Background

The genus Sulfitobacter is a member of the family Roseobacteraceae of the class Alphaproteobacteria. This genus was first proposed by Sorokin [1] describing two strains of heterotrophic bacteria with high sulfide oxidase activity isolated from the H2S-O2 interface of the Black Sea. As of March 10, 2023, the genus Sulfitobacter comprised 24 validated species according to NCBI taxonomy and LPSN (https://lpsn.dsmz.de/genus/sulfitobacter). Members of the family Roseobacteraceae constitute up to 20% of coastal marine bacterial populations [2], making this family one of the most abundant groups of marine bacteria. Sulfitobacter strains are abundant and widely distributed across diverse ocean habitats [3], including seawater [4, 5], sediment [6], tidal flat [7], starfish [8], seagrass [8], brown algae [9] and coral [10]. Besides oxidizing sulfite and thiosulfate, Sulfitobacter strains also have the capability to degrade diatom-derived dimethylsulfoniopropionate (DMSP), an organic sulfur-containing compound presents globally in very large amounts (109 tons or more per year) [11]. As a result, Sulfitobacter strains are considered significant contributors to the organic sulfur cycle in marine environments. Strains of Sulfitobacter also attract much attention as they could produce bioactive metabolites [12,13,14], accumulate tungsten [15], mitigate harmful algal blooms [16] and degrade hydrocarbon [17].

Taxonomy forms the fundamental basis for microbiology, and current microbial taxonomy relies solely on an approach known as polyphasic taxonomy [18]. Over the past several decades, 16S rRNA gene phylogeny has played a central role in polyphasic taxonomy. Recently, with advancements in next-generation whole-genome sequencing, more accurate genetic and phylogenetic methods such as phylogenomic analysis [19], average amino acid identity (AAI) [20], average nucleotide identity (ANI) [21], and digital DNA–DNA hybridization (dDDH) [22] have been adopted, significantly enhancing the accuracy of taxonomic assignments. As a result, many earlier taxonomic classifications have undergone re-evaluation and modification using genome-based analysis [23, 24]. It has been reported that compared to genome-based phylogenetic analyses, 16S rRNA gene-based phylogeny lacks the resolution necessary for proper phylogenetic reconstruction in Roseobacteraceae species [25]. In this study, the taxonomic status of the genus Sulfitobacter and the metabolism associated with organic sulfur cycling were explored based on genome analysis. The taxonomy of the genus Sulfitobacter has been re-evaluated, proposing its re-classified into Sulfitobacter sensu stricto, along with the establishment of a novel genus, Parasulfitobacter gen. nov. Our findings aim to offer a deeper insight into the genus Sulfitobacter, provide guidance for future taxonomic endeavors related to this genus, and mitigate potential inaccuracies in taxonomic classification.

Methods

16S rRNA gene and genome sequences collection

The 16S rRNA gene sequences of the type strains of the validated species within the genus Sulfitobacter and closely related species were downloaded from EzBioCloud database, and their accession numbers were shown in Fig. S1. Additionally, 18 genome sequences of type strains within the genus Sulfibobacter were downloaded from the NCBI GenBank assembly database. For a comprehensive phylogenomic analysis of the genus Sulfitobacter, 17 genome sequences of related type strains within the genera Roseobacter, Pseudosulfitobacter, Ruegeria, Roseivivax, Pelagimonas, Litorivita, Arenibacterium, Yoonia, Loktanella, and Brevirhabdus were obtained from the NCBI GenBank assembly database. The genome sequence of Hyphomonas polymorpha PS728T was also downloaded and employed as the outgroup in the phylogenomic analysis. Details regarding the genome sequence properties of the mentioned 36 type strains were presented in Table S1 within the supplemental material. To calculate AAI values between the genome sequences of S. algicola 1151T and the type strains of the type species within the genera belonging to the family Roseobacteraceae, 129 genome sequences of type strains of the type species within the genera belonging to family Roseobacteraceae were downloaded from the GenBank assembly database. The accession numbers of the genome sequences of these 129 type strains were listed in Table S2 in the supplemental material.

