Combined in vivo and in situ genome-resolved metagenomics reveals novel symbiotic nitrogen fixing interactions between non-cyanobacterial diazotrophs and microalgae

doi:10.21203/rs.3.rs-2361194/v1

Download PDF

Article

Combined in vivo and in situ genome-resolved metagenomics reveals novel symbiotic nitrogen fixing interactions between non-cyanobacterial diazotrophs and microalgae

https://doi.org/10.21203/rs.3.rs-2361194/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Non-cyanobacteria diazotrophs (NCDs) were shown to dominate in surface waters shifting the long-held paradigm of cyanobacteria dominance and raising fundamental questions on how these putative heterotrophic bacteria thrive in sunlit oceans. Here, we report an unprecedented finding in the widely used model diatom Phaeodactylum tricornutum (Pt) of NCDs sustaining diatom cells in the absence of bioavailable nitrogen. We identified PtNCDs using metagenomics sequencing and detected nitrogenase gene in silico and/or by PCR. We demonstrated nitrogen fixation in PtNCDs and their close genetic affiliation with NCDs from the environment. We showed the wide occurrence of this type of symbiosis with the isolation of NCDs from other microalgae, their identification in the environment, and predicted their associations with photosynthetic microalgae. Overall, this study provides evidence for a previously overlooked symbiosis using a multidisciplinary model-based approach, which will help understand the different players driving global marine nitrogen fixation.

Biological sciences/Microbiology/Environmental microbiology/Water microbiology

Earth and environmental sciences/Ocean sciences/Marine biology

Biological nitrogen fixation

non-cyanobacterial diazotrophs

Phaeodactylum tricornutum

microalgae

symbiosis

Nearly 80% of Earth’s atmosphere is in the form of molecular nitrogen (N₂). Similarly, the main form of nitrogen in the oceans which cover 71% of Earth surface is dissolved dinitrogen. The rest is reactive nitrogen in the form of nitrate, ammonia and dissolved organic compounds, which are very scarce rendering nitrogen one of the main limiting factors of productivity in most areas of the oceans¹. Despite its dominance, ironically, N₂ is unavailable for use by most organisms because there is a strong triple bond between the two nitrogen atoms that requires high energy input to split the molecule and supply bio-accessible nitrogen². Biological nitrogen fixation which is an energy intensive process is specific to only some prokaryotes and archaea termed diazotrophs. It is catalyzed by the nitrogenase enzyme complex including the nitrogenase reductase encoded by the nifH gene and the nitrogenase composed of two pairs of different subunits, α and β, respectively encoded by the nifD and nifKgenes³. Therefore, eukaryotes are only able to obtain fixed nitrogen through their symbiotic interactions with diazotrophs⁴.

In aquatic habitats, symbiosis between nitrogen fixing cyanobacteria and various eukaryotes, including several diatom genera such as Chaetoceros, Climacodium and Hemiaulus, are common and important in oligotrophic habitats of the ocean⁵. Because the nitrogenase complex is sensitive to oxygen, diazotrophs evolved protective nitrogen fixation alternatives that involve conditionally, temporally, or spatially separating oxidative phosphorylation or photosynthesis from nitrogen fixation^6-11. Cyanobacteria were considered for a very long time as the main diazotrophs in the ocean. However, recent molecular analyses indicated that non-cyanobacterial diazotrophs (NCDs) are also present and active showing an important presence of diverse nifH gene amplicons related to non-cyanobacteria, especially to Proteobacteria and Planctomycetes^12-16.

The overwhelming dominance of NCDs raises important questions on how these presumably heterotrophs proteobacteria thrive in the photic zones of the oceans. One emerging hypothesis is their association with microalgae in symbiotic interactions for nitrogen fixation. In our work, we provide first evidence of occurrence of such interactions between NCDs and several microalgae. More importantly, we identified this interaction in an attractive biological model, the diatom Phaeodactylum tricornutum (Pt), opening important and unique opportunities to investigate the molecular mechanisms of diazotrophy. We sequenced the bacterial meta-community associated with Pt and identified previously unknown diverse marine diazotrophs of mainly α-proteobacteria type, supporting the host cells in the absence of any form of usable nitrogen. Further, we isolated, cultured, sequenced and fully assembled most of the NCDs identified in lab culture. Their assembled genomes clustered phylogenetically to NCDs found in the environment and showed environmental co-occurrence patterns with microalgae including diatoms, green algae and haptophytes. Our work brings evidence for an ecologically important symbiotic interaction between microalgae and NCDs that might explain how both partners cope with periods of nitrogen scarcity, and how heterotrophic NCDs compromise for nitrogen fixation with its high energy cost in fully oxygenated areas of the water column. These new findings will have to be considered in future nitrogen and carbon biogeochemical cycling studies.

Identification of non-cyanobacterial diazotrophs symbiosis with diatoms

A screen of 10 xenic accessions of Pt in the absence of any source of usable nitrogen in the growth medium, identified one accession, P. tricornutum 15 (hereafter Pt15) that was able to survive in comparison to the axenic control (Fig. 1a), suggesting that diatom growth was supported by putative bacterial nitrogen fixation providing a usable source of nitrogen to diatom cells. After 3 months of nitrogen starvation, xenic Pt15 showed similar growth to non-starved cells upon transfer to nitrate replete medium. This indicated that Pt15 starved cells stayed alive and capable of growth in favorable conditions, evidenced by their exponential growth when transferred to nitrate replete medium (Fig. 1b). To demonstrate nitrogen fixation, we used the widely applied Acetylene Reduction Assay (ARA)¹⁷ that showed ethylene production, although low in xenic Pt15 culture starved for nitrogen, but not in plus N condition (+N). No ethylene was detected in axenic Pt15 confirming that the detected ethylene was not endogenous to the diatom cells but originated from bacterial nitrogen fixation (Supplementary Table 1).

To further investigate nitrogen fixing symbiosis in minus N condition (-N), we measured Particulate Organic Nitrogen quota (PON) in diatom cells and showed a stable level of PON even after 80 days of growth, supporting the maintenance of diatom cells with infinite amounts of usable nitrogen likely provided by Pt15 associated bacteria (Supplementary Fig. 1a). Simultaneously, measured biomass of Pt15, indicated a stable cell population suggesting a ground state photosynthetic activity that might support the energy demand of nitrogen fixation (Supplementary Fig. 1b). Secreted NH4 was found in important amounts into the medium (data not shown). Taken together, our results suggest bacterial nitrogen fixation that maintain diatom population cells stable and alive.

Light microscopy showed an important accumulation of bacteria in -N compared to +N condition (Fig. 1c,d,e,f). Transmission electron microscopy confirmed this observation and revealed a peculiar structure around diatom cells looking like a pellicle of exopolysaccharides exclusively present in starved cells (Fig. 1g,h). Similar structure, although less prominent occurred around some of bacterial cells either as a single bacterium or a group of several bacteria.

To identify species composition of Pt15 bacterial community in -N and +N conditions, bacterial DNA was extracted and shotgun sequenced revealing a total of 153,4 million reads. Using the GTDB reference database, we identified a total of 2 classes, 8 families, 12 orders and 90 genera. At class level, α-proteobacteria represented 98% ± 0,8 in +N and 99% ± 0,5 in -N, based on relative abundance estimations, while Gammaproteobacteria were detected with a relative abundance below 1% (Figure 2a). At family level, Sphingomonadaceae and Rhizobiaceae, both known to contain diazotrophic species dominated (Figure 2c). At genus level, among the 90 distinct taxa identified, 73 genera were only detected in -N with no reads in +N (Supplementary Table 2). Composition of the community microbiome at genus level showed a dominance of Sphyngopixys in +N, while -N showed a significant enrichment in Mesorhizobium sp., Bradyrhizobium sp., Sphingomonas sp., and Brevundimonas sp., all known as nitrogen fixing taxa (Fig. 2d). Compared to +N, Sphyngopixys richness decreased in -N although its dominance persisted in this condition. Interestingly, several presumably non-nitrogen fixing species (67 out of 73) were significantly enriched in -N (Supplementary Table 2), suggesting a role of these species in direct or indirect interactions within the bacterial community and/or with the host.

