreased gradually. This indicated that, although the protein location of PhSulf1 was predicted to be in the periplasm (Table S2), P. haloplanktis LL1 secreted PhSulf1, rather than PhSulf2, extracellularly. Thus, the removal of the DA2S sulfate group by PhSulf1 could take place both in the periplasmic space and outside the cell, while the desulfurization reaction by PhSulf2 toward G4S sulfate ester groups only occurred in the periplasmic space. Since the PhSulf1 catalyzed the desulfurization of ι-carrageenan extracellularly, the desulfurization-depolymerization sequential catabolism pathway in P. haloplanktis LL1 was further confirmed.

Based on the results, we concluded that P. haloplanktis LL1 employed different pathways for κ- and ι-carrageenan catabolism, especially for the initiation steps (Fig. 6). For ι-carrageenan, the strain processes a unique desulfurization-depolymerization sequential way triggered by the combined activity of PhSulf1 and PhSulf2. For κ-carrageenan metabolism, only PhSulf2 rather than PhSulf1 is involved, and the depolymerization and desulfurization processes can occur simultaneously.

Fig 6.

The proposed catabolic pathway of ι-carrageenan and κ-carrageenan in P. haloplanktis LL1.

Distribution of PhSULF1 and PhSULF2 across the global ocean

The geographical distribution analysis was carried out based on the metagenomes data set (OM_RGC_v2_metaG) and metatranscriptomes data set (OM_RGC_v2_metaT) associated with the marine environment from Tara Ocean database with a cut-off value of e⁻³⁰ (Fig. 7). Both PhSULF1 and PhSULF2 and their transcripts exhibited a broad distribution in marine bacteria in the Atlantic, Pacific, and Indian oceans, suggesting their widespread importance in marine bacteria and potential significance in global ocean carbon cycling. Additionally, we observed higher gene and transcript abundances of PhSULF2 compared to those of PhSULF1 (Fig. 7A vs B and C vs D), indicating the prevalence of 4S-ι/κ-carrageenan S1_19A sulfatases compared to 2S-ι-carrageenan S1_NC sulfatases for marine bacterial carrageenan degradation. We further detected that both PhSULF1 and PhSULF2 homologs showed similar or slightly higher relative abundance compared with other known S1_NC and S1_19A subfamily members, respectively (Table S3), indicating the environmental significance of the sulfatase PhSulf1 and PhSulf2.

Fig 7.

Geographical distribution of PhSULF1 and PhSULF2 in marine bacteria. Geographic distribution of genes (A and B) and transcripts (C and D) of PhSULF1 (A and C) and PhSULF2 (B and D) using a cut-off value of e⁻³⁰. The relative abundance at each station was represented by different sizes symbols. The maps were prepared using the online Tara Oceans data sets (https://tara-oceans.mio.osupytheas.fr/ocean-gene-atlas/).

DISCUSSION

The marine ecosystem is one of the most important ecological systems on Earth. Marine bacteria play significant roles in marine ecosystem dynamics and biogeochemical cycles (3, 4). The genus Pseudoalteromonas is a type of marine-specific bacteria that is widely distributed in different marine regions and exhibits high abundance. It is considered a representative carrageenan-degrading bacterial genus (27). Moreover, the growth phenotypes of Pseudoalteromonas strains display a wide range of diversity and complexity. For instance, P. carrageenovora 9^T can degrade both κ- and ι-carrageenan but lacks the ability to utilize them as the sole carbon source. Four strains of Pseudoalteromonas, namely P. fuliginea PS47, P. fuliginea PS2, P. distincta U2A, and Pseudoalteromonas sp. FUC4, have been reported to grow on both κ- and ι-carrageenan, which is only possible, however, when ι- or κ-carrageenan oligosaccharides were supplemented in the medium (12). In this study, P. haloplanktis LL1, in contrast, could utilize ι-carrageenan without any oligosaccharide supplement and showed similar growth efficiency compared to that on κ-carrageenan (Fig. 1A), indicating a distinct pathway for ι-carrageenan utilization in P. haloplanktis LL1.

