https://orcid.org/0000-0003-2429-8772
Figures
Abstract
Interactions between bacterial microbiota and epibenthic species of the dinoflagellate Prorocentrum may define the onset and persistence of benthic harmful algal blooms (bHABs). Chemical ecological interactions within the dinoflagellate phycosphere potentially involve a complex variety of organic molecules, metabolites, and toxins, including undefined bioactive compounds. In this study, the bacterial diversity and core members of the dinoflagellate-associated microbiota were defined from 11 strains of three epibenthic Prorocentrum species, representing three geographically disjunct locations within Mexican coastal waters. Microbiota profiles in stable monoclonal Prorocentrum cultures were obtained by sequencing amplicons of the V3-V4 region of the 16S rRNA gene. Thirteen classes of bacteria were identified among dinoflagellate clones, where Alphaproteobacteria, Gammaproteobacteria, and Bacteroidia were consistently dominant. The bacterial community structure exhibited significantly different grouping by the location of origin of dinoflagellate clones. No significant diversity difference was found among free-living or unattached bacteria in the dinoflagellate culture medium (M) compared with those in closer association with the dinoflagellate host cells (H). Twelve taxa were defined as core members of the bacterial assemblage, representing the genera Algiphilus, Cohaesibacter, Labrenzia, Mameliella, Marinobacter, Marivita, Massilia, Muricauda, Roseitalea, and an unclassified member of the Rhodobacteraceae. The core members are inferred to significantly contribute to primary and secondary metabolic functions, but no direct correlation with dinoflagellate toxigenicity was apparent. Overall the bacterial profile and implied gene functionality indicated a suite of positive interactions, suggesting either mutualism or commensalism with the dinoflagellate. The further characterization and interpretation of specific gene functions and interactions between bacteria and dinoflagellates, such as epibenthic members of genus Prorocentrum, are key to understanding their role in toxigenesis and bHAB development.
Citation: Martínez-Mercado MA, Cembella AD, Sánchez-Castrejón E, Saavedra-Flores A, Galindo-Sánchez CE, Durán-Riveroll LM (2024) Functional diversity of bacterial microbiota associated with the toxigenic benthic dinoflagellate Prorocentrum. PLoS ONE 19(7): e0306108. https://doi.org/10.1371/journal.pone.0306108
Editor: Raffaella Casotti, Stazione Zoologica Anton Dohrn, ITALY
Received: February 22, 2024; Accepted: June 11, 2024; Published: July 16, 2024
Copyright: © 2024 Martínez-Mercado et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The archived bacterial sequencing dataset is publically accessible from the Sequencing Read Archive (NCBI-SRA) (https://www.ncbi.nlm.nih.gov/sra) under BioProject PRJNA905463.
Funding: This work received financing from Cátedras CONACyT Project 1009 and CONACyT Ciencia Básica A1-S-8616 awarded to LD-R. The contribution of ADC was supported financially within Topic 6: Marine and Polar Life; Subtopic 6.2—Adaptation of marine life of the Alfred-Wegener-Institut, Helmholtz Zentrum für Polar-und Meeresforschung under the program “Changing Earth, Sustaining our Future,” Helmholtz Gemeinschaft, Germany. Sequencing was funded by CICESE institutional project 682136, managed by CG-S. The funders played no role in the study design, data analysis, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Marine species of the dinoflagellate genus Prorocentrum Ehrenberg are globally distributed, although epibenthic ecotypes are more prominent in subtropical and tropical regions [1]. The genus Prorocentrum comprises >80 defined morphospecies [2], but many morphotypes, particularly among the benthic contingent, remain only provisionally classified within various species complexes [3]. Recent efforts to resolve taxonomic and phylogenetic affinities, e.g., within the P. lima species complex, have been directed to target sequences of the 18S or 28S ribosomal RNA (rRNA) gene and internal transcribed spacer (ITS) region [4, 5].
In shallow coastal waters, Prorocentrum cell aggregates live attached to macroalgae, seagrasses, or other substrates, where cell proliferation can cause benthic harmful algal blooms (bHABs) [1, 4]. Recent attention on the ecological and human health consequences of bHABs has heightened scientific and biomedical awareness of these phenomena and their consequences [6]. In the benthos or attached to floating or suspended substrates, these Prorocentrum aggregates may negatively affect marine microfauna grazing directly upon the surfaces or living in the benthos. There is also potential for transfer of diarrhetic shellfish toxins (DSTs) produced by certain epibenthic and epiphytic Prorocentrum species [4] from dislodged cells to the water column [7, 8]. The relative impact of such toxigenic blooms on toxin accumulation and species interactions in benthic food webs, however, remains poorly defined.
Epibenthic dinoflagellates are naturally exposed to associated bacteria and other microeukaryotes within two broad compartments—the free-living and substrate-attached communities, respectively. Chemical interactions can occur in the extracellular space surrounding the dinoflagellate cells, denoted as the phycosphere [9], into which metabolites and other dissolved organics can diffuse and be exchanged with the aqueous medium. Chemically-mediated interactions with bacteria within the phycosphere could be beneficial or detrimental to the epibenthic dinoflagellate, or even mutually beneficial for both the dinoflagellate and the microbiota. In the mutually beneficial case, these interactions may involve the excretion or leakage of dissolved organic matter and shedding of nutrient-rich particular matter by the dinoflagellate in reciprocal exchange for vitamins or organic nutrients generated by the bacterial assemblage.
In negative interaction scenarios, “toxins” or other allelochemicals generated by the dinoflagellate may limit parasitism, colonization, or reduce competition for inorganic nutrients from the bacterial or microeukaryotic component in the microbiota [10]. Many epibenthic Prorocentrum species produce a suite of bioactive polyketide metabolites [11], including the DSTs responsible for diarrhetic shellfish poisoning (DSP). Production of toxic polyketides or other allelochemicals by Prorocentrum may be a functionally adaptative response–a chemical warfare strategy–particularly effective for epibenthic dinoflagellates as a defense against predators, and bacterial and microeukaryotic competitors for attachment space and nutrients [12]. Knowledge of the strategic importance of the bacterial community in toxin production or regulation in benthic dinoflagellates is scarce. While the toxin functions are still unknown, they may involve profound interactions with associated bacteria within the phycosphere and upon epibenthic substrates. Elucidation of the bacterial community composition and functional diversity is, therefore, the first step in defining these interactions.
In early studies of bacterial interactions with dinoflagellates, diversity was determined by culturing co-occurring bacteria [13, 14] and then by characterization of morphological or metabolic criteria and growth tolerance regimes. Later characterization studies for bacterial species in dinoflagellate cultures have been based on nucleic acid sequencing of a genetic marker, e.g. one or several hypervariable regions of the 16S rRNA gene [15, 16]. By combining morphological and sequencing approaches, it has been possible to demonstrate the influence of co-cultured bacteria associated with P. lima on growth, and toxin production rate and toxin cell quota of the dinoflagellate [16]. Such culture-based approaches allow detailed descriptions of specific bacterial members of the consortium but are limited in scope and biased for microbiome diversity studies because they are restricted to those taxa that may be readily co-cultured under defined laboratory conditions.
An alternative culture-independent approach to bacterial diversity can be obtained by high-throughput sequencing (HTS) of hypervariable regions of the 16S rRNA gene. In this way, the microbiome of several microeukaryotes (“phytoplankton”) has been characterized within a broader context, including species of planktonic dinoflagellates, such as Margalefidinium (= Cochlodinium) polykrikoides [17] and Prorocentrum minimum [18], and toxigenic benthic Gambierdiscus balechii [19].
The goal of this study was to evaluate the bacterial community and core membership in xenic Prorocentrum cultures by HTS sequencing of the V3-V4 regions of the bacterial 16S rRNA gene. The composition, diversity, and structure of the bacterial communities associated with two operationally defined fractions were compared among 11 strains of epibenthic Prorocentrum of differing toxigenic capacity and from distinct geographical origins from the coasts of Mexico. The comparative aspects focused primarily on P. lima because P. hoffmannianum and P. rhathymum were each represented by only a single strain. Core members of the Prorocentrum bacterial community were defined from fractions directly associated with the host dinoflagellate cells. The potential contribution of the core bacterial contingent to functional diversity interactions was explored by metabolic inference analysis.