16S rRNA gene-based and genome-based phylogenetic analysis

Multiple sequence alignment of the obtained 16S rRNA gene sequences was conducted using the Muscle program [26] integrated in MEGA software version X [27]. A phylogenetic tree was then established utilizing the maximum-likelihood (ML) method with MEAG X. The selected substitution model for this analysis was Kimura 2-parameter and Gama Distributed with Invariant sites (K2 + G + I), and the tree supported topologies were evaluated through bootstrap values calculated based on 1000 replications. In an effort to comprehensively analyze the taxonomy of the genus Sulfitobacter, genome-based phylogenetic trees were reconstructed using three sets of sequences: the nucleotide sequence of an up-to-date bacterial core gene set (UBCG) [19] consisting of 92 genes, the amino acid sequence of UBCG, and the amino acid sequence of single-copy orthologous clusters (OCs) comprising 488 proteins. For the reconstruction of the phylogenomic tree based on the nucleotide sequence of UBCG, a codon-based alignment file (Additional file 2) was generated using a JAVA program [19] with the ‘-a codon’ option from the 36 genome sequences (Table S1). This file was utilized to construct a ML tree by PhyML 3.0 [28] with the selected substitution model being GTR, and the tree supported topologies were evaluated through bootstrap values calculated based on 100 replications. For the reconstruction of the phylogenomic tree based on the amino acid sequence of UBCG, an alignment file (Additional file 3) was generated using the same JAVA program [19] with the ‘-a aa’ option. The resulting file was utilized to construct a ML tree following the same steps as with the nucleotide sequence of UBCG except that the selected substitution model was LG. During reconstruction of the phylogenomic tree based on OCs, the amino acid sequences were identified by comparing whole protein sequences pairwise with the execution of Proteinortho version 6 [29] with the command ‘-e = 1e-5 -cov = 50 -identity = 50’. Subsequently, single-copy OCs were filtered using an in-house perl script (Additional file 4), and the resulting file (Additional file 5) was used to construct a ML tree following the same steps as with the amino acid sequence of UBCG.

Calculation of genome-based similarity indices for taxa delineation

AAI was computed using the CompareM (https://github.com/dparks1134/ CompareM) program with the parameters of 40% amino acid identity and 50% coverage length. Alignment fractions (AF) and genome-wide ANI (gANI) values were calculated through the Microbial Species Identifier (MiSI) method using ANIcalculator 2014 − 127, version 1.0 (https://ani.jgi.doe.gov/html/anicalculator.php) [30]. Percentage of conserved proteins (POCP) was calculated based on an approach described by Qin et al. [31]..

Comparative genome analysis

Annotated genome files obtained from the NCBI GenBank assembly database were manually reviewed to identify genes related to sulfite oxidation and DMSP degradation pathways. Functional annotation of Open Reading Frames (ORFs) was also conducted using the KEGG automatic annotation server (KASS v2.1, https://www.genome.jp/tools/kaas/) [32] with the KEGG database (http://www.genome.jp/kegg/). The functional annotated genes were categorized using KEGG orthology (KO) numbers. Preparation of the Venn diagram and identification of the core genomes were conducted using EVenn (http://www.ehbio.com/test/venn) [33]. Genes sharing KEGG orthologs in the genomes of all strains were classified as the core genome.

Analysis of the phenotypic characteristics

Phenotypic characteristics of the Sulfitobacter species were collected and reviewed from the original descriptions in various studies.

Results and discussions

Phylogenetic and phylogenomic analysis of the genus Sulfitobacter

In order to assess the effectiveness of 16S rRNA gene sequence-based phylogenetic reconstruction in the taxonomy of the Sulfitobacter species, a ML tree was reconstructed based on 16 S rRNA sequences of the type strains within the genus Sulfitobacter and closely related genera (Fig. S1). Within this ML tree, the genus Sulfitobacter appears paraphyletic due to the presence of type strains from the genera Roseobacter and Arenibacterium. Furthermore, the ML tree demonstrates inadequate bootstrap support for the majority of branches. This observation leads us to conclude that 16S rRNA gene sequences lack the resolution required for precise phylogenetic reconstruction within Sulfitobacter species.