Carrageenan metabolism mainly contains two processes: depolymerization and desulfurization (15, 17). Generally, the utilization of κ-carrageenan and ι-carrageenan follows a sequential order of depolymerization and desulfurization, based on prior studies on carrageenan-degrading strains such as Z. galactanivorans Dsij^T, C. echini A3^T, F. algicola, and P. carrageenovora 9^T (15, 17, 23). This means that the κ- and ι-carrageenan were first depolymerized by κ- and ι- carrageenases to generate oligosaccharides, which were further desulfurized by carrageenan-specific sulfatases for intracellular metabolism. P. haloplanktis LL1 contains the genes encoding the GH16 κ-carrageenase PhCgk and the exo-acting G4S-sulfatase PhSulf3 in PhCarPUL (Fig. 2A), indicating that this strain also employs such depolymerization-desulfurization sequential process for κ-carrageenan catabolism. However, P. haloplanktis LL1 has no ι-carrageenases (Fig. 1B, C and 2A) and shows significant differences in the ι-carrageenan catabolism pathway.

Another Pseudoalteromonas strain, P. fuliginea PS47 was also reported to utilize ι-carrageenan but contain no genes encoding ι-carrageenase (12). Based on the enzymatic data in the previous study, P. fuliginea PS47 was proposed to import highly polymerized ι-carrageenan into the periplasm and use the combined action of S1_19A endo-G4S ι-carrageenan sulfatase and a lipid-anchored GH16B to generate the oligosaccharides with various sulfurization patterns. The hybrid oligosaccharides were further desulfurized by two exo-acting DA2S sulfatase S1_NC and G4S sulfatase S1_19B to generate β-carrageenan oligosaccharides for further metabolism (12). Here, based on the bioinformatic (Fig. 2; Fig. S1), biochemical (Fig. 4), as well as genetic evidence (Fig. 3 and 5), we concluded that P. haloplanktis LL1 employed a unique desulfurization-depolymerization sequential process for the ι-carrageenan catabolism (Fig. 6). The sulfatases PhSulf1 and PhSulf2 play an indispensable role to initiate the catabolism pathway.

The desulfurization-depolymerization sequential process was previously defined as “pathway II” for carrageenan catabolism in a previous study (20). In this process, the desulfurization of ι-carrageenan is the prerequisite for subsequent depolymerization catalyzed by glycoside hydrolases (23). Thus, sulfatases are critical enzymes for carrageenan degradation, particularly when ι-carrageenases are absent. All the characterized sulfatases involved in carrageenan metabolism can be classed into five subfamilies, which are S1_7, S1_NC, S1_17, S1_19B, and S1_19A (Fig. 2B). In which, both S1_17 and S1_NC subfamilies are composed of sulfatases with DA2S desulfurization activities toward ι-carrageenan. Differently, S1_17 members are derived from strains containing ι-carrageenase (17, 20, 23), while S1_NC sulfatases are from strains producing no ι-carrageenase, such as P. fuliginea PS47 and P. haloplanktis LL1 (12). S1_7 and S1_19B subfamilies have the conserved function as exo-G4S-sulfatases toward κ-carrageenan (Exo-G4S-KC) and are responsible for the generation of β-carrageenan oligosaccharides without sulfation (12, 20, 23). S1_19A members are derived from bacterial strains with or without ι-carrageenase and show various G4S desulfurization activities toward κ- and ι- carrageenan (12, 19). Based on the phylogenetic analysis of sulfatases with different activities (Fig. 2B), all known S1_19A sulfatases and the PhSulf1 of subfamily S1_NC have endo-sulfatase activity toward ι-carrageenan, which is required for the initiation of carrageenan degradation independent on carrageenase. Furthermore, it is noteworthy that endo-sulfatases were also detected in strains that could produce ι-carrageenase, such as Z. galactanivorans and F. algicola (Fig. 2B), suggesting that these strains might employ desulfurization-depolymerization and depolymerization-desulfurization sequential processes simultaneously for ι-carrageenan catabolism.