Materials and methods
Collection and culture of Prorocentrum strains
Clonal cultures of epibenthic Prorocentrum were initiated from field material collected from coastal sites in the southern Gulf of California (Bahía de La Paz, Baja California Sur), the central Gulf of Mexico (Veracruz Reef System, Veracruz) and the Caribbean coast of Mexico (Puerto Morelos, Quintana Roo) (Table 1). Epibenthic dinoflagellates were harvested from various substrates, including seagrasses (Tracheophyta), macroalgae (Rhodophyta, Ochrophyta, and Chlorophyta), and from a rope attached to a buoy. All substrate samples (<5 g) were collected from non-protected species, and from public areas for which permits are not required for Mexico. Nevertheless, collection by investigators from the lead institute (CICESE, Ensenada) followed standard careful field practices to minimize ecosystem impacts. Microbiological research in the laboratory did not involve human or animal subjects, or cultivation of pathogenic microbes, and thus no specific endorsement from a CICESE or CONAHCYT ethics panel was necessary. Field samples were processed to establish clonal dinoflagellate strains according to previously described methods [4]. The taxonomic assignments and species identification of the Prorocentrum strains for the present study (Table 1) were previously determined by morphological analysis by microscopy and sequence analysis of the ITS1-5.8S_rRNA-ITS2 and 28S large-subunit (LSU) rRNA gene operon, complemented with secondary structure modeling of ITS2 rRNA variation based on compensatory base changes [4].
In brief, for the current experiments, xenic monoclonal Prorocentrum cultures were grown in 60x15 mm plastic Petri plates on 50%-strength GSe growth medium [20], modified without soil extract supplement. Dinoflagellate cultures were incubated at 25 ± 1 °C on a 12:12 h light:dark cycle and under illumination of 50 μmol photons m-2 s-1. Exponentially growing Prorocentrum were transferred into 125 mL Erlenmeyer flasks for further growth on full-strength GSe medium under the same environmental conditions and harvested after 3 to 4 weeks.
Harvest and fractionation of dinoflagellate-associated bacteria
Dinoflagellate-associated bacteria were harvested from the Prorocentrum cells and growth medium by gentle sieving, with repeated washing with sterile seawater, through a sterile 20 μm-nylon mesh. Dinoflagellate cells with surface-attached and intracellular (e.g., endosymbiotic) bacteria, hereafter comprising the dinoflagellate host (H) microbiome (S1 Fig), were recovered from the upper mesh surface by gentle backwashing with sterile-filtered seawater into 15 mL centrifuge tubes. After centrifugation (524R Eppendorf, Hamburg, Germany) at 6,000 × g at 4 °C for 3 min, the supernatant was discarded, and the H-fraction pellets were transferred with sterile seawater into 2 mL microtubes, and re-centrifuged as above. Bacteria from the filtrate passing through the 20 μm-nylon mesh, containing free-living or loosely bound dinoflagellate-associated bacteria, were harvested as the medium (M)-microbiome (S1 Fig) by centrifugation (Megafuge 40R, ThermoScientific, Waltham, USA) (15,000 × g at 4 °C for 15 min) of 100 mL filtrate. After discarding the supernatants, the bacterial cell pellets of M-fractions were transferred into 2 mL microtubes. The pellets of H- and M-fractions were frozen at -20 °C until DNA extraction.
Amplification and sequencing of bacterial 16S rRNA gene
Total DNA was extracted with the DNeasy Power Soil Kit (QIAGEN, Hilden, Germany) following the manufacturer’s instructions, with quantification by NanoDrop 2000 spectrophotometer (Thermo Scientific, Waltham, MA, United States). Due to high biomass contribution of dinoflagellate cells to the H-fraction, approximately 100 mg of H-samples versus 10 mg of M-fractions were extracted for amplification of the 16S rRNA bacterial genes. Samples were homogenized in extraction solution provided in the DNA extraction kit (QIAGEN, Hilden, Germany) in a FastPrep-24® 5G instrument (MP Biomedicals, Santa Ana, CA, United States) over three cycles of 30 s at 6.0 m s-1; each cycle was followed by a 5 min incubation on ice to minimize thermal degradation.
DNA library preparation was based on the MiSeq 16S Metagenomic Sequencing Library Preparation B protocol (Illumina, San Diego, CA, United States), with modifications according to the respective primers. The hypervariable regions V3-V4 of the 16S rRNA gene were amplified using the primer sequences 341F/785R published by Klindworth et al. [21]. PCR reactions were performed in duplicate in a final volume of 20 μL reaction mix: 1X KAPA HiFi Buffer GC, 2.5 mM MgCl2, 0.3 mM KAPA dNTP Mix, 0.3 μM of each primer, 1 U of KAPA HiFi HotStart DNA Polymerase (KAPA Biosystems, Wilmington, United States) and 12.5 ng of total DNA. The thermal cycler program was: 3 min at 95 °C, 25 cycles of 30 s at 95 °C, 30 s at 55 °C, and 30 s at 72 °C, with a final extension of 5 min at 72 °C.
PCR products were visualized in 2% agarose gel and then purified with Agencourt AMPure XP paramagnetic beads (Beckman Coulter, Brea, CA, United States). The sequencing Illumina indexes were incorporated for dual indexing with the Nextera XT index kit according to the manufacturer’s instructions. The libraries were quantified with a Qubit™ dsDNA High Sensitivity assay in a Qubit® 3.0 fluorometer (Thermo Scientific, Waltham, MA, United States). Each library was normalized at 40 nM, and then pooled and diluted to 4 nM. Paired-end reads of 250 bp were generated on a MiSeq sequencing platform (Illumina, San Diego, CA, USA).
Sequencing data analysis
Sequencing reads were processed with the dada2 v.1.14 package [22] in R v3.6 [23] following the author’s recommendations and with the necessary parameter adjustments. Sequence processing began by removing primer sequences from both reads, followed by base quality filtering, error rate calculation, and dereplication of reads. Quality-controlled reads were then denoised with the dada2 algorithm. The consensus method was used to identify and remove chimeras. Afterward, successfully merged paired reads were interpreted to define Amplicon Sequence Variants (ASVs). ASV sequences were taxonomically assigned with the ‘assignTaxonomy’ function at a confidence threshold of 80 and the SILVA 138.1 database [24] formatted for dada2 (https://doi.org/10.5281/zenodo.4587955). Packages DECIPHER v.2.14.0 [25], and phangorn v.2.5.5 [26] were used to align ASV sequences based on an initial single-linkage guide tree constructed from distances on shared k-mers.
A neighbor-joining (NJ) tree was constructed as the starting point for Maximum Likelihood (ML) tree inference based on a GTR+I+G model. ASV count table, taxonomy assignment, and ML tree were combined in a phyloseq object (v.1.30.0) [27] for phylogenetic and ecological analysis in R. Sequence counts were randomly subsampled to the same depth (91,115 reads), corresponding to the minimum deep-sequencing among samples. UniFrac distances (weighted and unweighted) among samples were used to explore the bacterial community structure by Principal Coordinates Analysis. Evaluation of sample grouping by location of origin was tested by PERMANOVA (9,999 permutations; α = 0.05) and validated by analysis of the variance of the groups’ dispersions. Several alternative alpha diversity indexes—observed ASVs (richness), Shannon diversity index, and Faith’s phylogenetic distance—were compared between dinoflagellate host (H)- and medium (M)- bacterial communities by a Kolmogorov-Smirnov test (α = 0.05). Community composition was explored at the class level by summarizing sequence counts per sample and plotting taxa with relative abundance >0.1%. Shared genera between the H and M bacteria were visualized in a Venn diagram. Core members were defined as those ASVs present in >75% of the samples with a minimum relative abundance of 0.0001% per sample, as for the endosymbiotic dinoflagellate Symbiodinium [28].
Prediction of the presence and abundance of the Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolic pathways [29] from ASVs of the host microbiome was conducted with PICRUSt2 v.2.2.0_b stratified pipeline, using a Nearest Sequenced Taxon Index (NSTI) < 0.2 [30], by the phylogenetic placement of short sequences to identify ancestral states [31–33]. The inferred metabolic profiles are used to support and complement known functions of the core members by comparing their contribution versus the contribution of non-core members through a Mann-Whitney test (α = 0.05) with Bonferroni correction for multiple testing.
Analysis of dinoflagellate DST composition in relation with associated bacteria
The toxin composition and cell quota of DSTs of toxigenic strains of P. lima and one of P. hoffmannianum were determined from H-fraction samples extracted in 50% MeOH and analyzed by liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) by previously published methods [4, 16]. Identification of bacterial taxa whose abundance correlated with the toxin production was tested by Pearson correlation and corrected for multiple testing by False Discovery Rate (FDR< 0.05) with the microbiome package [34]. The possible association of metabolic functions related to DST production was restricted to the 10 toxigenic strains of P. lima and P. hoffmannianum, leaving out the non-toxigenic P. rhathymum (PA20) [4], represented by only a single strain.