It is accepted that the multigene-based phylogenomic approach is much more consistent and dependable, thus being the preferred method for inferring phylogenetic relationships among prokaryotes. In this investigation, genome-based phylogeny is used as the primary guideline for revisiting the taxonomic status of the genus Sulfitobacter. Phylogenomic trees were reconstructed using three sets of sequences: the amino acid sequence of OCs, the amino acid sequence of UBCG, and the nucleotide sequence of UBCG. The three phylogenomic trees (Figs. 1 and 2, and Fig. S2) display robust bootstrap support for the majority of the branches, suggesting that phylogenomic analysis should be more suitable for inferring relationships among Sulfitobacter species. All trees show that the majority of Sulfitobacter species, including the type species S. pontiacus DSM 10,014T, clustered together, except for S. algicola 1151T. In both the phylogenomic trees based on the amino acid sequences of OCs (Fig. 1) and UBCG (Fig. 2), S. algicola 1151T forms a distinct branch quite far away from other Sulfitobacter type strains. In the phylogenomic tree based on the nucleotide sequence of UBCG (Fig. S2), S. algicola 1151T forms a cluster with Pelagimonas varians DSM 23,678T and Litorivita pollutaquae FSX-11T with a very low bootstrap support value (7%), and this cluster is also quite far away from other Sulfitobacter species. The extended branch length indicates a distant genomic relationship between S. algicola 1151T and P. varians DSM 23,678T as well as L. pollutaquae FSX-11T. Based on these phylogenomic trees, it is suggested that S. algicola does not belong to the genus Sulfitobacter and should be reclassified into a novel genus. The phylogemonic trees also show that Pseudosulfitobacter pseudonitzschiae DSM 26,824T forms a separate cluster from the primary Sulfitobacter clade, supporting the proposal that this species does not belong to Sulfitobacter and should be reclassified into the novel genus Pseudosulfitobacter [34]. Moreover, the three phylogemonic trees demonstrate that genera such as Roseobacter, Ruegeria, Roseivivax and Yoonia are monophyletic, suggesting that these genera are well defined.

Fig. 1
figure 1

Maximum-likelihood phylogenetic tree based on the amino acid sequences of OCs consisting of 488 proteins of the type strains of validated species of the genus Sulfitobacter and members of closely related taxa whose genome sequences were available. H. polymorpha PS728T is used as an outgroup. S. algicola 1151T is shown in bold. Bootstrap percentages (> 70%) based on 100 replicates are shown at nodes. Bar, 0.1 substitutions per nucleotide position

Full size image
Fig. 2
figure 2

Maximum-likelihood phylogenetic tree based on the amino acid sequences of UBCG of the type strains of validated species of the genus Sulfitobacter and members of closely related taxa whose genome sequences were available. H. polymorpha PS728T is used as an outgroup. S. algicola 1151T is shown in bold. Bootstrap percentages (> 70%) based on 100 replicates are shown at nodes. Bar, 0.1 substitutions per nucleotide position

Full size image

Assessment of genome-based similarity indexes for genus

AAI is the most widely used genomic amino acid-level comparison for demarcating genera. In addition, percentage of conserved proteins (POCP), alignment fractions (AF) and genome-wide ANI (gANI) are also suggested to be used for genus delineation. As prokaryotic taxa display a continuum of AAI, POCP, AF and gANI values, the discrete boundaries for genus are difficult to define. Results of Luo et al. [20]. indicate that AAI values among members of related but different genera typically range between 60 and 80%, with a maximum not exceeding 85%. In a taxonomic study of species of the roseobacter group, Wirth et al. [35]. employed a gradient of AAI to delimit genera that is defined by two values: a minimum value (80%) below which separating species into different genera should be considered and a maximum value (85%) above which combining species into the same genus should be considered. Nicholson et al. [36]. proposed a similar strategy in using AAI to delimit genera: AAI between the type strain of one species and the type strain of the type species of one genus should be greater than 76% so as to assign this species to the same genus. In addition, the AAI among all type strains of one specific genus should be greater than 74%. AAI has been applied for delimiting genera in various prokaryotic families, including Flavobacteriaceae [37], Roseobacteraceae [35], Colwelliaceae [38] and Weeksellaceae [36]. It is recommended for application in other prokaryotic genera as well [35]. POCP has also been used to estimate evolutionary distance between two strains, with a cut-off for prokaryotic genera set at 50% [31]. However, many studies argued that this 50% POCP cut-off might be overly conservative [35]. Consequently, while POCP has been commonly used to estimate evolutionary distance, the 50% cut-off for genera is often not applied. In this study, POCP is employed for estimating evolutionary distance without applying the 50% cutoff. It is reported that a combination of gANI and AF between two genomes has been shown to accurately reflect genomic relatedness, aiding in the delineation of species or genus [30]. Barco et al. [39]. reported that the AF values of the estimated genus inflection points have a mean of 0.333, with a median of 0.349. Additionally, the gANI values of the estimated genus inflection points have a mean of 73.10%, with a median of 73.08%. In this study, AAI, POCP, gANI and AF were all used for analyzing the genomic relatedness of different species.