Isolation/identification of P. tricornutum associated non-cyanobacterial diazotrophs

Given the unique nature of the identified symbiosis involving a model diatom species and NCDs, it was important to isolate the diazotrophic bacteria identified in silico from metagenomics (MetaG) data in order to further explore this interaction. We used universal media and took advantage of the species assignation provided by MetaG to target the isolation of the desired bacteria according to literature reports^18-21. Therefore, different growth conditions (media, temperature) were tested (Supplementary file 1) and permitted the isolation of 12 bacterial species (Table 1). Bacterial colonies were identified using either full length 16S rRNA gene or two hypervariable regions of the 16S (V1V2, V3V4). The presence of nifH was assayed using three combinations of degenerate primers as well as nifH specific primers designed from MetaG data whenever possible (Supplementary Fig. 2a,b). Six bacteria were identified as nitrogen fixers based on a combination of nifH PCR amplification and/or nifH in silico detection (Table 1). Comparative metagenomics analysis identified two Sphyngopyxis and Bradyrhizobium species, which were distinguished using Metagenome-Assembled Genomes (MAGs) specific primers designed for each species (Supplementary Fig. 2). The whole set of nif genes including the nifHDK operon were identified in Bradyrhizobium MAG21, the only bacterium that was sequenced as a single species, while the other MAGs were assembled from either the MetaG data or sequences from a pool of 3 isolated bacteria. Two and four copies of nifH were detected in Bradyrhizobium MAG21 and Mesorhizobium MAG3, respectively. Among the 12 isolated bacteria, one species Methylobacterium was not detected in metagenomics sequences likely due to its low abundance emphasizing the importance of combining sequencing and isolation/culturing approaches (Table 1). Methylobacterium grew in nitrogen free medium (NFM) and nitrogen fixation was reported in this genus²². To further evidence nitrogen fixing ability of the isolated bacteria, we used nitrogen-free medium with bromothymol blue, a pH sensitive dye to grow the two most abundant bacteria, Mesorhizobium and Bradyrhizobium and found a switch in the color from green (pH=6.7) to blue (pH > 7) indicating nitrogen fixation and the release of NH3, which alkalinizes the medium (Figure 1i).

Diazotrophic MAGs are diverse and closely affiliated to environmentally sampled diazotrophs of Alphaproteobacteria type

To gain insights into Pt15 associated diazotrophs and their phylogenetic relationship with known taxa, we assembled their genomes and recovered several MAGs with completeness higher than 97% (Table 1, Supplementary Table 3). To further improve MAG assembly for some of the isolated bacteria, besides MetaG data initially used, MAGs were assembled with additional Illumina and Pacific Biosciences sequencing of the isolated bacteria including Bradyrhizobium, Georhizobium and Sphyngopyxis which were sequenced either alone or in a pool of 3 bacteria. Overall, chromosomal sequences were well assembled with higher completeness as compared to MAGs and nearly no cross contamination of bins. To achieve a better identification of the bacterial genetic material, we sought to recover plasmids from non-assembled reads, which are important elements for bacteria survival susceptible to contain nitrogen fixation genes. We used three different plasmid prediction tools and recovered 42 plasmids assigned to different bacterial species with an average length of 97013 bp (Supplementary Table 3). The nitrogenase gene (nifH) was in silico detected in six out of eight Pt15 associated MAGs among which, one plasmid located nifH (Table 1). Besides nifH, several additional nif genes were detected in most of the MAGs, among which two were found in plasmids, nifU and nifS (Table 1). Phylogenomic analysis of Pt15 MAGs, MAGs of known terrestrial and marine diazotrophs from the Microscope database²³ and MAGs from Tara Oceans data using Anvi’o revealed two clusters, A (Mesorhizobium, Bradyrhizobium MAGs, Georhizobium, Phylobacterium, and Bevundimonas) and B (2 Sphyngopyxis MAGs and Sphingobium). Cluster A was affiliated to terrestrial and marine diazotroph MAGs with a closer affiliation to terrestrial MAGs of the same genus (Fig. 3), while cluster B was related to genera from marine and terrestrial environments identified as nitrogen fixers. In both clusters, whenever the related MAGs were not identified as diazotrophs (likely due to incomplete assembly), they were found to belong to Rhizobiales, Parvibaculales, Caulobacterales and Sphingomonodales, orders known to contain nitrogen fixing species^24-26. Altogether, our results including (i) survival of the xenic diatom in the absence of usable forms of nitrogen, (ii) nitrogen fixation demonstrated by ARA and BTB, (iii) detection of nifH, and (iv) the isolation from the xenic Pt15 of NCDs clustering with known proteobacteria diazotrophs, indicated that a symbiotic nitrogen fixation likely occurred in Pt15 xenic culture supporting the survival of diatom cells population in -N condition.

Functional gene annotation of Pt15 associated NCDs in nitrate deplete versus replete conditions

To understand the biological functions of the microbial community, genes from both conditions were assigned to functional categories (Figure 4, supplementary Table S4). There was a significant enrichment of genes related to nitrogen fixation and metabolism in -N compared to +N condition. Typically, in -N, we exclusively identified nifH and nifD, two key components of the nitrogenase enzyme complex. Likewise, nifX and nifT were only detected in -N suggesting their critical role in nitrogen fixation. Part of the nif operon, NifX was reported to be important in FeMo-co maturation and its transfer to NifDK proteins^27,28. Other genes related to nitrogen fixation metabolism emerged from the analysis, namely HemN, Fpr, FixS, FixA, FixH, GroEL, NifS and FixJ (Figure 4), some of which are part of the NIF regulon.

The abundance of several transporters was striking in -N, indicating an important role in energy/substrates shuttling between the host and the bacterial community. Examples included EcfT gene (energy-coupling factor transporter transmembrane) known for its function in providing the energy necessary to transport a number of different substrates²⁹, AcpS, an acyl carrier gene, which functions as a multi-carbon units carrier into different biosynthetic pathways³⁰, TauB, a taurine transporter that might play a role in providing taurine, which is produced by a large number of algae and known as a potential source of carbon, nitrogen and energy³¹. Stress responsive genes were clearly more represented in -N compared to +N including SpoT, which is a general bacterial stress response³² that allows bacteria to increase its persistence in stressful conditions, here the lack of nitrogen. Other genes found almost exclusively in the starved bacterial community were the PAS domain containing genes (Per-ARNT-Sim)³³, which function as signal sensors by their capacity to bind small molecules, modulators and transducers. They are known to sense oxygen to prevent degradation of oxygen sensitive enzymes suggesting a crucial role in -N condition where there is an enrichment in nitrogenase containing bacteria. Overall, functional analysis of MetaG data corroborated with NCDs enrichment in -N and nitrogen fixation.

Geographical distribution and ecological relevance of NCD-microalgae symbiosis

Next, we sought to assess the prevalence and importance of similar interactions in other microalgae and in the environment. First, we screened a large collection of xenic microalgae as described for Pt15 and identified 5 species of microalgae (35%) that grew in the absence of nitrogen in the growth medium suggesting the presence of diazotrophs (Supplementary Table 5). As for Pt15, the retained microalgae were screened using different nitrogen deprived media under various conditions for diazotrophs isolation. NifH degenerate and 16S rRNA gene primers were used to identify the isolated bacterial colonies, which revealed five Alpha-proteobacteria including 2 species of Thalassiospira sp. (isolated from two different microalgae), Stenotrophomonas maltophilia, Stappia sp., Rhizobium and Rhodococcus qingshengii suggesting that the occurrence of NCDs in a symbiotic interaction with microalgae is frequent. PCR detection using degenerate primers was successful only in Stappia sp., and S. maltophilia. However, the remaining species grew in NFM and/or fixed nitrogen (Supplementary Table 5).

Identified NCDs were sequenced in a pool of 3 bacteria using Illumina and PacBio and their genome were assembled, showing overall, a high completeness (>87%, 2 MAGs with 100%) and low contamination (<2%) (Supplementary Table 3). We then used MAGs isolated from Pt15 and the other microalgae (Table 1, Supplementary Table 3) to reassess their phylogenetic relationship as described for Pt15 isolated NCDs. The analysis revealed 3 more clusters (C, D and E) in close proximity to diazotrophic MAGs from the environment and/or MAGs belonging to diazotrophic containing orders³⁴ ³⁵ ³⁶ (Supplementary Fig. 3).