Results
The microbiome associated with cultured strains of the dinoflagellate genus Prorocentrum among three species (P. lima, P. rhathymum, and P. hoffmannianum) from disjunct geographical regions and collected from different substrates revealed high molecular diversity in both the dinoflagellate host (H)- and medium (M)-fractions. High-throughput sequencing of the bacterial community targeted to the V3-V4 regions of the 16S rRNA gene generated an average of 15.4 x 104 reads for H-fractions and 14.1 x 104 reads for M-fractions. After quality control and taxonomic filtration, reads analysis yielded an average of 58 bacterial ASVs in H-fractions and 67 in M-fractions (S1 Table). The sequencing depth achieved was sufficient to obtain a good representation of the diversity of each sample, as observed by rarefaction curves approaching the asymptote (S2 Fig).
Composition of the bacterial community associated with Prorocentrum
A total of 14 classes from eight phyla were represented among the ASVs from the H- and M-fractions. At the class level, 10 were considered dominant as they were represented in relative ASV abundances higher than 0.1%. Alphaproteobacteria was the most abundant class (relative abundances of 47 to 93%) in both fractions (Fig 1); there was no significant difference in the dominance structure except for a switch in ranking order from the third to the fourth position of Gammaproteobacteria and Phycisphaerae between H- and M-fractions, respectively. The classes Bacteroidia and Gammaproteobacteria were present in all H- and M-fractions, but classes Phycisphaerae and Planctomycetes were relatively abundant (>5%) in only seven and eight samples, respectively (Fig 1). The least abundant ASV classes were represented by Actinobacteria, Bacilli, Desulforomonadia, and KD4-96.
Bacterial profiles were analyzed by sequencing of V3-V4 16S rRNA gene amplicons. Relative abundances per sample are shown at the class level for abundant bacterial taxa (>0.1%) in the fraction associated with the host (H) or free-living in the culture medium (M) of nine strains of P. lima and one strain each of P. rhathymum (PA20) and P. hoffmannianum (PA85). Location of dinoflagellate strain origin is indicated above in the figure and in Table 1.
An exploration of the taxonomic distribution at the genus level (S3 Fig) showed that the dominant classes were each represented by several genera, e.g. Labrenzia, Roseitalea, Cohaesibacter, Marivita, Thalassobaculum (among others) for Alphaproteobacteria; and Marinobacter, Massilia, Algiphilus, and generically unassigned-Salinisphaerales for Gammaproteobacteria. Bacteroidia were represented by only two genera: Muricauda and Phaeodactylibacter, with the former yielding more ASVs from both fractions among the dinoflagellate cultures. Seventy-three percent of genera (69 out of 95) were shared between the two bacterial fractions; only 20 genera (21%) were exclusively found in M-fractions and six genera (6%) only in H-fractions. The only non-toxigenic strain included in this study, P. rhathymum PA20, exhibited distinctive bacterial genera in higher abundance compared to all toxigenic Prorocentrum strains. In decreasing order of abundance, these genera were: SM1A02 (Phycisphaerales), Hoeflea (Alphaproteobacteria), Methylophaga (Gammaproteobacteria), and Roseicyclus (Alphaproteobacteria).
Core members of the bacterial community
The core members of the dinoflagellate-associated bacterial community were defined (as ASVs) by focusing only on the H-fractions. These H-bacteria are physically closer to the dinoflagellate cells, thereby establishing continuous and more intimate interactions than with free-living bacteria in the medium (M). Furthermore, there was no evidence from comparing H- and M-fractions of significant differences in compositional structure at the class level. Core membership defined by taking into account only the toxigenic Prorocentrum strains yielded 12 bacterial ASVs that met the criteria as core members (S4 Fig). Notably, eight core-ASVs were found in all samples.
The defined bacterial core members associated with Prorocentrum belonged to three classes (Alphaproteobacteria, Gammaproteobacteria, and Bacteroidia), seven families, and 10 genera (Fig 2). The core member assigned to genus Labrenzia had a consistently high abundance, except for Prorocentrum strain PA78. Other genera represented by ASVs with high abundance across samples were Roseitalea, Muricauda, and Marinobacter. There were also ASVs with low abundance but high occurrence (i.e., rare taxa), representing the genera Massilia (2 ASVs) and Cohaesibacter (1 ASV). The core-ASVs belonging to Algiphilus, Marivita, Mameliella, and unclassified-Rhodobacteraceae were absent from a few strains: ASVs associated with Algiphilus were not found in Prorocentrum PA2, Marivita was absent from PA26, whilst Mameliella ASVs were not detected from PA44 and PA61, nor Rhodobacteraceae from PA53 and PA85. The core member ASV_004 assigned as an unclassified-Rhodobacteraceae was the most abundant taxon in strain PA78, yielding a very distinctive pattern among core members. A cluster analysis, however, indicated no apparent grouping by location of origin based only on the abundance of core members (Fig 2).
Heatmap shows core ASVs and their associated taxonomy (at family and genus levels) per sample. Dendrogram of hierarchical cluster analysis of samples is shown at the top.
Potential metabolic functions were predicted with the ASV sequences of the H-fractions to explore the functional contribution of the core members. A reliable functional profile was inferred for all ten H-fractions, according to the NSTI-sample values in the range of 0.025 to 0.125 (S1 Table). In these profiles, the core members contribute significantly more to nine major metabolic pathways compared to the contribution of non-core members (S2 Table). The key functional subcategories significantly supported by the core members include basal metabolism of different primary carbon sources (e.g., carbohydrates, amino acids and associated cofactors and vitamins), and secondary metabolism (including biosynthesis of terpenoids and polyketides, and degradation of xenobiotics).
Differences in bacterial community diversity among Prorocentrum strains
The alpha-diversity and community structure of the bacterial community in the H- and M- fractions overlap. The values of alpha-diversity indices (Observed ASVs, Shannon index, and Faith’s phylogenetic distance) compared by the Kolmogorov-Smirnov test (α = 0.05) were not significantly different between the H- and M-fractions (Fig 3). The UniFrac distance indicates that the samples group in compartments by the presence of taxa (PERMANOVA unweighted, p = 0.024), but not depending on the relative abundance of these taxa (PERMANOVA: weighted, p = 0.49, Fig 4). Nevertheless, despite the fact that the Prorocentrum strains were established and maintained under identical conditions, the bacterial community associated with these dinoflagellate strains was significantly different among the locations of origin: Gulf of California, Gulf of Mexico, and the Caribbean (PERMANOVA: unweighted, p = 1e-04; weighted, p = 0.0103; Fig 4).
Violin plots show the distribution of alpha diversity indices (A) observed ASVs, (B) Shannon index and (C) phylogenetic distance calculated for the bacterial community directly associated with Prorocentrum (Host = H) and free-living in culture medium (Medium = M). Comparison between compartments was by a Kolmogorov-Smirnov test (α = 0.05).
Principal coordinates analysis (PCA) of (A) unweighted and (B) weighted UniFrac distances among bacterial communities from Prorocentrum strains. Differences in community associated with the location of origin (shown by colors) were found by PERMANOVA test (α = 0.05, permutations = 9,999). Shapes indicate compartments associated directly with the host (H, circles) or free-living in the culture medium (M, triangles).
Effects of associated bacteria on toxigenicity of Prorocentrum strains
The linear polyketide okadaic acid (OA) and its respective diol-ester (OA-D8) tended to be the dominant DST analogs among all 10 toxigenic Prorocentrum strains considered for the bacterial community analysis herein. These two analogs typically comprised >50% of the total DST composition on a molar basis; other minor derivatives were present but varied widely among strains (S5 Fig and S1 Table). These minor variants included dinophysistoxin 1 (DTX1), its respective diol-ester (DTX1-D8), and two structurally undefined analogs, DTX1a and DTX1a-D8, in minor proportions. Strain P. lima PA53 from the Gulf of California produced mostly OA (about 75%), whereas PA49 cells from the same location yielded a similar proportion of OA and OA-D8 (<50%). In contrast, strain PA78 from the Gulf of Mexico produced mainly DTX1 (50%) but among all strains was the only one expressing such a relatively high DTX1 content.