By now, according to the known bacterial strains following the desulfurization-depolymerization sequential process, such as P. fuliginea PS47 and F. algicola, ι-carrageenan catabolism is initiated by the subfamily S1_19A sulfatases, PsS1_19A and OUC-S1_19B, respectively (12, 20). Depending on the endo-G4S-sulfatase activity of these sulfatases (Fig. 2B), the sulfate groups in the D-Gal residues of ι- carrageenan were removed to generate partially desulfurized ι/α- carrageenan hybrid. The products are then hydrolyzed into various types of oligosaccharides under the action of GH16 enzymes with ι/α- and κ/β-carrageenan hybrid activities, such as OUC-FaGH16A and OUC-FaGH16B from F. algicola (20) and GH16B from P. fuliginea PS47 (12). The obtained ι-carrageenan oligosaccharides are subsequently desulfurized by exo-acting DA2S sulfatases from S1_19B or S1_NC subfamilies before further metabolism (12, 15). In P. haloplanktis LL1, such desulfurization-depolymerization sequential process for ι-carrageenan degradation is also present and initiated by PhSulf2 (Fig. 6) because of its endo-G4S-sulfatase activity toward ι-carrageenan (Fig. 2B and 4D). However, unlike all these known DA2S sulfatases, P. haloplanktis LL1 possessed a novel sulfatase PhSulf1, which catalyzes the desulfurization of ι-carrageenan rather than oligosaccharides, suggesting that PhSulf1 acting as an endo-DA2S sulfatase rather than exo-sulfatase (Fig. 4; Fig. S3). Therefore, the ι-carrageenan catabolism pathway in P. haloplanktis LL1 contains distinct features. It can be initiated by a unique endo-acting DA2S sulfatase, generating ι/κ- carrageenan hybrid (Fig. 6). Despite the unique features, widespread occurrence of PhSULF1 homologs are detected, indicating the prevalence of ι-carrageenan degraders following such endo-acting DA2S sulfatase initiated catabolism pathway in various marine bacteria (Fig. 7) as well as its indispensable contribution in global ocean carbon cycling.

In contrast to the genome-wide distribution observed in Z. galactanivorans DsijT and F. algicola, our analysis revealed that all genes in the strain LL1 potentially associated with carrageenan depolymerization are located in the PhCarPUL cluster (Fig. 2A). Specifically, three genes, LL1_GM000391, LL1_GM000393, and LL1_GM000414, were identified to participate in carrageenan depolymerization, and they were all concentrated within PhCarPUL. As indicated in Fig. 6, after the desulfurization by PhSulf1 and PhSulf2, PhGH16 (LL1_GM00393) is predicted to act on carrageenan hybrid and produce carrageenan oligosaccharides as an endo-α/β-carrageenase (GH16_13) for further desulfurization by PhSulf3. β-NC2 (neo-carrabiose) is subsequently liberated from carrageenan oligosaccharides through the action of a putative β-galactosidase (PhGal and GH167) encoded by LL1_GM000391 and converted to D-galactose and 3,6-anhydro-D-galactose by a candidate α−1,3-(3,6-anhydro)-D-galactosidase (LL1_GM00394). It is noteworthy that such degradation process pattern is conserved across various strains of Pseudoalteromonas according to the CarPUL analysis (Fig. 2A). In contrast, in Z. galactanivorans DsijT and F. algicola from the Flavobacteriales order, exo-β-galactosidase (GH2) and the subsequent exo-α−3,6-anhydro-D-galactosidase (GH127 and GH129) have been determined to drive the depolymerization process (17, 20, 23). This further highlights the diversity of carrageenan metabolism processes in marine bacteria.

In addition, the ι-carrageenan desulfurization process in P. haloplanktis LL1 could occur outside the cell as evidenced by the decrease of carrageenan sulfate content during the cultivation in ι-carrageenan (Fig. 5D). Such extracellular desulfurization process by PhSulf1 would significantly contribute to the efficient ι-carrageenan utilization of the strain LL1. In addition, the strain LL1 could also provide desulfurized products for other strains lacking the sulfatases in the environment and play a role as a “sharing pioneer” rather than a selfish organism within heterotrophic bacterial communities (28). Therefore, although P. haloplanktis LL1 cannot produce ι-carrageenases, thanks to the specific sulfatases, it employs two distinct pathways for κ- and ι-carrageenan catabolism, which provides the strain the enhanced capability to acquire nutrients from the marine environment, leading to improved survival. The ability to utilize both hydrolyzed and desulfurized products expands the range of available carbon sources, thereby increasing the adaptability and competitive advantage of P. haloplanktis LL in marine ecosystems.