Next, we assessed the occurrence of Pt15 MAGs in the environment in surface water and Deep Chlorophyll Maximum layer (DCM) using Tara Ocean datasets^37,38. Among the 130 sampling stations, 92 stations showed the occurrence of at least one of the Pt15 MAGs reflecting a high prevalence (Fig. 5a, Supplementary Fig. 4). Pt15 closely affiliated Tara MAGs were more frequently occurring with 129 recorded stations. The distribution of Pt15 MAGs was extensive, spanning all latitudes from south Pacific, Indian and Atlantic Oceans to polar regions. The MAGs were found in both surface and DCM samples, mainly within the 0.8-5µm, 0.8-2000, 3-5-20µm and 180-2000 size fractions (Fig. 5a, supplementary Fig. 4). Interestingly, Pt15 MAGs co-occurred in surface and DCM samples in most cases, although their dominance in surface samples was prominent. Except for few stations, clusters A and B seemed to co-occur suggesting potential ecological interactions between their respective bacteria.

We then asked whether the occurrence of the MAGs isolated from Pt15 correlated with environmental factors monitored during the Tara Oceans campaign and whether they co-occur with other organisms. We considered only two Pt15 MAGs (Mesorhizobium MAG3 and Bradyrhizobium MAG8) to assess their correlation with four environmental factors used as indicators of diazotrophy, namely nitrate, nitrate/phosphate ratio, iron, and oxygen. Interestingly, Pt15 MAGs were detected in low nitrate and almost no detection above 5 µM. Similarly, the detection of Pt15 MAGs coincided with low nitrate to phosphate ratio (Fig. 5b,c,d,e). Both measurements are in line with nitrate limitation and phosphate abundance as markers of diazotrophs presence^39,40. Iron availability seemed to affect differently the abundance of both MAGs. While, Mesorhizobium was observed in low to no detectable levels of iron, Bradyrhizobium MAG8 was detected in higher iron concentrations suggesting different requirement for iron in these two species. By contrast to what is known about oxygen inhibition of nitrogenase, both MAGs seemed to thrive in relatively well oxygenated zones (200 µM) compared to less oxygenated areas or oxygen minimum zones (OMZ, 20- 50 µM O₂). This implies adaptive means by which these bacteria fix nitrogen (Supplementary Fig. 4)

Strikingly, a co-occurrence analysis revealed a clear association of the NCDs isolated from Pt15 (Fig 5f) with mainly chromista including diatoms, haptophytes, and chlorophytes. Bradyrhizobium MAG8 showed the highest co-occurrence overall followed by Sphyngopixys MAG7. Most shared co-occurrences were with haptophytes, Bikosea and Pycnococcus. Although, nitrogen fixation is not demonstrated for these predicted associations, this analysis comforted a likely symbiotic interaction for nitrogen fixation between investigated NCDs and different species of mainly photosynthetic micro-eukaryotes.

Recent reports of NCDs dominance in the euphotic zones of the ocean questioned the long-held dogma that symbiotic N₂ fixation is mainly carried by Cyanobacteria, but also raised fundamental questions on how these presumably heterotroph NCDs can thrive in the sunlit ocean^12-14,41. The nitrogenase enzyme complex responsible for nitrogen fixation is sensitive to oxygen that irreversibly inactivates the enzyme. Therefore, diazotrophs evolved several mechanisms to protect nitrogenase from oxygen degradation including hyper-respiration, uncoupling of photosynthesis in space and time in photosynthetic diazotrophs, and attachment/interaction with microalgae or debris which were found to be hotspots of nitrogen fixation⁴². Using a multidisciplinary approach, the study described herein brings unprecedented evidence of symbiotic interactions for nitrogen fixation between microalgae in particular the model diatom Phaeodactylum tricornutum and NCDs. Our study revealed among several screened xenic accessions of P. tricornutum in the absence of usable form of nitrogen, one accession, Pt15 interacting with nitrogen fixing proteobacteria. Although low, this nitrogen fixation and uptake seemed to maintain the viability of the diatom and its associated bacteria. To compensate for the allocation of energy to the bacterial community, Pt15 cells seemed to slow down growth. This lab culture interaction may reflect natural environments, where such symbiotic interactions, in extreme conditions (low to no nitrogen) take place until favorable factors (e.g., nitrate repletion with upwelling or Sahara dust) permit the bloom of cells we observed in our experimental cultures when transferred to nitrate replete medium.

MetaG sequencing of the associated bacteria revealed not only one diazotrophic species but a group of nitrogen fixers almost exclusively Alpha-proteobacteria mainly represented by Rhizobiaceae, Sphingomonadaceae and Xanthobacteraceae. The nitrogenase gene was identified in all of them either in silico or by PCR amplification, supporting further diazotrophy in these bacteria. The nifH gene was found in most cases in the chromosome, except for Georhizobium where it was predicted to be plasmidic. Interestingly the chromosomal nif genes are clustered over 30-40 Kb regions forming genomic islands, a feature found in symbiotic bacteria where these clusters are known as symbiotic islands, which corresponds to an evolutionary strategy adopted by symbiotic bacteria in their quest to interact with eukaryotic hosts⁴³. Demonstrated by ARA, the rates of nitrogen fixation in xenic Pt15 in -N are low compared to the used controls but higher comparable to non-cyanobacteria diazotrophs from the environment, although monitoring techniques and conditions, environment versus lab cultures are different^44,45.

Our study reveals that NCDs interaction with Pt15 and other microalgae is likely not an isolated phenomenon and is more broadly distributed than previously anticipated. Our MAGs were detected in the environment as closely affiliated with previously identified terrestrial and marine NCDs supporting diazotrophy and their ecological relevance. These MAGs were identified at global scale in surface and DCM samples in microalgae containing size fractions, which strongly supported the hypothesis of an interaction with them as evidenced by the co-occurrence analysis. This interaction might be taking advantage of hypo-oxic biomes microalgae and bacteria may provide with the secretion of extracellular polymers, mainly polysaccharides, that form biofilm ensuring low O₂ diffusion, which was suggested by our TEM analysis with the formation of pellicles around Pt cells and bacteria. Such pellicles were reported to form around two nitrogen fixing bacteria, P. stutzeri and Azospirillum brasilense in oxygenated surroundings implying the importance of oxygen stress in the formation of these structures⁴⁶. The presence of these sac like structures might have hindered diatom uptake of released ammonium leading to growth inhibition. Besides biofilm formation, this symbiotic interaction may combine different strategies for both microalgae and NCDs to avoid/limit oxygen inactivation of nitrogenase such as a conformational switch that protects nitrogenase from O₂ degradation^10,47 or hyper-respiration⁴⁸ among other protective mechanisms that need further investigations. In line with our discovery of NCDs interaction with microalgae for nitrogen fixation, a recent study⁴⁹ suggested that anaerobic processes, including heterotrophic N₂ fixation, can take place in anoxic microenvironments inside sinking particles even in fully oxygenated marine waters.

The unique interaction we described here between a model diatom and a bacterial community of diazotrophs available as lab culture, avoided environmental sampling related issues allowing high quality sequencing and complete to near-complete assembly of several MAGs with taxonomic assignation to the genus/species level. The high-resolution taxonomic assignation of our MAGs represents a reliable reference for genus assignation of the closely affiliated Tara Oceans MAGs with unknown taxonomic affiliation. The identification of different diazotrophic taxa with a model diatom opens wide perspectives to investigate the role of individual and combined interactions of bacterial species with the host and within the bacterial community. Associated and often overlooked non-diazotrophic bacteria enriched in -N, likely contribute to the symbiosis through different means such as buffering the competition between several diazotrophs in the interaction with the host, producing microalgae growth compounds and keeping away antagonists to secure the interaction with the host. Thus, Pt15 accession with its associated NCDs represent a unique Eukaryotic-bacteria model system and an exciting opportunity to study molecular mechanisms underlying this symbiotic interaction for nitrogen fixation and beyond.

Microalgae species used and culture media

Screening of Phaeodactylum tricornutum accessions

Ten xenic P. tricornutum accessions were screened in presence (545 µM) and absence of nitrate in Enhanced Artificial Sea Water (EASW) in order to select xenic Pt ecotype growing in the absence of nitrate. Starting concentration for all the accessions was 100,000 cells per ml. The cells were counted and washed twice with nitrate free EASW in order to remove all residual nitrate and resuspended in respective media for screening over a period of one month. A growth curve was made by counting the cells in periodic time intervals. The cultures were grown at 19°C with a 12/12 hours photoperiod at 30 to 50 µE/s/m2. A cocktail of antibiotics (Ampicillin 100 µg/µl, Chloramphenicol (14 µg/µl) and Streptomycin 100 µg/µl) was used to make Pt15 axenic in liquid EASW culture medium with refreshment of antibiotic containing medium every 5 days.