The relationship between the relative abundance patterns of the bacterial community and DST production of dinoflagellate strains was evaluated by Pearson correlation analysis to identify bacterial taxa associated with the toxigenicity of studied strains. The members of the H-fractions showed a significant correlation with two of six DST analogs produced by the analyzed Prorocentrum strains (Fig 5): dinophysistoxins DTX1a and DTX1a-D8 showed a significant positive correlation (Pearson correlation r with FDR correction <0.05) with a set of 15 and 16 ASVs, respectively. Among these, only ASV_004 (an unclassified Rhodobacteraceae) belongs to the core members. Considering the taxa assigned to correlate ASVs with DST composition, four other genera were also significantly related to the core group: Labrenzia with DTX1a; and Marivita, Muricauda, and Roseitalea with DTX1a-D8. Of note, no significant correlations were found between members of the bacterial community and the other DSTs produced by Prorocentrum (OA, OA-D8, DTX1, DTX1-D8), and negative correlations were not significant.
Pearson correlation test with false discovery rate correction (FDR< 0.05) was calculated between the relative ASV abundance of H-associated bacteria and the relative compositional abundance (fmol cell-1) of DSTs produced by ten Prorocentrum strains. For clarity, the heatmap shows the correlations only for significant ASVs (*), but for all six DSTs determined among the strains: OA, okadaic acid; OA-D8, okadaic acid diol-ester; DTX1, dinophysistoxin 1; DTX1-D8, dinophysistoxin 1 diol-ester; DTX1a, undescribed DTX1 isomer; DTX1a-D8, undescribed DTX1a isomer.
Discussion
Composition of the bacterial community
Alphaproteobacteria dominated the bacterial community of the Prorocentrum strains, and was the dominant class among all samples of both H- and M-compartments analyzed. Alphaproteobacteria, together with Bacteroidia and Gammaproteobacteria are also the dominant lineages of microbial communities in natural planktonic phytoplankton blooms [35], and are the most common classes represented in the microbiome of benthic dinoflagellates [16, 19, 36]. Other classes with abundance >0.1% identified as part of Prorocentrum microbial community are worthy of future research because most of them (Phycisphaerae, Planctomycetes, Polyangia, Myxococcia, and Rhodothermia) occur in similar or even higher proportions in the microbiota of other epibenthic dinoflagellates, such as Gambierdiscus australes, G. balechii, G. caribeaus, G. carolinianus, or Ostreopsis ovata [19, 37–39]. These bacterial classes are frequently found as microbial community components of a wide variety of cultured Symbiodiniaceae–most often parasitic or endosymbiotic dinoflagellates known colloquially as members of the zooxanthellae (e.g., of corals) [40]. To our knowledge, the two exceptions to this apparent cosmopolitan distribution would be the class OM190 (phylum Planctomycetota), found only associated with cultures of various species of the dinoflagellate Alexandrium [41] and phaeophytes (“kelp”, e.g., Laminaria hyperborea) [42], and Kapabacteria (phylum Ca. Kapabacteria). The latter taxon (previously OPB56) is an extremophile whose mixed-cultured member was obtained from a bioreactor processing mining wastewater [43] and is not typically found in shallow-water marine benthic ecosystems.
Roseobacter was previously found as the dominant lineage at the genus level associated with benthic P. lima by bacterial plate isolation techniques [13, 14], and by Denaturing Gradient Gel Electrophoresis (DGGE) with the planktonic species P. minimum [18]. In the current HTS-bacterial community analysis, the taxa closest to Roseobacter were unclassified members of the parental Rhodobacteraceae family, also in agreement with a previous study by our group [16] based on a different set of Prorocentrum strains (except for PA2 and PA26) from the Gulf of Mexico and the Caribbean. The bacterial taxa recovered through the culturing approach in the latter study were not always among the most abundant as presented herein from HTS. For example, the highly abundant core member of the genus Labrenzia, commonly associated with dinoflagellates [19, 28, 38] and other eukaryotic microalgae such as diatoms [44], was not recovered in the aforementioned P. lima study. On the other hand, the genus Marinobacter was indeed identified by both approaches, and even as a core member in the present HTS study. In fact, the four classes identified in the previous study [16] (Actinobacteria, Flavobacteriia [herein reported as order Flavobacteriales], Alphaproteobacteria, and Gammaproteobacteria) were effectively recovered here, but along with 10 other classes, reflecting the broader scope of HTS.
Composition and diversity of bacterial communities in compartments
In xenic cultures of benthic dinoflagellates, bacteria can inhabit two broad compartments: free-living or aggregated in the culture medium (M) or directly associated with the dinoflagellate host (H). The bacteria in H-fractions may be attached to the dinoflagellate cell surface, occur as endosymbionts, or remain bound to mucus within the phycosphere. These compartments promote different lifestyle strategies, and thus, in a natural setting, they are typically inhabited by different bacteria [45]. A deeper analysis of this relationship focused, therefore, on the host associated bacterial community (H) of toxigenic strains, as the interactions are expected to be continuous and direct, with an expected significant impact on the ecophysiology of the dinoflagellate host.
The bacterial community of the Prorocentrum strains was dominated by Alphaproteobacteria, and to a lesser extent, Bacteroidia and Gammaproteobacteria, in both H- and M-fractions. Among the set of 11 Prorocentrum strains, the bacterial diversity and composition overlapped among clones and was only differentiated by a few genera each in both the H- and M-fractions. Major differences in the bacterial community between these fractions were found in HTS-based studies of P. minimum [18] and a metagenomic analysis on dynamics of free-living and attached bacterial assemblages in the centric diatom Skeletonema sp. [46]. In both cases, however, the microbiomes were associated with planktonic species and may be expected to differ in structure and composition of their respective compartments because of less spatially restricted microenvironments. Culture-based studies of the microbiome associated with strains of benthic Prorocentrum species [14, 16, 47], however, have shown dramatic differences in composition and structure of these compartments. The inherent selectivity of culture-based methods, i.e. the unculturable microbial majority [48], may be the cause for the apparent discrepancies between HTS- and culture-based approaches to define benthic dinoflagellate microbiomes.
The alpha- and beta- diversity overlap between bacteria in the culture medium (M-fraction) and host associated (H-fraction) may arise from an expanded but more loosely defined phycosphere created under stable culture conditions. The relatively limited space in an enclosed culture vessel can result in a more homogeneous distribution of bioactive compounds and nutrients than otherwise would be available by turbulence and diffusion from the water column or from benthic substrates in the natural environment. During different growth stages, specific types of organic molecules are released by dinoflagellate cells through leakage and excretion [18, 49], and, presumably, these metabolites reach equilibrium within the phycosphere in dinoflagellate cultures. Cultured benthic dinoflagellates also frequently produce abundant polysaccharide (mucus) surrounding the cell, serving as an organic substrate for bacterial growth. In support of this interpretation, the bacteria in the M-fractions of P. minimum were reported to be considerably more affected by changes in the growth stage of the host than those in the H-fractions [18]. In such cases, bacteria do not require direct interaction with the dinoflagellate to obtain certain nutrients or receive allelochemical signals, but retain the option to interact directly with the dinoflagellate for very specific needs within defined timeframes.
Functional dependence of locations of origin and microenvironmental selection
Factors such as specific location of origin and time of collection related to seasonal or other environmental cycles are considered to play an important role in the composition and structure of the microbiome developed in eukaryotic microalgal culture [50]. Differences in bacterial community composition related to the location of origin were also observed in the current study, but reflected only as crude biogeographical profile. Communities of marine bacteria vary latitudinally [51] and seasonally [52], yet the microbiome associated to microalgae has been determined to be conservative and species-specific [53, 54]. What may be true for phytoplanktonic assemblages that are not spatially restricted may not be valid for microbiomes associated with benthic dinoflagellates, which reflect a more stable microenvironment. For example, two benthic/epiphytic dinoflagellate species (Gambierdiscus australes and Ostreopsis cf. ovata) collected and isolated from same beach presented different bacterial profiles [38], but several strains of the planktonic dinoflagellate Alexandrium ostenfeldii collected from diverse locations in the Baltic Sea developed a similar bacterial profile in culture [41]. It is not surprising, therefore, that benthic Prorocentrum strains of two disjunct locations, i.e. North East Pacific Ocean and the Gulf of Mexico, exhibit a generally different bacterial community structure, but with enough common taxa to define the core members.