MATERIALS AND METHODS

Strains and cultivation

The strains used in this study are listed in Table 1. The P. haloplanktis LL1 strain (collection number 2E01119 at the Marine Microorganisms Culture Collection of China, MCCC) (21), originally isolated from the decayed seaweed collected from Yellow Sea, China, was cultivated at 28°C in 2216E medium (10.0 g/L of tryptone, 5.0 g/L of peptone, and 0.01 g/L of FePO₄) and ST medium (1.0 g/L of κ-/ι-carrageenan, 2.0 g/L of meat extract, 1.0 g/L of K₂HPO₄·3H₂O, 20.0 g/L of NaCl, 0.5 g/L of MgSO₄·7H₂O, 0.01 g/L of FePO₄, and pH 8.0). B. choshinensis (Takara Bio Inc., Dalian, China) was cultivated at 30°C in TMNm medium consisting of 40.0 g/L of glucose, 40.0 g/L of polypeptone, 5.0 g/L of meat extract, 2.0 g/L of yeast extract, 10.0 mg/L of FeSO₄·7H₂O, 10.0 mg/L of MnSO₄·4H₂O, 1.0 mg/L of ZnSO₄·7H₂O, 10.0 µg/mL of neomycin, and pH 7.0 (22). Escherichia coli JM109, Top10, and WM3064 (Takara Bio Inc., Dalian, China) were cultivated at 37°C in Luria-Bertani medium. One hundred microgram per milliliter ampicillin, 10 µg/mL neomycin, 50 µg/mL meso-2,6-diaminopimelic acid (DAP), and 25 µg/mL chloramphenicol were supplemented when necessary.

TABLE 1.

Bacterial strains and plasmids used in this study

Strains/plasmids	Relevant characteristic	Sources
Strains
E. coli
Top10	F- mcrAΔ(mrr-hsdRMS-mcrBC) φ80lacZ ΔM15 ΔlacX74 nupG araD139Δ(ara-leu)7697 galE15 galK16 rpsL(StrR) endA1 λ-	Takara Bio Inc
JM109	F- endA1 glnV44 thi-1 gyrA96 recA1 mcrB+Δ(lac-proAB) e14- (traD36 proAB+ lacIq lacZΔM15) hsdR17(rK- rK+)	Takara Bio Inc
WM3064	ThrB1004 pro thi rpsL hsdS lacZΔM15 RP4-1360 Δ(araBAD)567ΔdapA1341::(erm pir)	Gift of Xiaoxue Wang from the South China Sea Institute, Chinese Academy of Sciences
B. choshinensis
B. choshinensis	Wild-type strain	Takara Bio Inc
B. choshinensis-PhSULF1	B. choshinensis with the plasmid expressing PhSULF1	This work
B. choshinensis-PhSULF2	B. choshinensis with the plasmid expressing PhSULF2	This work
P. haloplanktis
P. haloplanktis LL1	Wild-type strain	MCCC
∆SULF1	P. haloplanktis LL1 with deleted PhSULF1	This work
∆SULF2	P. haloplanktis LL1 with deleted PhSULF2	This work
Plasmids
pNCMO2	Expression vector	Takara Bio Inc
pNCMO2-PhSULF1	pNCMO2 derivative for expressing PhSULF1	This work
pNCMO2-PhSULF2	pNCMO2 derivative for expressing PhSULF2	This work
pK18mobsacB-Cm	Suicide vector	Gift of Xiaoxue Wang from the South China Sea Institute, Chinese Academy of Sciences
pK18mobsacB-Cm-SULF1	pK18mobsacB-Cm derivative for deleting PhSULF1	This work
pK18mobsacB-Cm-SULF2	pK18mobsacB-Cm derivative for deleting PhSULF2	This work

Genome DNA isolation, sequencing, and annotation

The P. haloplanktis LL1 cells grown in 2216E medium at 28°C for 12 hours were transferred to ST medium and cultured for 24 hours. Then, the cells were harvested and used for genomic DNA extraction using a TIANamp bacteria DNA kit according to the manufacturer’s instructions (Tiangen Biotech, Beijing, China). The purity and concentration of the extracted genomic DNA were tested using agarose gel electrophoresis. The complete genome sequencing and assembly were performed using the Illumina HiSeq system (BGI Genomics, Beijing, China). Gene sequences were obtained using Glimmer 3.0 software. Gene annotation was carried out through Gene Ontology (29), Clusters of Orthologous Groups (COG) (30), Swiss-Prot (31), Kyoto Encyclopedia of Genes and Genomes (32, 33), and Carbohydrate-Active Enzymes Database (CAZy) (34).