Screening of other microalgae

A total of fourteen microalgae were screened in nitrate-deplete and nitrate-replete enhanced artificial sea water for growth (Supplementary Table 5). About 100,000 cells were counted, resuspended in their respective media and grown as described above. Their growth was assessed for a period of two weeks at regular intervals.

Acetylene reduction assay

Acetylene reduction assay (ARA)¹⁷ was used to quantify the conversion of acetylene to ethylene (C₂H₄) which allows the assessment of nitrogenase activity. ARA was conducted on the following samples: two positive controls, Azotobacter. vinelandii (NCIMB 12096), Pseudomonas stutzeri BAL361, xenic Pt15 grown in NFM^50,51 or nitrogen replete medium, all incubated at three different temperatures (19, 25, 30°C). Axenic Pt15 culture grown in NFM medium + 5 sugars (final concentration of 1% each: D-glucose, D-sucrose, D-fructose, D-galactose and D-mannose) was used in every measurement as a negative control to ensure there was no endogenous ethylene produced. All the cultures were grown at their respective growth media and conditions. After reaching an appropriate O.D. between 0.2-0.3, 1 ml of the bacterial cells were collected in an Eppendorf’s tube and centrifuged to 11,000 rpm for 10 minutes at RT to remove the medium. The cells were re-suspended in NFM with five sugars and grown at their respective medium. When the bacterial growth reached an O.D. of 0.2-0.3, the cultures were prepared for acetylene reduction assay. The cultures were secured with a rubber stopper in order to make them air tight. Ten percent of the air was removed from the culture flasks and supplemented with 10% acetylene gas. The cultures were incubated at their appropriate temperatures for a period of 72 hours before monitoring ARA with gas chromatography. A volume of 500 µl were taken from the headspace of the samples using a Hamilton gas tight syringe (SYR 500 ul 750 RN no NDL, NDL large RN 6/pk) and injected into a gas chromatograph 7820 model Agilent equipped with a Flame Ionization Detector (GC-FID) and using Hydrogen as carrier gas at a constant flow of 7.7 mL/min. Sample were injected into a split/splitless injector, set at 160 ̊C in splitless mode. A GS-Alumina capillary column, 50 m × 0.53 mm × 0.25 μm (Agilent, 115-3552) was used. The programmed oven temperature was at 160 ̊C (isotherm program) during 2 min. The FID detector was set at 160 ̊C. A standard curve was made with ethylene gas using 5 volumes (100, 150, 200 µl). The raw data generated from the GC provides value of y, Area [pA*s]. This value is supplemented in the slope-intercept formula, y = mx + c, with ‘m’ and ‘c’ representing slope and constant values respectively, denoted in the calibration curve. The amounts of ethylene were calculated as described previously⁵². Acetylene reduction rates were converted to fixed N using the common 3:1 (C2H2 :N2) molar conversion ratio^53,54.

Particulate organic nitrogen measurements

One liter of 200,000 cells/ml culture of xenic Pt15 was grown in nitrate-free EASW for over a period of 80-days at 19°C and 12h/12h light/dark photo period at 70µE/s/m2. Six time points (day(s)) used for harvesting samples: 1, 4, 12, 59, 76 and 80. The cells were counted at each time point using cell counter. Twenty ml of Pt15 culture were harvested and injected through a sterile (treated at 450 deg/4hrs) Whatman GF/F glass microfibre filter papers to collect Pt15 cells and the filters were dried overnight at 60°C incubator and stored at -80°C. The 20 ml filtrate from the cultures was passed through 0.22 um filters (Sigma ca. GVWP10050), in order to remove bacterial cells and stored at -80°C, before being processed for particulate organic nitrogen measurement analysis. 545 uM nitrate-EASW, 0 uM nitrate-EASW and three heat treated Whatman GF/F glass microfibre filter were used as blanks for comparison with test samples.

Microscopy

Light microscopy

Living cells were directly used for preparations. Samples were observed with differential interference contrast (DIC) microscopy by BX53 microscope (Olympus, Japan) and Axiocam 705 color (Carl Zeiss, Germany).

Transmission electron microscopy

Samples were prepared as described previously ⁵⁵. Embedded samples were cut for making pale yellow ultra-thin sections (70-80nm thickness) with a microtome EM UC6 (Leica Microsystems, Germany). Sections were counter- stained with TI blue (Nisshin EM, Japan) and lead citrate for inspection with an electron microscope JEM-1011KM II (JEOL, Japan).

Bacterial isolation and culture media used

A total of eight media recipe were used for bacteria isolation and culture under different conditions, all summarized in supplementary file 1. All isolated bacteria were checked for purity by repeated streaking and 16S sequencing. Bacterial species were maintained in 20% glycerol at - 80°C for long-term storage.

Determination of nitrogen fixing ability using NfB medium

Bacterial isolates were incubated in liquid nitrogen free medium with bromothymol blue (NfB)⁵⁶ at 30°C for 16 days and monitored for color change every day. Two positive controls, a slow and fast growing strains, respectively, Pseudomonas stutzeri and Azotobacter vinelandii were used. Isolates free NfB medium was used as a negative control. The NfB medium contains a pH sensitive dye, bromothymol blue which switches from green (pH= 6-6.7) to blue (pH>6.7)⁵⁷. The switch in the color is due to nitrogen fixation which consumes H+ protons alkalinizing the medium.

DNA extraction and PCR amplification of marker genes

For initial identification of bacteria, the isolated colonies were picked under sterile conditions using filter tips, resuspended and mixed thoroughly in 50 µl nuclease free water (NFW). The suspension was heat/cold treated at 95°C following 4°C, 10 min each and twice for cell lysis. Post-treatment, an additional 50 ul NFW was added to the suspension for a total volume of 100 µl. Five µl of the colony DNA was used as template for PCR.

Wizard® Promega Genomic DNA extraction kit was used for extracting high-quality genomic DNA from bacterial cells. The bacteria were grown in their appropriate media until the culture optical density (OD) reached 0.3. The cells were harvested and washed in 1X PBS twice. The cell pellets were resuspended in 480 µl 50 mM EDTA and treated with 120 µl of 20mg/ml lyzozyme (VWR cat no. 0663-5G), incubated at 37°C for 60 min in order to lyze possible gram-positive bacteria. Post incubation, the cells were centrifuged and the supernatant was removed. Nuclear Lysis solution, 600 µl (Promega cat no. A7941), was used to re-suspend and gently mix the pellet for incubation at 80°C for 5 min. The suspension was allowed to cool at room temperature (RT) for 10 min before adding 1 µl RNAse at 10 mg/ml (VWR cat no. 0675-250MG) and incubation at 37 for 60 min. After incubation, the suspension was cooled at RT for 10 min. Protein lysis solution, 200 ul (cat no. VWR 0663-5G) was added, vortexed and incubated on ice (4°C for 5 min). The mixture was centrifuged and the clear supernatant was transferred to a clean Eppendorf with 600 µl RT isopropanol and the tube was mixed gently before being centrifuged. The cell pellet was washed with freshly prepared 70% ethanol and re-centrifuged. The ethanol was carefully removed without disturbing the pellet and air dried before resuspension in appropriate amount of NFW. The quality of the DNA was checked using Nanodrop 260/280 and 230/260 ratios and on 1% agarose gel. The identity of the bacteria was re-confirmed using 16S rRNA gene amplification and sanger sequencing.

Three degenerate primers (F1 – nifH3R, nifH2F – nifh3R, PolF – PolR)⁵⁸ for nifH gene detection by PCR were standardized. The PCR was done as follows, 95°C for 5 minutes (1 cycle), followed by 40 cycles of 95°C at 30 seconds, standardized annealing temperature at 30 seconds, elongation at 72°C for 30 seconds and a final extension at 72°C for 7 minutes. The annealing temperature used for F1 – nifH3R, nifH2F – nifh3R, PolF – PolR combinations were 48, 52 and 54°C respectively. Fifty ng of DNA was added to a 20 μl PCR reaction containing, 10X Dream Taq buffer, 0.2 mM dNTPs, 2 pmol primers and 1.25 U of Taq (Fischer Scientific ca. n° 15689374). The reaction was performed with 35 cycles of 95 °C (1 min), 55 °C (1 min) and 72 °C (1 min). For species identification, 16S PCR fd1-rd1, variable regions PCR v1v2 v3v4 and v5v7 were used. All primer sequence details are reported in supplementary Table 7.