The AAI value between S. algicola 1151T and the type strain of the type species of Sulfitobacter, S. pontiacus DSM 10,014T, was calculated to be 68.0%, apparently below the 76.0% threshold. AAI values between S. algicola 1151T and the other 16 type strains of the genus Sulfitobacter ranged from 67.3 to 68.5%, again falling below the suggested 74.0% cutoff. Consequently, the AAI indexes support our proposal derived from phylogenomic analysis that S. algicola does not belong to Sulfitobacter. The AAI values between S. pontiacus DSM 10,014T and the other 16 type strains of the genus Sulfitobacter ranged from 73.5 to 92.4%, indicating a closer genomic relatedness with S. pontiacus DSM 10,014T than with S. algicola 1151T. Althrough AAI values between S. pontiacus DSM 10,014T and some of these 16 type strains were calculated to be lower than 76.0%, as the phylogenomic trees show that all of the selected type strains of Sulfitobacter other than S. algicola 1151T cluster together, we think there is no necessity to split these Sulfibacter species into different genera.

The POCP values between S. algicola 1151T and the selected 17 type strains of Sulfitobacter ranged from 51.8 to 57.1%, while the POCP values among the 17 type strains of Sulfitobacter were calculated to be between 61.9 and 84.9%. The relatively lower POCP values suggest that S. algicola 1151T exhibits reduced genomic relatedness with the 17 selected type strains of Sulfitobacter, supporting our proposal derived from phylogenomic analysis that S. algicola does not belong to Sulfitobacter.

The gANI and AF values between S. algicola 1151T and S. pontiacus DSM 10,014T were calculated to be 71.835% and 0.325, respectively, both lower than the reported genus inflection points of 73.08% and 0.333. The gANI and AF values between S. pontiacus DSM 10,014T and the 16 type strains of the genus Sulfitobacter were calculated to be between 74.17 and 87.19% and 0.495–0.81, higher than the reported genus inflection points. Consequently, gANI and AF values further support our proposal that S. algicola 1151T does not belong to the genus Sulfitobacter.

The phylogenomic analysis suggests that S. algicola 1151T should be reclassified into a novel genus. To further analyze this proposal, the available genomes of the type strains of the type species of the family Roseobacteraceae from GenBank assembly database were downloaded. Subsequently, the AAI values between all the selected type strains and S. algicola 1151T were calculated. These AAI values ranged from 55.62 to 69.52%, apparently lower than the 76% threshold, supporting the proposal derived from phylogenomic analysis that S. algicola should be reclassified into a novel genus.