The bacterial compositional profile is typically reduced in diversity by selection and adaptation processes during co-culture with dinoflagellates. The “genetics of survivors” issue must also be considered from the temporal perspective–as a function of time in culture. Nevertheless, long-term cultures (since 1994) of Ostreococcus tauri (Chlorophyta) have exhibited the same dominant bacterial taxa (Marinobacter) through the different culture stages of the alga, as well as when the strains was grown under different conditions in alternative media [55]. While we are aware of the time-series selection issue in dinoflagellate culture, a relatively stable bacterial community will arise under long term growth in homogeneous conditions. Previous analyzes of the microbiome throughout the culture adaptation process in diatoms and other dinoflagellates have found that once the microalgal culture is established, the companion microbiota is stable as well [44, 50, 54, 56]. In fact, an increase in microbiome evenness was observed in the culture of diatoms over propagation time due to the reduction of the dominant taxa [44].
In the current experiments with Prorocentrum, cultures were maintained under uniform standardized environmental conditions designed to optimize dinoflagellate growth success, but on growth media richer in both macronutrients and micronutrients [20] than those typically found in their natural environment. Therefore, any bacteria carried over from the origin pass through the same selective factors of culture acclimation through successive transfer cycles. The bacterial profile is then an indication of the epibenthic bacterial community present in the locations of origin and also capable of surviving interaction with the dinoflagellates. The transitional selection process that inevitably occurs upon dinoflagellate isolation argues strongly for early determination of bacterial diversity by high-selectivity techniques such as HTS to minimize culture artifacts. The high cryptic bacterial diversity among multiple P. lima strains cannot be interpreted further to compare variation at the dinoflagellate species level–P. hoffmannianum and P. rhathymum are represented only by a single strain each in the current study.
Known and potential functions of core members
The inferred metagenome serves only as a guide to focus the exploration of interactions, and thus we employ only the metabolic functional categories defined from the literature cited below. The metabolic strategies reported in similar or representative taxa related to the twelve bacterial ASVs identified as core members and their inferred metabolic contributions to interactions with the dinoflagellate may be grouped into three broad categories: stress tolerance, nutrient acquisition, and molecular bioactive interactions. The two most abundant core members Labrenzia and Roseitalea are known for the production of antioxidants. Labrenzia is capable of biosynthesis and degradation of dimethylsulfoniopropionate (DMS) [57]–a potent allelochemical, which may be produced as a stress response in coral holobionts. Roseitalea is characterized by the production of carotenoids [58]. Muricauda representatives produce the carotenoid zeaxanthin, suggested to protect corals from environmental stress [59], such as the high solar irradiance impacting benthic communities in shallow coastal waters. Other studies have reported the association of Labrenzia, Roseitalea and Muricauda, as well as Algiphilus, to thermo-tolerance in dinoflagellates of the endosymbiont family Symbiodiniaceae [60, 61]. The stress tolerance capabilities of these core members may be a feature leveraged by the dinoflagellate in specific environmental circumstances, thereby allowing the species to thrive in challenging environments.
The metabolic and signalling interactions between associated bacteria and eukaryotic phytoplankton have been investigated in several studies [62, 63]. Nutrient acquisition is among the central interactions. The exchange of fixed nitrogen for amino acids and organic carbon is known in terrestrial and aquatic members of Alphaproteobacteria [35, 64]. From the benthic marine environment, this lineage was represented by core members Labrenzia, Roseitalea, and Cohaesibacter. However, little is known about the metabolism of the latter genus, and these functions have not been experimentally reported. Among micronutrients and trace metals made bioavailable to microeukaryotes by bacteria, iron (Fe) is considered highly prominent. In bacteria, Fe can be captured by secretion of siderophores like vibrioferrin in Marinobacter [65] or massiliachelin in Massilia [66], both identified as core members of the Prorocentrum microbiome in this study. Transfer of sequestered bioavailable Fe to the dinoflagellate may be a critical function under Fe-limited environmental conditions. Bacterial metabolic activity in xenic dinoflagellate cultures and the known preference for ammonium ion (NH4+) instead of nitrate (NO3-) uptake for growth and toxin biosynthesis in Prorocentrum [67] demonstrates a potentially critical role of the bacterial community in nutrient recycling and conversion of inorganic to organic N and P as growth substrates. Although N is generally considered the primary limiting macronutrient for eukaryotic microalgae growth in marine ecosystems, in certain circumstances (e.g. in Norwegian fjords and the Ligurian Sea, Mediterranean), inorganic P may be scarce and recalcitrant and hence become the limiting nutrient (reviewed in Cembella et al. [68, 69]). In any case, bacterial mobilization of inorganic P-nutrients and conversion to low molecular weight organic P compounds which may be readily assimilated by eukaryotic microalgae are a major factor in both freshwater and marine systems [69]. In the marine environment, a possible role for Massilia could be the solubilization of phosphate as has been reported for terrestrial representatives [70].
The third category of known functions among bacterial core members covers a variety of secondary metabolism pathways. These functions likely serve as co-adaptation advantages and may change depending on the environment or during the growth stages or changes in the health status of the microalga in co-culture. In early stationery and senescence growth phases in culture, microalgal cells tend to become progressively leakier, releasing cell contents as bacterial load increases. Nevertheless, even during early growth phases, microalgae typically release low molecular weight molecules, such as amino acids or organic carbon [49]. High cell abundances of Marivita have been found to be associated with the lag phase of dinoflagellate Margalefidinium (Cochlodinium) polykrikoides, suggesting their participation as growth promoters through the production of vitamins [17]. Marivita has been found in association with epibenthic [38] and planktonic [41] dinoflagellates, and coccolithophorids [71]. High Marivita cell abundances were also part of the microbial community of the planktonic dinoflagellate Lingulodinium polyedra, a B1 and B12-vitamin auxotroph [72]. Machinery for the production of energy reservoirs was identified in Cohaesibacter and Mameliella. In these bacteria, stored poly-β-hydroxybutyrate (PHB) granules can be metabolized when common energy sources are in short supply [73, 74]. Moreover, eukaryotic microalgae can accumulate polycyclic aromatic hydrocarbons directly from seawater, and core members Mameliella and Algiphilus can metabolize a series of aromatic compounds [74, 75]. In fact, Algiphilus cell abundance in cultures of Lingulodinium increased after phenanthrene enrichment [76]. Finally, antibiotic production by bacteria is a strategy to deter the co-settling of competitors, including pathogenic bacteria, and microeukaryotes. Mameliella strains may produce the antibiotics bacteriocins and microcins [77], whereas the secondary pigment violacein synthesized by Massilia and explored as an anti-cancerogenic agent [78] is also highly toxic to bactivorous nanoflagellates.
Role of bacterial community diversity in the regulation of toxigenicity
Studies of the bacterial community in non-toxigenic and toxigenic strains originating from the same sites may help elucidate if specific bacterial profiles promote the toxigenicity of the host dinoflagellate. Prorocentrum strains from diverse locations in Mexico were previously shown [4, 16] to produce mainly the linear polyketide okadaic acid (OA) and its respective diol-ester (OA-D8), but the relationship with associated bacteria remained unclear. The single P. hoffmannianum strain PA85 yielded only four DST analogs, with approximately equimolar OA and OA-D8 as dominant analogs, followed by DTX1>DTX1-D8. This composition is unusual compared to the P. lima strains in this study, but the total cell toxin composition (fmol cell-1) could not be distinguished from other Prorocentrum strains from the Veracruz reef system.
In several cases, benthic Prorocentrum culture conditions have been optimized for maximum DST production [79], and even studied through the cell cycle [67]. The effects of toxigenicity on the bacterial assemblages, however, are not usually investigated, with rare exceptions [16, 19]. Previous studies report that depletion of bacteria by antibiotics did not affect the toxin production in P. lima cultures [16], providing evidence that complete DST synthesis by the dinoflagellate is not bacteria-dependent [80], nor subject to associated bacterial polyketide synthase (PKS) genes [47]. Inorganic nutrient limitation in P. lima may lead to an increase in DST cell quota due to uncoupling of toxin biosynthesis from growth [67], but with negligible effect on the DST toxin composition profile. In all cases, the experiments were conducted with non-axenic Prorocentrum cultures, therefore the nutrient-dependent role of bacteria is unknown. The question remains open of whether effects of associated bacteria on toxigenicity of Prorocentrum cells are indirectly manifested through general effects on growth and metabolism, or via direct transcriptomic or post-transcriptomic regulation of dinoflagellate PKS activity.