Sample preparation and DNA extraction for Illumina and Pacific Pac Bio sequencing

Duplicate with a starting concentration of 200,000 cells/ml of 3-liter culture volumes of xenic Pt15 in nitrate-replete and nitrate-deplete enhanced artificial sea water were grown for a period of 2 months (60-days) at 19°C and 12h/12h light/dark photoperiod at 70µE/s/m2 before collection for sequencing. The xenic Pt15 cells were filtered through a 3 µm filter (MF-millipore hydrophobic nitrocellulose, diameter 47 mm (cat log no. SSWP04700) to retain the cells and allow passage of the filtrate containing the bacterial community. The filtrate was filtered twice in order to avoid contamination of the bacterial cells with residual Pt15 cells. The double filtered bacterial community was collected in 0.22-micron filter (MF-millipore hydrophobic nitrocellulose, diameter 47 mm (cat log no. GSWP04700) and resuspended in 1X PBS. DNA was extracted for metagenome sequencing using Quick-DNA fungal/bacterial mini prep kit (cat no. DG005) and according to manufacturer instructions. The bacterial DNA quality was checked using 1% agarose gel to assess intact DNA and quantified using Qubit broad range DNA kit, before being sequenced. Additionally, 16S rRNA sequencing was performed to assess the presence of bacterial DNA and 18s rRNA PCR was run as a negative control to inspect absence of P. tricornutum DNA. When bacterial isolates are available, they are sequenced either as a single species or in a pool of 3 bacteria using Illumina and PacBio.

Data processing

Metagenomic analysis

Two different strategies were used to analyze the metagenomic reads : (i) a “Single assemblies” strategy, assembly and binning was performed for each sample individually. This strategy is known to produce less fragmented assemblies with a limitation of chimeric contigs formation⁵⁹. (ii) a co-assembly strategy was also performed to complete the results with additional MAGs. The idea is to pool all the available samples reads and build a unique and common assembly for all samples. Samples from 2 experiments were used. An experiment with sequences obtained at 2g/L salinity (one sample in +N condition / one sample in -N condition) and the other experiment at 20 g/L salinity includes duplicates for each of the conditions. Both data sets are similar.

Single assembly strategy

DNA raw reads were analyzed using a dedicated metagenomic pipeline ATLAS v2.7⁵⁹. This tools includes four major steps of shotgun analysis: quality control of the raw reads, assembly, binning, taxonomic and functional annotations. Briefly, quality control of raw sequences is performed using utilities in the BBTools suite (https://sourceforge.net/projects/bbmap/): BBDuk for trimming and removing adapters, BBSplit for decontamination. In our case Phaeodactylum tricornutum reference genome, including mitochondria and chloroplast sequences, was used for this specific step. Quality-control (QC) reads assemblies were performed using metaSPADES⁶⁰ with ATLAS default parameters. The assembled contigs shorter than 500 bases length and without mapped reads, were removed to keep high-quality contigs.

Metagenome-assembled genomes (MAGs) were obtained using two distinct binners : maxbin2⁶¹ and metabat2⁶². The bins obtained by the different binning tools were combined using DASTool⁶³ and dereplicated with dRep⁶⁴ using default parameters. Completeness and contamination were computed using CheckM⁶⁵. Only bins with a completeness above 50 % and a contamination rate under 10 % were kept. Prediction of open reading frames (ORFs) was performed, at the assembly level (not only at the MAGs level) using Prodigal⁶⁶. Linclust⁶⁷ was used to cluster the translated gene product obtained to generate gene and protein catalogs common to all samples.

Co-assembly strategy

QC reads and co-assembly were analyzed using a shotgun dedicated sequential pipeline MetaWRAP v1.3.2⁶³. Briefly, QC reads are pulled together and addressed to the chosen assembler MetaSPADES using MetaWRAP default parameters. Metagenome-assembled genomes (MAGs) were obtained using two distinct binners: maxbin2 and metabat2 and the bins obtained were combined using an internal MetaWRAP dedicated tool. Completeness and contamination were computed using CheckM⁶⁵. Only bins with a completeness above 50 % and a contamination rate under 10 % were kept. MetaWRAP also includes a re-assemble bins module that collect reads belonging to each bin, and then reassemble them independently with a "permissive" and a "strict" algorithm. Only improved bins are kept in the final set. Using this strategy, we were able to add an additional MAG.

MAGs relative abundances

The QC reads were mapped back to the contigs, and bam files were generated to organize downstream analysis. Median coverage of the genomes was computed to evaluate MAGs relative abundance estimations.

Taxonomic assignation of metagenomic data

We utilized Anvio version 7.2 (https://anvio.org) for taxanomic profiling of the P. tricornutum associated metagenomes in nitrate deplete and replete condition. The workflow uses The Genome Taxonomy Database (GTDB, doi:10.1038/nbt.4229) to determine the taxanomy based on 22 single copy core genes (scgs, anvio/anvio/data/misc/SCG_TAXONOMY). Sequence alignment search is done using DIAMOND (doi:10.1038/nmeth.3176).

Hybrid assembly of MAGs

Assembly
Metagenomic assembly of Illumina technology reads does not yield complete
genomes due to the difficulty of assembling repetitive regions⁶⁸. Whenever Illumina and Pac bio reads are available, we used a strategy to improve MAGs assembly by combining both type of reads. Metagenomic assembly of Illumina reads paired
end and PacBio reads post quality trimming, were assembled using metaSPAdes or
SPAdes(v3.15.5) with option flag (--meta). To validate the newly assembled
genomes, we verified that the long reads covered regions that were not supported by
the mapped Illumina sequences only and that we get high quality bins. We obtained
genomes in which some regions had not been supported by the assembly of Illumina
reads only.

Binning
The contigs were grouped and assigned to the individual genome. To do this, we used MaxBin2 and metaBAT2 which are clustering methods based on compositional features or on alignment (similarity), or both (https://anaconda.org/bioconda/maxbin2 and https://anaconda.org/bioconda/metabat2)⁶⁹.

Refinement
Bins from metabat2 and maxbin2, were consolidated into one more solid set of bins. To do this we used Binning_refiner which improves genomic bins by combining different binning programs available at:https://github.com/songweizhi/Binning_refiner ⁷⁰.

Taxonomic Assignation

GTDB-Tk has already been independently and positively evaluated for the classification of MAGs⁷¹. After consolidating the bins in the refinement step, we determined the taxonomy of each bin. The gtdbtk classify_wf module was used in this work using a GTDB database https://github.com/ecogenomics/gtdbtk ⁷².

Gene annotation of metagenomic data and MAGs

Gene annotation of metagenomics data in two conditions was done using Anvio version 7.2. Three databases were used for a robust gene observations: COG20 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4383993,http://www.ncbi.nlm.nih.gov/COG/), KEGG and PROKKA (https://github.com/tseemann/prokka). DIAMOND to search NCBI’s Database. In built, Anvio program annotates a contig database with HMM hits from KOfam, a database of KEGG Orthologs (KOs, https://www.genome.jp/kegg/ko.html). For PROKKA, the annotation was done externally and the gene calls were imported to Anvio platform for the individual metagenome contigs databases.

Phylogenomic analysis of MAGs

The phylogenomic tree was built using Anvio-7.1 (https://github.com/merenlab/anvio/releases) with database of 1888 TARA ocean MAGs (contig-level FASTA files) submitted to Genoscope (https://www.genoscope.cns.fr/tara/). Additional closest terrestrial and marine genomes (supplementary table 6) to the individual MAGs were selected using MicroScope database (mage.genoscope.cns.fr/microscope/) using genome clustering, utilizing MASH (https://github.com/marbl/Mash) for ANI (Average nucleotide identity) for calculating pairwise genomic distancing, neighbour-joining (https://www.npmjs.com/package/neighbor-joining) for tree-construction and computing the clustering using Louvain Community Detection (https://github.com/taynaud/python-louvain). Anvi'o utilizes ‘prodigal’⁶⁶ to identify open reading frames for the contig database. The HMM profiling for 71 single core-copy bacterial genes (SCGs) is done using in-built Anvi'o database, Bacteria_71 (https://doi.org/10.1093/bioinformatics/btz188) and utilizes MUSCLE for sequence alignment (http://www.drive5.com/muscle). The bootstrap values are represented from 0 – 1 with replicates of 100.