Genomic and phenotypic features analysis

Genomic and phenotypic features have also been widely used in bacterial taxonomy. In this study, we carefully reviewed the genomic and phenotypic features of the Sulfitobacter species. The distinctive characteristics between S. algicola 1151T and the selected 17 type strains of Sulfitobacter are listed in Table 1. Significant differences were observed in features such as growth temperature, polar lipid composition, fatty acid compositions and genomic DNA G + C content. Specifically, the grow temperature of S. algicola 1151T is higher than that of the selected 17 type strains of Sulfitobacter. The proportion of C20:1ω7c in the fatty acid profile of S. algicola 1151T was measured to be 29.7%, whereas C20:1ω7c was not detected or the proportion was measured to be less than 0.5% in the fatty acid profiles of the selected 17 type strains of Sulfitobacter. The proportion of C18:0 in the fatty acid profile of S. algicola 1151T was measured to be 11.7%, whereas C18:0 was not detected or the proportion was measured to be less than 2.0% in the fatty acid profiles of the selected 17 type strains of Sulfitobacter. The proportion of summed feature 8 (C18:1ω7c and/or C18:1ω6c) in the fatty acid profile of S. algicola 1151T was measured to be 44.1%, whereas the proportions of summed feature 8 in the fatty acid profiles of the selected 17 type strains of Sulfitobacter ranged between 50.0 and 89.6%. The genomic DNA G + C content of S. algicola 1151T was measured to be 51.8 mol%, whereas the genomic DNA G + C contents of the selected 17 type strains of Sulfitobacter range from 54.7 mol% to 61.2 mol%. In summary, results of the genomic and phenotypic features analysis support our proposal that S. algicola does not belong to the genus Sulfitobacter.

Table 1 Characteristics that differentiate Sulfitobacter algicola 1151T and the selected 17 type strains of Sulfitobacter
Full size table

To investigate the metabolic features of S. algicola 1151T and the selected 17 type strains of Sulfitobacter, we conducted functional analyses utilizing KEGG database categories. The overall relative abundances of KEGG functional genes in both S. algicola 1151T and the selected 17 type strains of Sulfitobacter were found to be similar (Fig. S3). Carbohydrate metabolism, amino acid metabolism, membrane transport, transport and catabolism, and translation-associated genes exhibit high abundance in both S. algicola 1151T and the selected 17 type strains. The core genome of the selected 17 type strains of Sulfitobacter comprises 1090 genes (Fig. S4), with translation, amino acid metabolism and carbohydrate metabolism-associated genes being notably abundant in the core genome (Fig. S5).

Sulfur oxidation and DMSP degradation pathways analysis

Sulfur oxidation is a critical component of the Earth’s sulfur cycle. The sulfur element can exist in a variety of oxidation states ranging from − 2 to + 6, yielding various sulfur compounds such as thiosulfate (S2O32−), sulfite (SO32−), sulfide (S22−) and sulfate (SO32−). Reduced sulfur compounds are oxidized to sulfur or sulfate by a community of bacteria which are called sulfur-oxidizing bacteria (SOB). Within this bacterial community, various enzymes and proteins involved in sulfur oxidation have been discovered. A central sulfur oxidization pathway, known as the sulfur oxidization (Sox) system, possesses the capability to oxidize thiosulfate, sulfide, sulfite, and elemental sulfur to sulfate [40]. Additionally, some other enzymes participating in sulfur oxidation have also been identified. For sulfide oxidation, there are two related enzymes that belong to the disulfide reductase protein family: flavocytochrome c sulfide dehydrogenase (FCC) and sulfide: quinone reductase (SQR). Oxidation of sulfur to sulfite by the dissimilatory sulfite reductase (rDSR) system has been experimentally proven [41]. Regarding the oxidation of sulfite to sulfate, two pathways have been reported: one (SorAB) involves a sulfite dehydrogenase reducing cytochrome c, while the other (SoeABC) involves reducing a quinone [42].