The decision to remove the only non-toxigenic strain P. rhathymum PA20 from specific analyses was justified by the fact that not a single bacterial taxon in its microbiome stood out by absence of toxigenicity. Nevertheless, among the bacterial taxa that were represented in higher relative abundance in PA20 compared to the toxigenic ones included functions related to the metabolism of C1 compounds (e.g., methylamine and methanol) by Methylophaga representatives [81, 82], or the inability to reduce nitrate by members of the genus Hoeflea isolated from a P. lima culture [83]. In the current analysis herein, the most abundant DSTs (OA and OA-D8) produced by these Prorocentrum strains were not significantly associated with the presence of particular bacteria. This result agrees with the growth data and DST production by a different set of Prorocentrum strains from the Gulf of Mexico and the Caribbean [16]. Only two closely related DST analogs (DTX1a and DTX1a-D8) presented significant correlations with bacterial ASVs. Taking into account that maximum taxonomic assignment of ASVs is at the genus level, these significantly-correlated taxa indicate that different bacterial species may be involved in the core assemblage than in the correlation with toxin production. Among ASVs correlating with toxin production are taxa from recognized families or genera with algicidal activity (Alteromonadaceae, Kordia, and Saprospiraceae), suggesting allelochemical interactions between these groups. As both the toxins and the bacteria involved in this correlation are present in low abundance, a bacteriostatic effect is plausible.
Conclusions
A more diverse bacterial community than previously characterized was detected by HTS techniques in xenic Prorocentrum cultures and apparently is maintained in both operationally defined compartments–free living in the medium (M) and in close association with the host (H). These H- and M-compartments were similar in composition of the bacterial microbiome. The Prorocentrum microbiota is partially structured based on the location of origin and may vary depending on the collection time (season). The dominant classes in strains from three disjunct locations within Mexican coastal waters were Alphaproteobacteria, Gammaproteobacteria, and Bacteroidia. The Alphaproteobacteria members of the genera Labrenzia and Roseitalea were the most abundant in the dinoflagellate-culture conditions in this study. Along with other seven genera, they comprised the Prorocentrum core members. Known interactions and metabolism of the core members, supported by inferred metabolism, suggest several pathways of mutualistic interaction covering carbon acquisition, nutrient exchange, and secondary metabolism of bioactive compounds. Together, the known functions of core members and the pathways to which they contribute show a variety of strategies available for a successful mutualistic relationship between the dinoflagellate host and its microbiota. Axenic dinoflagellate cultures and a variety of bacterial isolates will be required to design and perform experiments for the elucidation of specific interactions.
Supporting information
S1 Fig. Representation of the compartments analyzed from the dinoflagellate host (H) and culture medium (M) fractions.
A Prorocentrum cell is shown surrounded by the bacterial community from these two compartments. The host-associated (H) bacteria are endosymbiotic and/or bound to the dinoflagellate cell and included within the phycosphere (orange shading) whereas bacteria within the dinoflagellate culture-medium (M) are free-living or only loosely associated with the dinoflagellate.
https://doi.org/10.1371/journal.pone.0306108.s001
(PDF)
S2 Fig. Bacterial community sequencing effort.
Rarefaction curves of V3-V4 16S rRNA gene amplicon sequencing data processed into ASVs from bacteria of 11 Prorocentrum strains analyzed as the community directly associated with the dinoflagellate host (H) (top) and the free-living bacteria in the medium (M) (bottom).
https://doi.org/10.1371/journal.pone.0306108.s002
(PDF)
S3 Fig. Top abundant bacterial genera in Prorocentrum cultures.
Relative abundances per sample are shown at the genus level for the top 20 abundant genera in the fraction associated with the host (H) or free-living in the culture medium (M). Taxa labels include the order name and the genus name when possible, otherwise the last assigned level is used (f = family, p = phylum). Location of dinoflagellate strain origin is indicated on the top.
https://doi.org/10.1371/journal.pone.0306108.s003
(PDF)
S4 Fig. Determination of Prorocentrum core bacterial members.
A combination of high prevalence (tile color, 10 strains = 100%) and relative abundance (column-wise) were used to define the bacterial core members (indicated by black dots). ASV identifiers and associated taxonomy are shown at the left.
https://doi.org/10.1371/journal.pone.0306108.s004
(PDF)
S5 Fig. Radar plot of the compositional toxigenic profile of Prorocentrum strains.
Distribution of the relative DST composition (% molar) of 10 Prorocentrum strains with different location of origin. OA, okadaic acid; OA-D8, okadaic acid diol-ester; DTX1, dinophysistoxin 1; DTX1-D8, dinophysistoxin 1 diol-ester; DTX1a, undescribed DTX1 isomer; DTX1a-D8, undescribed DTX1a isomer.
https://doi.org/10.1371/journal.pone.0306108.s005
(PDF)
S1 Table. Summary data on Prorocentrum species and strains, location of origin, transfer cycles, toxin composition, and associated bacterial sequencing and data analysis by fraction (NA, not available).
https://doi.org/10.1371/journal.pone.0306108.s006
(XLSX)
S2 Table. Contribution of core ASVs to major KEGG pathways.
Result of a Mann-Whitney test (with Bonferroni correction, ns = non-significant) between contribution of core and non-core ASVs.
https://doi.org/10.1371/journal.pone.0306108.s007
(XLSX)
Acknowledgments
We thank Yuri Okolodkov (U. Veracruzana) for support with dinoflagellate taxonomy and Manuel Victoria-Muguira of the Dorado Buceo SCUBA agency (Veracruz) for help with field sampling and logistics. Sequencing was performed with the MiSeq sequencing platform at the Functional Genomics Laboratory and data analysis in the OMICA supercomputing facility of CICESE (Ensenada, Baja California, Mexico).
References
- 1. Durán-Riveroll LD, Cembella AD, Okolodkov YB. A review on the biodiversity and biogeography of toxigenic benthic marine dinoflagellates of the coasts of Latin America. Front Mar Sci. 2019;6:148.
- 2.
Guiry MD, Guiry GM. World-wide electronic publication, National University of Ireland, Galway. https://wwwalgaebaseorg, accessed on 05 November 2023.
- 3. Hoppenrath M, Chomérat N, Horiguchi T, Schweikert M, Nagahama Y, Murray S. Taxonomy and phylogeny of the benthic Prorocentrum species (Dinophyceae)—A proposal and review. Harmful Algae. 2013;27:1–28.
- 4. Cembella AD, Durán-Riveroll LM, Tarazona-Janampa UI, Okolodkov YB, García-Sandoval R, Krock B, et al. Phylogeography and diversity among populations of the toxigenic benthic dinoflagellate Prorocentrum from coastal reef systems in Mexico. Front Mar Sci. 2021;8.
- 5. Chomérat N, Bilien G, Zentz F. A taxonomical study of benthic Prorocentrum species (Prorocentrales, Dinophyceae) from Anse Dufour (Martinique Island, eastern Caribbean Sea). Mar Biodivers. 2019;49(3):1299–1319.
- 6. Berdalet E, Tester PA, Chinain M, Fraga S, Lemée R, Litaker W, et al. Harmful algal blooms in benthic systems: Recent progress and future research. Oceanography. 2017;30(1):36–45.
- 7. Lawrence JE. Prorocentrum lima: a putative link to diarrhetic shellfish poisoning in Nova Scotia. Harmful Algae. 1998:78–79.
- 8. Foden J, Purdie DA, Morris S, Nascimento S. Epiphytic abundance and toxicity of Prorocentrum lima populations in the Fleet Lagoon, UK. Harmful Algae. 2005;4(6):1063–1074.
- 9. Seymour JR, Amin SA, Raina JB, Stocker R. Zooming in on the phycosphere: The ecological interface for phytoplankton-bacteria relationships. Nat Microbiol. 2017. pmid:28555622
- 10. Ianora A, Bentley MG, Caldwell GS, Casotti R, Cembella AD, Engström-Öst J, et al. The relevance of marine chemical ecology to plankton and ecosystem function: an emerging field. Mar Drugs. 2011;9(9):1625–1648. pmid:22131962
- 11. Hu W, Xu J, Sinkkonen J, Wu J. Polyketides from marine dinoflagellates of the genus Prorocentrum, biosynthetic origin and bioactivity of their okadaic acid analogues. Mini Rev Med Chem. 2010;10(1):51–61.
- 12. Cembella AD. Chemical ecology of eukaryotic microalgae in marine ecosystems. Phycologia. 2003;42(4):420–447.
- 13. Lafay B, Ruimy R, De Traubenberg CR, Breittmayer V, Gauthier MJ, Christen R. Roseobacter algicola sp. nov., a new marine bacterium isolated from the phycosphere of the toxin-producing dinoflagellate Prorocentrum lima. Int J Syst Evol Microbiol. 1995;45(2):290–296.