Plasmid prediction and annotation

Plasmids were predicted from MetaG data using the following plasmid assembly tools: SCAPP⁷³ (Sequence Contents-Aware Plasmid Peeler), https://github.com/Shamir-Lab/SCAPP, metaplasmidSPAdes⁷⁴ (GitHub - ablab/spades: SPAdes Genome Assembler) and PlaScope⁷⁵ (GitHub - labgem/PlaScope: Plasmid exploration of bacterial genomes). Predicted plasmids were annotated using Dfast⁷⁶ (https://github.com/nigyta/dfast_core), eggNOG-mapper v2⁷⁷ and PROKKA 1.14.6.

Tara Oceans metagenome-assembled genomes counts tables

MAGs abundance count tables and fasta files were downloaded from open-source released Tara Oceans databases at <https://www.genoscope.cns.fr/tara/> ^13,78. Data included 2601 MAGs of which 1888 prokaryotes and 713 eukaryotes MAGs in 937 samples at 2 depths (Surface and DCM) and 6 size fractions (0.22-3μm, 3-20μm, 0.8-5μm, 20-200μm, 0.8-2000μm, 180-2000μm).

Biogeography of metagenome-assembled genomes

Total sum scaling (TSS) normalisation was performed on reads counts mapped to MAGs, by dividing all counts by the total number of reads in a given sample. Different depths (surface and DCM) and size fractions (0.22-1.6/3µm, 3/5-20µm, 0.8-5µm, 20-180/200µm, 0.8-2000µm, 180-2000µm) were analyzed separately to project relative abundances on global maps. Relative abundances of Tara Oceans MAGs phylogenetically close to Pt15 and belonging to the same cluster were summed up for each sample. Sizes of the pie charts reflect the total amount of relative abundances present in each sample.

Environmental parameters preferences

Tara Oceans physico-chemical parameters were downloaded from Pangaea³⁷. MAGs TSS relative abundances were projected and analyzed in relationship with ammonium (µmol/L), chlorophyll A (mg/m3), iron (µmol/L), nitrite (µmol/L), nitrate (µmol/L), phosphate (µmol/L), nitrate/phosphate, oxygen (µmol/kg), salinity (PSS-78), silicate (µmol/L), temperature (°C) and photosynthetically available radiation (mol quanta/m2/day).

Co-occurrence network inference

Co-occurrence network inference was performed using FlashWeave v0.18.0⁷⁹ with Julia v1.5.3, for eukaryotic and prokaryotic Tara MAGs linked to Pt15 MAGs. To account for the compositional nature of MAGs abundance data, a multiplicative replacement method⁸⁰ was applied to replace zeros, and a Centered Log Ratio transformation⁸¹ was applied separately on both prokaryotic and eukaryotic abundance matrices. FlashWeave parameters used were max_k=3, n_obs_min=10, and normalize=False.

Acknowledgements

We acknowledge Clara Guillouche and Irene Romero Rodriguez for their technical assistance with PCR screening and Romain le Balch for his help with GC runs. LT acknowledges support from the region of Pays de la Loire (ConnecTalent EPIALG project), Epicycle ANR project (ANR-19-CE20- 0028-02) and µAlgaNIF France-Japan International Research Project. UC was supported by grant 998UMR6286 Connect Talent EpiAlg from Région Pays de la Loire to LT. SC acknowledges support from the CNRS MITI through the interdisciplinary program Modélisation du Vivant (GOBITMAP grant), and the H2020 European Commission project AtlantECO (award number 862923). We are grateful to the bioinformatics core facility of Nantes University (BiRD Biogenouest) for its technical support. The LABGeM (CEA/Genoscope &CNRS UMR8030), the France Génomique and French Bioinformatics Institute national infrastructures (funded as part of Investissement d’Avenir program managed by Agence Nationale pour la recherche, contrats ANR-10-INBS-09 and ANR-11-INBS-0013) are acknowledged for support within the Microscope annotation platform.

1 Moore, C. M., Mills, M.M., Arrigo, K.R. I. Berman-Frank, L. Bopp, P. W. Boyd, E. D. Galbraith, , R. J. Geider, C. G., S. L. Jaccard, T. D. Jickells, J. La Roche, T. M. Lenton, N. M. Mahowald, , E. Marañón, I. M., J. K. Moore, T. Nakatsuka, A. Oschlies, M. A. Saito, T. F. Thingstad, & Ullo, A. T. O. Processes and patterns of oceanic

nutrient limitation. Nature Geoscience 6, 701-710 (2013).

2 Lee, C. C., Ribbe, M. W. & Hu, Y. Cleaving the n,n triple bond: the transformation of dinitrogen to ammonia by nitrogenases. Met Ions Life Sci 14, 147-176 (2014). https://doi.org:10.1007/978-94-017-9269-1_7

3 Einsle, O. & Rees, D. C. Structural Enzymology of Nitrogenase Enzymes. Chem Rev 120, 4969-5004 (2020). https://doi.org:10.1021/acs.chemrev.0c00067

4 Kneip, C., Lockhart, P., Voss, C. & Maier, U. G. Nitrogen fixation in eukaryotes--new models for symbiosis. BMC Evol Biol 7, 55 (2007). https://doi.org:10.1186/1471-2148-7-55

5 Fiore, C. L., Jarett, J. K., Olson, N. D. & Lesser, M. P. Nitrogen fixation and nitrogen transformations in marine symbioses. Trends Microbiol 18, 455-463 (2010). https://doi.org:10.1016/j.tim.2010.07.001

6 Cornejo-Castillo, F. M. & Zehr, J. P. Hopanoid lipids may facilitate aerobic nitrogen fixation in the ocean. Proc Natl Acad Sci U S A 116, 18269-18271 (2019). https://doi.org:10.1073/pnas.1908165116

7 Inomura, K. et al. Quantifying Oxygen Management and Temperature and Light Dependencies of Nitrogen Fixation by Crocosphaera watsonii. mSphere 4 (2019). https://doi.org:10.1128/mSphere.00531-19

8 Inomura, K., Wilson, S. T. & Deutsch, C. Mechanistic Model for the Coexistence of Nitrogen Fixation and Photosynthesis in Marine Trichodesmium. mSystems 4 (2019). https://doi.org:10.1128/mSystems.00210-19

9 Munoz-Marin, M. D. C. et al. The Transcriptional Cycle Is Suited to Daytime N2 Fixation in the Unicellular Cyanobacterium "Candidatus Atelocyanobacterium thalassa" (UCYN-A). mBio 10 (2019). https://doi.org:10.1128/mBio.02495-18

10 Schlesier, J., Rohde, M., Gerhardt, S. & Einsle, O. A Conformational Switch Triggers Nitrogenase Protection from Oxygen Damage by Shethna Protein II (FeSII). J Am Chem Soc 138, 239-247 (2016). https://doi.org:10.1021/jacs.5b10341

11 Bothe, H., Tripp, H. J. & Zehr, J. P. Unicellular cyanobacteria with a new mode of life: the lack of photosynthetic oxygen evolution allows nitrogen fixation to proceed. Arch Microbiol 192, 783-790 (2010). https://doi.org:10.1007/s00203-010-0621-5

12 Delmont, T. O. et al. Nitrogen-fixing populations of Planctomycetes and Proteobacteria are abundant in surface ocean metagenomes. Nat Microbiol 3, 804-813 (2018). https://doi.org:10.1038/s41564-018-0176-9

13 Delmont, T. O. P. K., J.S. Veseli, I., Fuessel, J. Murat Eren, A., Foster, R.A., Bowler, C., Wincker, P., Pelletier, E. Heterotrophic bacterial diazotrophs are more abundant than their cyanobacterial counterparts in metagenomes covering most of the sunlit ocean. doi: https://doi.org/10.1101/2021.03.24.436778 (2021).