Bioinformatic analysis

Genes that are potentially related to the carrageenan metabolism pathway were determined based on protein homology using the SulfAtlas database (http://abims.sb-roscoff.fr/sulfatlas/index.html) (35, 36), the CAZymes database (http://www.cazy.org/), and the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov). The conserved structural domains were predicted and analyzed using the CD-Search module of NCBI. The amino acid sequence alignment was performed by Clustal Omega and ESPript online websites. The structures of sulfatases were predicted using the AlphaFold server. CELLO v.2.5 (http://cello.life.nctu.edu.tw/) was used for protein localization prediction, and SignaIP 5.0 (http://services.healthtech.dtu.dk/) was used to predict signal peptides. The phylogenetic analyses of PhSulf1 and PhSulf2 were performed with MEGA11.0 (37) using the Neighbor Joining method with 500 bootstrap replications. The distribution of PhSULF1 and PhSULF2 genes and transcripts were analyzed in Tara Oceans metagenomes data set (OM_RGC_v2_metaG) and metatranscriptomes data set (OM_RGC_v2_metaT; https://tara-oceans.mio.osupytheas.fr/ocean-gene-atlas/) with the stringency of e⁻³⁰ (38, 39). The distribution abundances of the homologs of PhSULF1, EU509_08820, PhSULF2, EU509_08825, FA_2979, FA_3000, and ZGAL_3145 (Table S3) were analyzed following the normalization method as described in the user guide of the Ocean Gene Atlas (38), and the gene’s read coverage is divided by the median of the coverages of a set of 10 universal single-copy marker genes (COG0012, COG0016, COG0018, COG0172, COG0215, COG0495, COG0525, COG0533, COG0541, and COG0552) (40, 41).

Plasmid construction

The genes PhSULF1 (LL1_GM000411) and PhSULF2 (LL1_GM000412) encoding carrageenan sulfatases PhSulf1 and PhSulf2 were amplified from the genome of P. haloplanktis LL1 using the primer pairs of 411Bam/411Eco and 412Bam/412Eco, respectively. The signal-peptide sequences were not included. The obtained fragments were then digested with BamHI and EcoRI and ligated into the expression vector pNCMO2 (Takara Bio Inc., Dalian, China). The signal peptide of pNCMO2 was used for the secretory protein expression in B. choshinensis. The obtained plasmid pNCMO2-PhSULF1 and pNCMO2-PhSULF2 were used for the expression of PhSulf1 and PhSulf2, respectively.

The plasmids pK18mobsacB-Cm-SULF1 and pK18mobsacB-Cm-SULF2 were constructed for the inactivation of genes PhSULF1 (LL1_GM000411) and PhSULF2 (LL1_GM000412), respectively, based on the suicide vector pK18mobsacB-Cm (26). Approximately 1 kb regions upstream and downstream of PhSULF1 and PhSULF2 genes were amplified from the genomic DNA of P. haloplanktis LL1 to serve as homology arms for the knockout of 973 bp and 1,116 bp within the coding sequences of PhSULF1 and PhSULF2, respectively. The obtained fragments were subsequently ligated into pK18mobsacB-Cm using the ClonExpress II One Step Cloning Kit (Vazyme Biotech, Nanjing, China). All primers used in this study are shown in Table S4.

Heterologous expression and protein purification

The expression plasmids pNCMO2-PhSULF1 and pNCMO2-PhSULF2 were transformed into B. choshinensis according to the chemical transformation method previously reported (24). The transformation of empty plasmid was used as the negative control. The transformants of B. choshinensis confirmed by PCR and sequencing were cultivated in TMNm medium at 30°C for 4 days. Because the expressed sulfatases PhSulf1 and PhSulf2 were predicted to be secretory proteins with 6 × His Tag, we purified the proteins directly from the supernatants by Ni-affinity chromatography as previously described (24). The supernatants were obtained by centrifuging the culture at 8,000 rpm for 10 min. The expression and purity of PhSulf1 and PhSulf2 were verified by SDS-PAGE.

Genetic manipulation

The PhSulf1 and PhSulf2 encoding gene deletion mutants were constructed by conjugation-based knockout experiment as previously described (26). In detail, the suicide gene deletion plasmids pK18mobsacB-Cm-SULF1 and pK18mobsacB-Cm- SULF2 were first transformed into E. coli WM3064. The obtained transformants were used as the donor to transfer the suicide plasmids to the receiver P. haloplanktis LL1 via conjugation according to a published protocol (26). The chloramphenicol-resisting transformants obtained from the 2216E solid medium containing 25 µg/mL of chloramphenicol were identified as PhSulf1- or PhSulf2-deleted mutants. The transformants were further screened on the 2216E medium plate containing 2% sucrose and verified by PCR and sequencing using primers shown in Table S4.