- 14. Prokic I, Brümmer F, Brigge T, Görtz HD, Gerdts G, Schütt C, et al. Bacteria of the genus Roseobacter associated with the toxic dinoflagellate Prorocentrum lima. Protist. 1998;149(4):347–357. pmid:23194717
- 15. Jasti S, Sieracki ME, Poulton NJ, Giewat MW, Rooney-Varga JN. Phylogenetic diversity and specificity of bacteria closely associated with Alexandrium spp. and other phytoplankton. Appl Environ Microbiol. 2005;71(7):3483–3494.
- 16. Tarazona-Janampa UI, Cembella AD, Pelayo-Zárate MC, Pajares S, Márquez-Valdelamar LM, Okolodkov YB, et al. Associated bacteria and their effects on growth and toxigenicity of the dinoflagellate Prorocentrum lima species complex from epibenthic substrates along Mexican coasts. Front Mar Sci. 2020;7:569.
- 17. Park BS, Guo R, Lim WA, Ki JS. Pyrosequencing reveals specific associations of bacterial clades Roseobacter and Flavobacterium with the harmful dinoflagellate Cochlodinium polykrikoides growing in culture. Mar Ecol. 2017;38(6):1–8.
- 18. Park BS, Guo R, Lim W-A, Ki J-S. Importance of free-living and particle-associated bacteria for the growth of the harmful dinoflagellate Prorocentrum minimum: evidence in culture stages. Mar Freshwater Res. 2018;69(2):290–299.
- 19. Wu Z, Lee WH, Liu Z, Lin S, Lam PKS. Microbiome associated with Gambierdiscus balechii cultures under different toxicity conditions. Front Mar Sci. 2022;9:1–11.
- 20. Blackburn SI, Bolch CJS, Haskard KA, Hallegraeff GM. Reproductive compatibility among four global populations of the toxic dinoflagellate Gymnodinium catenatum (Dinophyceae). Phycologia. 2001;40(1):78–87.
- 21. Klindworth A, Pruesse E, Schweer T, Peplies J, Quast C, Horn M, et al. Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 2013;41(1):1–11. pmid:22933715
- 22. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nature Methods. 2016;13(7):581–583. pmid:27214047
- 23.
R Core Team. R: A language and environment for statistical computing. In: Computing RFfS, editor. Vienna, Austria: R Foundation for Statistical Computing; 2020.
- 24. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Res. 2013;41(D1):590–596. pmid:23193283
- 25. Wright ES, Yilmaz LS, Noguera DR. DECIPHER, a Search-Based Approach to Chimera Identification for 16S rRNA Sequences. AEM. 2012;78(3):717–725. pmid:22101057
- 26. Schliep KP. phangorn: Phylogenetic analysis in R. Bioinformatics. 2011;27(4):592–593. pmid:21169378
- 27. McMurdie PJ, Holmes S. Phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8(4). pmid:23630581
- 28. Lawson CA, Raina JB, Kahlke T, Seymour JR, Suggett DJ. Defining the core microbiome of the symbiotic dinoflagellate, Symbiodinium. Env Microbiol Rep. 2018;10(1):7–11. pmid:29124895
- 29. Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 2016;44(D1):D457–D462. pmid:26476454
- 30. Douglas GM, Maffei VJ, Zaneveld JR, Yurgel SN, Brown JR, Taylor CM, et al. PICRUSt2 for prediction of metagenome functions. Nat Biotechnol. 2020;38(6):685–688. pmid:32483366
- 31. Ye Y, Doak TG. A parsimony approach to biological pathway reconstruction/inference for genomes and metagenomes. PLoS Comput Biol. 2009;5(8):1–8. pmid:19680427
- 32. Louca S, Doebeli M. Efficient comparative phylogenetics on large trees. Bioinformatics. 2018;34(6):1053–1055. pmid:29091997
- 33. Barbera P, Czech L, Lutteropp S, Stamatakis A. SCRAPP: A tool to assess the diversity of microbial samples from phylogenetic placements. Mol Ecol Resour. 2020;(July):1–10. pmid:32996237
- 34.
Lahti L, Shetty S. Tools for microbiome analysis in R. Microbiome package v1.15.3. 2017.
- 35. Buchan A, LeCleir GR, Gulvik CA, González JM. Master recyclers: features and functions of bacteria associated with phytoplankton blooms. Nat Rev Microbiol. 2014;12(10):686–698. pmid:25134618
- 36. Vanucci S, Guidi F, Pistocchi R, Long RA. Phylogenetic structure of bacterial assemblages co-occurring with Ostreopsis cf. ovata bloom. Harmful Algae. 2016;55:259–271.
- 37. Rambo IM, Dombrowski N, Constant L, Erdner D, Baker BJ. Metabolic relationships of uncultured bacteria associated with the microalgae Gambierdiscus. Environ Microbiol. 2020;22(5):1764–1783.
- 38. Denaro R, Crisafi F, Smedile F, Soprano V, Rossi R, Zingone A, et al. Microbiomes associated with cultures of Gambierdiscus australes and Ostreopsis cf. ovata, two epibenthic dinoflagellates from the NE Atlantic Ocean (Las Palmas, Gran Canaria). Mar Ecol 2022;(November 2021):1–13.
- 39. Deng Y, Wang K, Hu Z, Tang Y-Z. Abundant species diversity and essential functions of bacterial communities associated with dinoflagellates as revealed from metabarcoding sequencing for laboratory-raised clonal cultures. IJERPH. 2022;19(8):4446–4446. pmid:35457312
- 40. Maire J, Girvan SK, Barkla SE, Perez-Gonzalez A, Suggett DJ, Blackall LL, et al. Intracellular bacteria are common and taxonomically diverse in cultured and in hospite algal endosymbionts of coral reefs. ISME J. 2021;15(7):2028–2042. pmid:33558689
- 41. Sörenson E, Bertos‐Fortis M, Farnelid H, Kremp A, Krüger K, Lindehoff E,et al. Consistency in microbiomes in cultures of Alexandrium species isolated from brackish and marine waters. Env Microbiol Rep. 2019;11(3):425–433. pmid:30672139
- 42. Bengtsson MM, Øvreås L. Planctomycetes dominate biofilms on surfaces of the kelp Laminaria hyperborea. BMC Microbiology. 2010;10:1–12. pmid:20950420
- 43. Kantor RS, van Zyl AW, van Hille RP, Thomas BC, Harrison STL, Banfield JF. Bioreactor microbial ecosystems for thiocyanate and cyanide degradation unravelled with genome-resolved metagenomics. Environ Microbiol. 2015;17(12):4929–4941. pmid:26031303
- 44. Barreto Filho MM, Walker M, Ashworth MP, Morris JJ. Structure and long-term stability of the microbiome in diverse diatom cultures. Microbiol Spectr. 2021;9(1):e00269–00221. pmid:34190604
- 45. Fei C, Ochsenkühn MA, Shibl AA, Isaac A, Wang C, Amin SA. Quorum sensing regulates ‘swim-or-stick’ lifestyle in the phycosphere. Environ Microbiol. 2020;22(11):4761–4778. pmid:32896070
- 46. Deng Z, Chen S, Zhang P, Zhang X, Adams JM, Luo Q, et al. Dynamics of free-living and attached bacterial assemblages in Skeletonema sp. diatom cultures at elevated temperatures. Front Mar Sci. 2021;8(July):1–14.
- 47. Perez R, Liu L, Lopez J, An T, Rein KS. Diverse bacterial PKS sequences derived from okadaic acid-producing dinoflagellates. Mar Drugs. 2008;6(2):164–179. pmid:18728765
- 48. Rappé MS, Giovannoni SJ. The uncultured microbial majority. Annu Rev Microbiol. 2003;57(1):369–394. pmid:14527284
- 49. Azam F, Malfatti F. Microbial structuring of marine ecosystems. Nat Rev Microbiol. 2007;5(10):782–791. pmid:17853906
- 50. Ajani PA, Kahlke T, Siboni N, Carney R, Murray SA, Seymour JR. The microbiome of the cosmopolitan diatom Leptocylindrus reveals significant spatial and temporal variability. Front Microbiol. 2018;9(NOV):1–12. pmid:30498485
- 51. Fuhrman JA, Steele JA, Hewson I, Schwalbach MS, Brown MV, Green JL, et al. A latitudinal diversity gradient in planktonic marine bacteria. PNAS. 2008;105(22):7774–7778. pmid:18509059
- 52. Giovannoni SJ, Vergin KL. Seasonality in ocean microbial communities. Science. 2012;335(6069):671–676. pmid:22323811
- 53. Sapp M, Schwaderer AS, Wiltshire KH, Hoppe H-G, Gerdts G, Wichels A. Species-specific bacterial communities in the phycosphere of microalgae? Microb Ecol. 2007;53:683–699. pmid:17264999
- 54. Behringer G, Ochsenkühn MA, Fei C, Fanning J, Koester JA, Amin SA. Bacterial communities of diatoms display strong conservation across strains and time. Front Microbiol. 2018;9(APR):1–15. pmid:29681892
- 55. Lupette J, Lami R, Krasovec M, Grimsley N, Moreau H, Piganeau G, et al. Marinobacter dominates the bacterial community of the Ostreococcus tauri phycosphere in culture. Front Microbiol. 2016;7:1414.