14 Farnelid, H. et al. Nitrogenase gene amplicons from global marine surface waters are dominated by genes of non-cyanobacteria. PLoS One 6, e19223 (2011). https://doi.org:10.1371/journal.pone.0019223

15 Moisander, P. H. et al. Chasing after Non-cyanobacterial Nitrogen Fixation in Marine Pelagic Environments. Front Microbiol 8, 1736 (2017). https://doi.org:10.3389/fmicb.2017.01736

16 Harding, K. J. et al. Cell-specific measurements show nitrogen fixation by particle-attached putative non-cyanobacterial diazotrophs in the North Pacific Subtropical Gyre. Nat Commun 13, 6979 (2022). https://doi.org:10.1038/s41467-022-34585-y

17 Sullivan, B. W. et al. Spatially robust estimates of biological nitrogen (N) fixation imply substantial human alteration of the tropical N cycle. Proc Natl Acad Sci U S A 111, 8101-8106 (2014). https://doi.org:10.1073/pnas.1320646111

18 de Lajudie, P. et al. Characterization of tropical tree rhizobia and description of Mesorhizobium plurifarium sp. nov. Int J Syst Bacteriol 48 Pt 2, 369-382 (1998). https://doi.org:10.1099/00207713-48-2-369

19 Yang, X. et al. Mesorhizobium alexandrii sp. nov., isolated from phycosphere microbiota of PSTs-producing marine dinoflagellate Alexandrium minutum amtk4. Antonie Van Leeuwenhoek 113, 907-917 (2020). https://doi.org:10.1007/s10482-020-01400-x

20 Zhang, X., Tong, J., Dong, M., Akhtar, K. and He, B. Isolation, identification and characterization of nitrogen fixing endophytic bacteria and their effects on cassava production. PeerJ. 10: e12677 10 (2022).

21 Bostrom, K. H., Riemann, L., Kuhl, M. & Hagstrom, A. Isolation and gene quantification of heterotrophic N2-fixing bacterioplankton in the Baltic Sea. Environ Microbiol 9, 152-164 (2007). https://doi.org:10.1111/j.1462-2920.2006.01124.x

22 Sy, A. et al. Methylotrophic Methylobacterium bacteria nodulate and fix nitrogen in symbiosis with legumes. J Bacteriol 183, 214-220 (2001). https://doi.org:10.1128/JB.183.1.214-220.2001

23 Medigue, C. et al. MicroScope-an integrated resource for community expertise of gene functions and comparative analysis of microbial genomic and metabolic data. Brief Bioinform 20, 1071-1084 (2019). https://doi.org:10.1093/bib/bbx113

24 Lesser, M. P., Morrow, K. M., Pankey, S. M. & Noonan, S. H. C. Diazotroph diversity and nitrogen fixation in the coral Stylophora pistillata from the Great Barrier Reef. ISME J 12, 813-824 (2018). https://doi.org:10.1038/s41396-017-0008-6

25 Videira, S. S., de Araujo, J. L., Rodrigues Lda, S., Baldani, V. L. & Baldani, J. I. Occurrence and diversity of nitrogen-fixing Sphingomonas bacteria associated with rice plants grown in Brazil. FEMS Microbiol Lett 293, 11-19 (2009). https://doi.org:10.1111/j.1574-6968.2008.01475.x

26 Skorupska, A., Janczarek, M., Marczak, M., Mazur, A. & Krol, J. Rhizobial exopolysaccharides: genetic control and symbiotic functions. Microb Cell Fact 5, 7 (2006). https://doi.org:10.1186/1475-2859-5-7

27 Hernandez, J. A. et al. NifX and NifEN exchange NifB cofactor and the VK-cluster, a newly isolated intermediate of the iron-molybdenum cofactor biosynthetic pathway. Mol Microbiol 63, 177-192 (2007). https://doi.org:10.1111/j.1365-2958.2006.05514.x

28 Jasniewski, A. J., Lee, C. C., Ribbe, M. W. & Hu, Y. Reactivity, Mechanism, and Assembly of the Alternative Nitrogenases. Chem Rev 120, 5107-5157 (2020). https://doi.org:10.1021/acs.chemrev.9b00704

29 Wang, T. et al. Structure of a bacterial energy-coupling factor transporter. Nature 497, 272-276 (2013). https://doi.org:10.1038/nature12045

30 Geiger, O. & Lopez-Lara, I. M. Rhizobial acyl carrier proteins and their roles in the formation of bacterial cell-surface components that are required for the development of nitrogen-fixing root nodules on legume hosts. FEMS Microbiol Lett 208, 153-162 (2002). https://doi.org:10.1111/j.1574-6968.2002.tb11075.x

31 Cook, A. M. & Denger, K. Metabolism of taurine in microorganisms: a primer in molecular biodiversity? Adv Exp Med Biol 583, 3-13 (2006). https://doi.org:10.1007/978-0-387-33504-9_1

32 Boutte, C. C. & Crosson, S. Bacterial lifestyle shapes stringent response activation. Trends Microbiol 21, 174-180 (2013). https://doi.org:10.1016/j.tim.2013.01.002

33 Stuffle, E. C., Johnson, M. S. & Watts, K. J. PAS domains in bacterial signal transduction. Curr Opin Microbiol 61, 8-15 (2021). https://doi.org:10.1016/j.mib.2021.01.004

34 Madigan, M., Cox, S. S. & Stegeman, R. A. Nitrogen fixation and nitrogenase activities in members of the family Rhodospirillaceae. J Bacteriol 157, 73-78 (1984). https://doi.org:10.1128/jb.157.1.73-78.1984

35 Singh, R. K. et al. Unraveling Nitrogen Fixing Potential of Endophytic Diazotrophs of Different Saccharum Species for Sustainable Sugarcane Growth. International Journal of Molecular Sciences 23, 6242 (2022).

36 Sellstedt, A., Richau, K.H. Aspects of nitrogen-fixing Actinobacteria, in particular free-living and symbiotic Frankia. FEMS Microbiology Letters 342, 179–186 (2013).

37 Pesant, S. et al. Open science resources for the discovery and analysis of Tara Oceans data. Sci Data 2, 150023 (2015). https://doi.org:10.1038/sdata.2015.23

38 Karsenti, E. et al. A holistic approach to marine eco-systems biology. PLoS biology 9, e1001177 (2011). https://doi.org:10.1371/journal.pbio.1001177

39 Sanudo-Wilhelmy, S. A. et al. Phosphorus limitation of nitrogen fixation by Trichodesmium in the central Atlantic Ocean. Nature 411, 66-69 (2001). https://doi.org:10.1038/35075041

40 Ward BA, D. S., Moore CM, Follows MJ. Iron, phosphorus, and nitrogen supply ratios define the biogeography of

nitrogen fixation. . Limnol Oceanogr. 58, 2059–2075 (2013).

41 Bombar, D., Paerl, R. W. & Riemann, L. Marine Non-Cyanobacterial Diazotrophs: Moving beyond Molecular Detection. Trends Microbiol 24, 916-927 (2016). https://doi.org:10.1016/j.tim.2016.07.002

42 Geisler, E., Bogler, A., Rahav, E. & Bar-Zeev, E. Direct Detection of Heterotrophic Diazotrophs Associated with Planktonic Aggregates. Sci Rep 9, 9288 (2019). https://doi.org:10.1038/s41598-019-45505-4

43 Ochman, H. & Moran, N. A. Genes lost and genes found: evolution of bacterial pathogenesis and symbiosis. Science 292, 1096-1099 (2001). https://doi.org:10.1126/science.1058543

44 Farnelid, H. et al. Active nitrogen-fixing heterotrophic bacteria at and below the chemocline of the central Baltic Sea. ISME J 7, 1413-1423 (2013). https://doi.org:10.1038/ismej.2013.26

45 Hamersley R, T. K., Leinweber A, Gruber N, Zehr JP et al . . Nitrogen fixation within the water column associated with two hypoxic basins within the Southern California Bight. Aquat Microb Ecol. 63, 193–205 (2011).

46 Wang, D., Xu, A., Elmerich, C. & Ma, L. Z. Biofilm formation enables free-living nitrogen-fixing rhizobacteria to fix nitrogen under aerobic conditions. ISME J 11, 1602-1613 (2017). https://doi.org:10.1038/ismej.2017.30

47 Maier, R. J. & Moshiri, F. Role of the Azotobacter vinelandii nitrogenase-protective shethna protein in preventing oxygen-mediated cell death. J Bacteriol 182, 3854-3857 (2000). https://doi.org:10.1128/JB.182.13.3854-3857.2000

48 Brian, A. C. J., C. W. . The respiratory system of Azotobacter vinelandii. Eur. J. Biochem. 20, 29–35 (1971).

49 Chakraborty, S. et al. Quantifying nitrogen fixation by heterotrophic bacteria in sinking marine particles. Nat Commun 12, 4085 (2021). https://doi.org:10.1038/s41467-021-23875-6

50 Jorgensen, J. H., Pfaller, M.A., Carroll, K.C., Funke, G., Landry, M.L., Richter, S.S and Warnock., D.W. . Manual of Clinical Microbiology. 11th edn, Vol. 1 (2015).