It has been documented that certain Sulfitobacter species are positive for oxidizing reduced sulfur compounds [1, 43]. In our investigation, we searched the enzymes involved in the sulfur oxidation process in the genomes of Sulfitobacter species. Initially, our focus was on the sox gene cluster. It was found that the sox gene cluster could be identified in all the 17 type strains of Sulfitobacter species except for S. guttiformis KCTC 32,187T. Notably, the sox gene cluster could not be identified in S. algicola 1151T. The modular organization of the sox gene clusters varies among the selected type species (Fig. 3). All identified sox gene clusters comprise soxRSVWXYZABCD genes, with the presence of soxT, soxE, soxF, soxG and soxH in some of these clusters (Fig. 3, Table S3). Additionally, five genes (orf1-5) with unknown function were identified in certain sox gene clusters (Fig. 3, Table S3). In the sox gene cluster of S. brevis DSM 1143T, orf1, the product of which is annotated as heme-binding protein, is positioned between soxR and soxS. In the sox gene clusters of S. noctilucicola NB-77T and S. undariae DSM 102,234T, orf2, the product of which is annotated as DsrE family protein, occupies the space between soxB and soxC. In the sox gene clusters of S. noctilucicola NB-77T, S. noctilucae NB-68T, S. aestuariivivens TSTF-M16T and S. donghicola DSW-25T, orf3, orf4 and orf5, the products of which are annotated as DUF302 domain-containing protein, 5-aminolevulinate synthase and YeeE/YedE family protein, respectively, are situated between soxF and soxH. In the sox gene clusters of S. geojensis MM-124T and S. mediterraneus DSM 12,244T, orf3 and orf5 are situated between soxF and soxH. In the sox gene cluster of S. undariae DSM 102,234T, orf3 and orf4 are situated between soxF and soxH. In the sox gene cluster of S. marinus DSM 23,422T, orf3 is positioned between soxF and soxH. In the sox gene cluster of S. litoralis DSM 17,584T, orf3 is located between soxF and soxG. Lastly, in the sox gene cluster of S. pontiacus DSM 10,014T, orf3 is positioned downstream of soxD. The physiological functions of these five genes in sulfur oxidation warrant further research. In the sox gene clusters of S. dubius DSM 16,472T, S. profundi SAORIC-263T, S. indolifex HEL-45T, S. delicatus DSM 16,477T, S. maritimus S0837T and S. brevis DSM 1143T, certain genes, namely soxF and/or soxH, locate separately from the other genes of this gene cluster. Conversely, in the remaining selected Sulfitobacter species, all genes of the sox cluster are co-located (Fig. 3, Table S3).

Fig. 3
figure 3

The modular organization of sox gene cluster of the selected 17 type strains of Sulfitobacter and S. algicola 1151T. The five genes with unknown function (orf1-5) are marked by hollow arrows

Full size image

Regarding sulfide oxidation, the genes (soxEF) encoding flavocytochrome c sulfide dehydrogenase (FCC) were identified in six type strains of Sulfitobacter, namely S. noctilucae NB-68T, S. noctilucicola NB-77T, S. geojensis MM-124T, S. mediterraneus DSM 12,244T, S. aestuariivivens TSTF-M16T and S. donghicola DSW-25T, but not in the other 11 type strains of Sulfitobacter or S. algicola 1151T (Fig. 3, Table S3). The gene encoding sulfide: quinone reductase was not identified in the selected 17 type strains of Sulfitobacter or S. algicola 1151T. Concerning sulfite oxidation, three genes (soeABC) encoding a sulfite dehydrogenase were identified in the selected type strains of Sulfitobacter species, excluding S. pontiacus DSM 10,014T and S. litoralis DSM 17,584T (Table S4). Notably, soeABC were also identified in S. algicola 1151T. Genes encoding the dissimilatory sulfite reductase (rDSR) system that could be used for oxidizing sulfur to sulfite were not identified in any of the selected type strains. These findings suggest that nearly all Sulfitobacter species possess enzymes facilitating sulfite oxidation, which will offer energy for their survival.

The sulfonium compound DMSP is produced in the oceans at petagram levels mainly by marine phytoplankton, macroalgae and bacteria for its anti-stress functions [44, 45]. DMSP could be utilized as important sulfur and carbon sources by many bacteria, among which the Roseobacteraceae species and SAR11 clade are the most prominent members [46]. The genus Sulfitobacter belongs to Roseobacteraceae and it is reported that some Sulfitobacter strains are involved in DMSP degradation. For instance, Sulfitobacter sp. EE-36 possesses a DMSP lyase (DddL), which facilitates the conversion of DMSP into the gas dimethylsulfide (DMS) [47]. Moreover, it is reported that Sulfitobacter sp. D7 could consume and metabolize algal DMSP to produce high amounts of methanethiol, and DMSP could mediate the bacterial virulence of Sulfitobacter sp. D7 against an oceanic bloom-forming phytoplankter [48]. Therefore, our investigation delves into the genomes of Sulfitobacter species to identify enzymes involved in DMSP degradation.