- 56. Jackrel SL, Yang JW, Schmidt KC, Denef VJ. Host specificity of microbiome assembly and its fitness effects in phytoplankton. ISME J. 2021;15(3):774–788. pmid:33097853
- 57. Zhong H, Sun H, Liu R, Zhan Y, Huang X, Ju F, et al. Comparative genomic analysis of Labrenzia aggregata (Alphaproteobacteria) strains isolated from the Mariana Trench: Insights into the metabolic potentials and biogeochemical functions. Front Microbiol. 2021;12(December). pmid:34970235
- 58. Hyeon JW, Jeong SE, Baek K, Jeon CO. Roseitalea porphyridii gen. Nov., sp. nov., isolated from a red alga, and reclassification of Hoeflea suaedae Chung et al. 2013 as Pseudohoeflea suaedae gen. nov., comb. nov. Int J Syst Evol. 2017;67(2):362–368. pmid:27902230
- 59. Motone K, Takagi T, Aburaya S, Miura N, Aoki W, Ueda M. A zeaxanthin-producing bacterium isolated from the algal phycosphere protects coral endosymbionts from environmental stress. mBio. 2020;11(1):1–13. pmid:31964724
- 60. Grottoli AG, Dalcin Martins P, Wilkins MJ, Johnston MD, Warner ME, Cai W-j, et al. Coral physiology and microbiome dynamics under combined warming and ocean acidification. PLoS One. 2018;13(1):e0191156–e0191156. pmid:29338021
- 61. Buerger P, Vanstone RT, Maire J, van Oppen MJ. Long-term heat selection of the coral endosymbiont Cladocopium C1acro (Symbiodiniaceae) stabilizes associated bacterial communities. Int J Mol Sci. 2022;23(9):4913.
- 62. Amin S, Hmelo L, Van Tol H, Durham B, Carlson L, Heal K, et al. Interaction and signalling between a cosmopolitan phytoplankton and associated bacteria. Nature. 2015;522(7554):98–101. pmid:26017307
- 63. Tang YZ, Koch F, Gobler CJ. Most harmful algal bloom species are vitamin B1 and B12 auxotrophs. PNAS. 2010;107(48):20756–20761. pmid:21068377
- 64.
Wang ET, Tian CF, Chen WF, Young JPW, Chen WX. Ecology and evolution of Rhizobia: Principles and applications. Singapore: Springer Nature; 2019.
- 65. Amin SA, Green DH, Hart MC, Küpper FC, Sunda WG, Carrano CJ. Photolysis of iron-siderophore chelates promotes bacterial-algal mutualism. PNAS. 2009;106(40):17071–17076. pmid:19805106
- 66. Diettrich J, Kage H, Nett M. Genomics-inspired discovery of massiliachelin, an agrochelin epimer from Massilia sp. NR 4–1. Beilstein J Org Chem. 2019;15(4):1298–1303. pmid:31293678
- 67. Pan Y, Cembella AD, Quilliam MA. Cell cycle and toxin production in the benthic dinoflagellate Prorocentrum lima. Mar Biol. 1999;134(3):541–549.
- 68. Cembella AD, Antia NJ, Harrison PJ. The utilization of inorganic and organic phosphorous compounds as nutrients by eukaryotic microalgae: A multidisciplinary perspective: Part I. Crit Rev Microbiol. 1982;10(4):317–391.
- 69. Cembella AD, Antia NJ, Harrison PJ. The utilization of inorganic and organic phosphorous compounds as nutrients by eukaryotic microalgae: a multidisciplinary perspective: Part 2. Crit Rev Microbiol. 1984;11(1):13–81.
- 70. Zheng BX, Bi QF, Hao XL, Zhou GW, Yang XR. Massilia phosphatilytica sp. nov., a phosphate solubilizing bacteria isolated from a long-term fertilized soil. Int J Syst Evol. 2017;67(8):2514–2519. pmid:28853679
- 71. Green DH, Echavarri-Bravo V, Brennan D, Hart MC. Bacterial diversity associated with the coccolithophorid algae Emiliania huxleyi and Coccolithus pelagicus f. braarudii. BioMed Res Int. 2015:1–15. pmid:26273594
- 72. Cruz-López R, Maske H. The vitamin B1 and B12 required by the marine dinoflagellate Lingulodinium polyedrum can be provided by its associated bacterial community in culture. Front Microbiol. 2016;7(MAY):1–13. pmid:27199906
- 73. Hwang CY, Cho BC. Cohaesibacter gelatinilyticus gen. nov., sp. nov. a marine bacterium that forms a distinct branch in the order Rhizobiales, and proposal of Cohaesibacteraceae fam. nov. Int J Syst Evol. 2008;58(1):267–277.
- 74. Yamamoto T, Liu Y, Hasegawa Y, Iwaki H. Complete genome sequence of Mameliella alba Strain KU6B, a cyclohexylamine-utilizing marine bacterium. Microbiol Resour Announc. 2020;9(19):17–18. pmid:32381613
- 75. Gutierrez T, Green DH, Whitman WB, Nichols PD, Semple KT, Aitken MD. Algiphilus aromaticivorans gen. nov., sp. nov., an aromatic hydrocarbon-degrading bacterium isolated from a culture of the marine dinoflagellate Lingulodinium polyedrum, and proposal of Algiphilaceae fam. nov. Int J Syst Evol. 2012;62(11):2743–2749.
- 76. Thompson HF, Lesaulnier C, Pelikan C, Gutierrez T. Visualisation of the obligate hydrocarbonoclastic bacteria Polycyclovorans algicola and Algiphilus aromaticivorans in co-cultures with micro-algae by CARD-FISH. J Microbiol Methods. 2018;152(July):73–79. pmid:30063956
- 77. Danish-Daniel M, Han Ming G, Noor MEM, Yeong YS, Usup G. Draft genome sequence of Mameliella alba strain UMTAT08 isolated from clonal culture of toxic dinoflagellate Alexandrium tamiyavanichii. Genom Data. 2016;10:12–14. pmid:27625991
- 78. Myeong NR, Seong HJ, Kim HJ, Sul WJ. Complete genome sequence of antibiotic and anticancer agent violacein producing Massilia sp. strain NR 4–1. J Biotechnol. 2016;223(4):36–37. pmid:26916415
- 79. Hashimoto K, Uchida H, Nishimura T, Oikawa H, Funaki H, Honma C, et al. Determination of optimal culture conditions for toxin production by a Prorocentrum lima complex strain with high diarrhetic shellfish toxins yield. Harmful Algae. 2021;103:102025.
- 80. De Traubenberg CR, Soyer-Gobillard M-O, Géraud M-L, Albert M. The toxic dinoflagellate Prorocentrum lima and its associated bacteria. Eur J Protistol. 1995;31(4):383–388.
- 81. Neufeld JD, Schäfer H, Cox MJ, Boden R, McDonald IR, Murrell JC. Stable-isotope probing implicates Methylophaga spp. and novel Gammaproteobacteria in marine methanol and methylamine metabolism. ISME J. 2007;1(6):480–491.
- 82. Thompson HF, Summers S, Yuecel R, Gutierrez T. Hydrocarbon-degrading bacteria found tightly associated with the 50–70 μm cell-size population of eukaryotic phytoplankton in surface waters of a Northeast Atlantic region. Microorganisms. 2020;8(12):1955.
- 83. Biebl H, Tindall BJ, Pukall R, Lünsdorf H, Allgaier M, Wagner-Döbler I. Hoeflea phototrophica sp. nov., a novel marine aerobic alphaproteobacterium that forms bacteriochlorophyll a. IJSEM. 2006;56(4):821–826. pmid:16585702