RNA isolation and RT-qPCR analysis

P. haloplanktis LL1 was cultivated in 50 mL of 2216E medium at 28°C, 180 rpm for 24 h. The seed culture was inoculated into ST liquid medium supplemented with κ-carrageenan, ι-carrageenan, or galactose at an inoculum size of 7% (vol/vol). Following cultivation at 28°C, 180 rpm for 28 hours, bacterial cells were harvested for total RNA isolation using the Bacterial RNA Kit (OMEGA Bio-Tek, USA) according to the manufacturer’s introduction. The isolated RNA was monitored by agarose electrophoresis and purity analysis (A₂₆₀/A₂₈₀). The first strand cDNA synthesis was implemented using HiScript III RT SuperMix for qPCR (+gDNA wiper; Vazyme, China) according to the manufacturer’s introduction. The primers used for RT-­qPCR, as listed in Table S5, were synthesized by Beijing Tsingke Biotech Co., Ltd. and purified by high performance liquid chromatography (HPLC). RT-qPCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme, China) on a Rotor-Gene Q Real-time PCR Cycler (QIAGEN Hilden, Germany). The PCR was performed under the following conditions: 95°C for 15 min, then 40 cycles of 95°C for 5 s, 60°C for 30 s, and 72°C for 30 s. Each sample underwent four replicates. The melting curve was analyzed by Rotor-Gene Q 2.0.2 Real-time Data Acquisition and Analysis Software, and the Cq value was calculated. The relative expression of related genes was calculated according to the formula Rate = 2^-ΔΔCq. Although without experimental validation according to the Minimum Information for Publication Quantitative Real-Time PCR Experiments guidelines (42), we used the gene encoding the ρ-factor protein as the internal reference for calculation since it has been considered a relatively stable reference to normalize qPCR mRNA expression in various bacterial strains (43–46).

Analytic methods

Because the presence of polysaccharides severely influenced the analysis of the cell dry weight, the cell growth of P. haloplanktis LL1 was determined by quantifying the cell proteins. In detail, the cells grown in ST medium were centrifuged at 4°C, 5,000 rpm for 10 min. The cell pellets were suspended using 1 mL of pre-cooled phosphate-buffered saline (PBS) (pH 7.4) at 4°C and disrupted using a cell crusher. The obtained supernatants were then centrifuged at 4°C, 4,000 rpm for 5 min to separate the cell lysates, which were used to determine protein concentration following the Coomassie Brilliant Blue method (47).

Carrageenan oligosaccharides were obtained by incubating purchased κ-carrageenan and ι-carrageenan (Sigma-Aldrich) with κ- or ι-carrageenase at 40°C overnight. The carrageenases were purified according to our previous work (22, 25, 38). The sulfate content in κ-carrageenan and ι-carrageenan polysaccharides was determined following the Barium-chloride turbidimetric method (48). The change of sulfate content during the growth of P. haloplanktis LL1 was also performed using TLC. In detail, 0.3 µL of the samples was spotted on the silica gel plate three times with internals to allow the samples to be blown dry. Then, the layer was spread by a solvent consisting of n-butanol:formic acid:water [4:6:1, (vol/vol/vol)], and the orcinol-sulfuric acid reagent was rapidly splashed on the plate for color development. The silica gel plates were heated at 120°C for 3–5 min to reveal the oligosaccharide products (49).

To determine the activity of κ- and ι-carrageenase, the strain was cultured in ST medium at 28°C, 180 rpm for 28 h. The broth supernatants were then obtained by centrifuging at 4°C, 5,000 rpm for 10 min. The reaction mixtures containing 20 µL of the supernatants and 180 µL of a solution containing 0.5% (wt/vol) κ- or ι-carrageenan were incubated at 40°C for 15 min and were used to determine the production of reducing sugar based on the dinitrosalicylic acid (DNS) method (50).

DATA AVAILABILITY

The genome sequence of LL1 has been deposited to GenBank with the accession number JAVIFY000000000 (BioProject PRJNA1009502, BioSample SAMN37155428).

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.00255-24.

Supplemental material. aem.00255-24-s0001.docx.

Figures S1 to S4; Tables S1 to S5.

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.