51 Isenberg, H. D. Clinical Microbiology Procedures Handbook. 2nd Edition edn, Vol. 1 (Amer Society for Microbiology, 1992).

52 Paudel, D. et al. Isolation, Characterization, and Complete Genome Sequence of a Bradyrhizobium Strain Lb8 From Nodules of Peanut Utilizing Crack Entry Infection. Front Microbiol 11, 93 (2020). https://doi.org:10.3389/fmicb.2020.00093

53 Seitzinger, S. P., and Garber, J. H. . Nitrogen-fixation and N-15(2) calibration

of the acetylene-reduction assay in coastal marine-sediments. . Mar. Ecol. Progr. Ser 65–73, 65–73 (1987).

54 Hardy RWF, H. R., Jackson EK, Burns RC. . The

Acetylene-Ethylene Assay for N2 fixation: laboratory and

field evaluation. Plant Phys 43, 1185–1207 (1968).

55 Tanaka A, D. M. A., Amato A, Montsant A, Mathieu B, Rostaing P, Tirichine L & Bowler C. . Ultrastructure and Membrane Traffic During Cell Division in the Marine Pennate Diatom Phaeodactylum tricornutum. Protist 166, 506-521 (2015).

56 Baldani, J. I., Reis, V.M., Videira, S.S. et al. The art of isolating nitrogen-fixing bacteria from non-leguminous plants using N-free semi-solid media: a practical guide for microbiologists. Plant Soil 384, 413–431

57 Jha, C. K., Patel, D., Rajendran, N. & Saraf, M. Combinatorial assessment on dominance and informative diversity of PGPR from rhizosphere of Jatropha curcas L. J Basic Microbiol 50, 211-217 (2010). https://doi.org:10.1002/jobm.200900272

58 Gaby, J. C. & Buckley, D. H. A comprehensive aligned nifH gene database: a multipurpose tool for studies of nitrogen-fixing bacteria. Database (Oxford) 2014, bau001 (2014). https://doi.org:10.1093/database/bau001

59 Kieser, S., Brown, J., Zdobnov, E. M., Trajkovski, M. & McCue, L. A. ATLAS: a Snakemake workflow for assembly, annotation, and genomic binning of metagenome sequence data. BMC Bioinformatics 21, 257 (2020). https://doi.org:10.1186/s12859-020-03585-4

60 Nurk, S., Meleshko, D., Korobeynikov, A. & Pevzner, P. A. metaSPAdes: a new versatile metagenomic assembler. Genome Res 27, 824-834 (2017). https://doi.org:10.1101/gr.213959.116

61 Wu, Y. W., Simmons, B. A. & Singer, S. W. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics 32, 605-607 (2016). https://doi.org:10.1093/bioinformatics/btv638

62 Kang, D. D. et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ 7, e7359 (2019). https://doi.org:10.7717/peerj.7359

63 Sieber, C. M. K. et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy. Nat Microbiol 3, 836-843 (2018). https://doi.org:10.1038/s41564-018-0171-1

64 Olm, M. R., Brown, C. T., Brooks, B. & Banfield, J. F. dRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication. ISME J 11, 2864-2868 (2017). https://doi.org:10.1038/ismej.2017.126

65 Parks, D. H., Imelfort, M., Skennerton, C. T., Hugenholtz, P. & Tyson, G. W. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res 25, 1043-1055 (2015). https://doi.org:10.1101/gr.186072.114

66 Hyatt, D. et al. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics 11, 119 (2010). https://doi.org:10.1186/1471-2105-11-119

67 Steinegger, M. & Soding, J. Clustering huge protein sequence sets in linear time. Nat Commun 9, 2542 (2018). https://doi.org:10.1038/s41467-018-04964-5

68 Giguere, D. J., Bahcheli, A.T., Joris, B.R., Paulssen, J.M., Gieg, L.M., Flatley, M.W., Gloor, G.B. Complete and validated genomes from a metagenome. doi: https://doi.org/10.1101/2020.04.08.032540 (2020).

69 Meyer, F. et al. Tutorial: assessing metagenomics software with the CAMI benchmarking toolkit. Nat Protoc 16, 1785-1801 (2021). https://doi.org:10.1038/s41596-020-00480-3

70 Song, W. Z. & Thomas, T. Binning_refiner: improving genome bins through the combination of different binning programs. Bioinformatics 33, 1873-1875 (2017). https://doi.org:10.1093/bioinformatics/btx086

71 Coil, D. A. et al. Genomes from bacteria associated with the canine oral cavity: A test case for automated genome-based taxonomic assignment. PLoS One 14, e0214354 (2019). https://doi.org:10.1371/journal.pone.0214354

72 Chaumeil, P. A., Mussig, A. J., Hugenholtz, P. & Parks, D. H. GTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database. Bioinformatics (2019). https://doi.org:10.1093/bioinformatics/btz848

73 Pellow, D. et al. SCAPP: an algorithm for improved plasmid assembly in metagenomes. Microbiome 9, 144 (2021). https://doi.org:10.1186/s40168-021-01068-z

74 Antipov, D., Raiko, M., Lapidus, A. & Pevzner, P. A. Plasmid detection and assembly in genomic and metagenomic data sets. Genome Res 29, 961-968 (2019). https://doi.org:10.1101/gr.241299.118

75 Royer, G. et al. PlaScope: a targeted approach to assess the plasmidome from genome assemblies at the species level. Microb Genom 4 (2018). https://doi.org:10.1099/mgen.0.000211

76 Tanizawa, Y., Fujisawa, T., Kaminuma, E., Nakamura, Y. & Arita, M. DFAST and DAGA: web-based integrated genome annotation tools and resources. Biosci Microbiota Food Health 35, 173-184 (2016). https://doi.org:10.12938/bmfh.16-003

77 Cantalapiedra, C. P., Hernandez-Plaza, A., Letunic, I., Bork, P. & Huerta-Cepas, J. eggNOG-mapper v2: Functional Annotation, Orthology Assignments, and Domain Prediction at the Metagenomic Scale. Mol Biol Evol (2021). https://doi.org:10.1093/molbev/msab293

78 Delmont, T., Gaia, M., Hinsinger DD., et. al. . Functional repertoire convergence of distantly related eukaryotic plankton lineages revealed by genome-resolved metagenomics. Cell Genomics 2 (2022).

79 Tackmann, J., Matias Rodrigues, J. F. & von Mering, C. Rapid Inference of Direct Interactions in Large-Scale Ecological Networks from Heterogeneous Microbial Sequencing Data. Cell Syst 9, 286-296 e288 (2019). https://doi.org:10.1016/j.cels.2019.08.002

80 Fernández, J. A. M., Vidal, C.B.& Glahn, V.P. . Dealing with Zeros and Missing Values in Compositional Data Sets Using Nonparametric Imputation. Mathematical Geology 35, pages253–278 (2003).

81 Quinn, T. P. et al. A field guide for the compositional analysis of any-omics data. Gigascience 8 (2019). https://doi.org:10.1093/gigascience/giz107

Table 1 is available in the supplementary files section.

There is NO Competing Interest.

Table1.ods
Table 1
SupplementaryTable1.xls
Supplementary Dataset 1
SupplementaryFig1.pdf
Supplementary Dataset 1
SupplementaryFig2.pdf
Supplementary Dataset 2
SupplementaryTable2.xls
Supplementary Dataset 2
SupplementaryFig3.pdf
Supplementary Dataset 3
SupplementaryTable3.xlsx
Supplementary Dataset 3
SupplementaryTable4.xls
Supplementary Dataset 4
SupplementaryFig4.pdf
Supplementary Dataset 4
SupplementaryTable5.xlsx
Supplementary Dataset 5
SupplementaryTable6.xlsx
Supplementary Dataset 6
SupplementaryTable7.xls
Supplementary Dataset 7
SupplementaryfiguresTableslegend.docx
Supplementary figures and tables legend
SupplementaryFile1.doc
Supplementary Info 1

Download PDF

Version 1

posted

You are reading this latest preprint version

Combined in vivo and in situ genome-resolved metagenomics reveals novel symbiotic nitrogen fixing interactions between non-cyanobacterial diazotrophs and microalgae

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Methods

Declarations

Acknowledgements

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1