It is reported that bacteria employ two pathways for DMSP decomposition [49]: the demethylation pathway and the cleavage pathway (Fig. S6). Our findings reveal that genes encoding enzymes involved in DMSP degradation could be detected in all of the selected 17 type strains of Sulfitobacter and S. algicola 1151T (Fig. 4, Table S5). Some strains possess all the complete two pathways while the other strains do not. In the demethylation pathway, S. mediterraneus DSM 12,244T, S. noctilucae NB-68T, S. noctilucicola NB-77T, S. guttiformis KCTC 32,187T and S. marinus DSM 23,422T were identified to have all the four genes encoding DmdA, DmdB, DmdC and DmdD responsible for degrading DMSP to acetaldehyde and methanethiol. However, the complete demethylation pathway was not identified in the other type strains of the selected type stains of Sulfitobacter species, as they lack either DmdD or DmdA (Fig. 4, Table S5). In the case of S. algicola 1151T, neither DmdA nor DmdB was identified.

Fig. 4
figure 4

The distribution of enzymes involved in DMSP degradation in the selected 17 type strains of Sulfitobacter and S. algicola 1151T. Filled cycle indicates that the corresponding enzyme is present and the opened cycle indicates that the corresponding enzyme is absent

Full size image

In the cleavage pathway, the gene encoding DddD, responsible for directly degrading DMSP to DMS and 3-hydroxypropionic acid (3-HP), was not identified in the selected 17 type strains of Sulfitobacter or S. algicola 1151T. However, genes encoding DddL/DddP, AcuK, DddA and DddC, capable of degrading DMSP to DMS and acetyl-CoA, were identified in 10 type strains of the genus Sulfitobacter, These include S. brevis DSM 1143T, S. aestuariivivens TSTF-M16T, S. geojensis MM-124T, S. mediterraneus DSM 12,244T, S. noctilucae NB-68T, S. noctilucicola NB-77T, S. donghicola DSW-25T, S. guttiformis KCTC 32,187T, S. undariae DSM 102,234T and S. litoralis DSM 17,584T. The complete cleavage pathway was not identified in the other selected type strains of the genus Sulfitobacter or S. algicola 1151T as they lack DddL/DddP or DddA (Fig. 4, Table S5). These findings suggest that members of the genus Sulfitobacter decompose DMSP in either cleavage pathway or demethylation pathway. Notably, seven type strains of the genus Sulfitobacter contain DddL (Fig. 4, Table S5), a membrane-associated DMSP lyase capable of breaking down DMSP into DMS and acrylate. This indicates that these strains may also employ DMSP degradation as a defense strategy by shifting the predation pressure to non-DddL-containing bacteria [50].

Conclusion

In this study, based on 18 publically available genomes labeled as type strains of Sulfitobacter, we delved into the taxonomic status of this genus and its involvement in organic sulfur cycling. Employing whole-genome phylogeny as a guideline, and supplementing it with pairwise genome comparisons, our study suggests that S. algicola should be reclassified into a novel genus, for which the name Parasulfitobacter gen. nov. is proposed. This proposal finds support in the analysis of genomic and phenotypic features. Employing such an approach ensures a consistent and reliable classification of the genus Sulfitobacter, a group of bacteria that is both abundant and widely distributed, garnering increasing interest in terms of organic sulfur cycling, bioactive metabolites and biotechnical investigations.

This study also highlights the widespread presence of the sox gene cluster in nearly all the type strains of Sulfitobacter species, indicating the potential of the majority of Sulfitobacter species to oxidize reduced sulfur compounds, thereby deriving energy for their survival. Furthermore, our findings reveal the identification of both the demethylation pathway and the cleavage pathway for degrading DMSP in many Sulfitobacter species. This suggests that these bacteria can utilize DMSP as important sulfur and carbon sources or employ it as a defense strategy. These insights contribute to our understanding of how Sulfitobacter species participate in global sulfur cycle.

Description of Parasulfitobacter gen. nov. and Parasulfitobacter algicola comb. nov. are shown in Table 2.

Table 2 Description of Parasulfitobacter gen. nov. and Parasulfitobacter algicola sp. nov
Full size table