Genetic and physiological insights into the diazotrophic activity of a non-cyanobacterial marine diazotroph

Abstract

Nitrogen (N₂ ) fixation, or diazotrophy, supports a large part of primary production in oceans. Culture-independent approaches highlighted the presence in abundance of marine non-cyanobacterial diazotrophs (NCD), but their ecophysiology remains elusive, mostly because of the low number of isolated NCD and because of the lack of available genetic tools for these isolates. Here, a dual genetic and functional approach allowed unveiling the ecophysiology of a marine NCD affiliated to the species Vibrio diazotrophicus. Physiological characterization of the first marine NCD mutant obtained so far was performed using a soft-gellan assay, demonstrating that a ΔnifH mutant is not able to grow in nitrogen-free media. Furthermore, we demonstrated that V. diazotrophicus produces a thick biofilm under diazotrophic conditions, suggesting biofilm production as an adaptive response of this NCD to cope with the inhibition of nitrogen fixation by molecular oxygen. Finally, the genomic signature of V. diazotrophicus is essentially absent from metagenomic data of Tara Ocean expeditions, despite having been isolated from various marine environments. We think that the genetically tractable V. diazotrophicus strain used in this study may serve as an ideal model to study the ecophysiology of these overlooked procaryotic group.

FIGURE 1

Microscopy picture showing the obtention of V. diazotrophicus carrying the GFP‐expressing plasmid pFD086. Images were acquired from LB‐grown overnight cultures.

FIGURE 2

Genetic and physiological characterization of the V. diazotrophicus NS1 and the ΔnifH mutant. (A) Gel electrophoresis showing the deletion of nifH in V. diazotrophicus NS1 (Lanes 2 and 4). (B) Generation time of V. diazotrophicus NS1 and derivative mutants. Vibrio diazotrophicus NS1 and V. diazotrophicus NS1 ΔnifH contained the pFD085 plasmid and V. diazotrophicus NS1 compl‐nifH corresponds to V. diazotrophicus NS1 ΔnifH containing pFD120. A one‐way ANOVA with Tukey test was done to compare the different mutants. Depicted here are the mean + SEM of four replicates.

FIGURE 3

Soft‐gellan assay demonstrating bacterial growth under appropriate microoxic conditions. The ‘+’ and ‘−’indicate the presence or absence of NH₄ ⁺, respectively. White arrows indicate the location of the growing ring. Note the absence of ring in the ΔnifH mutant and in the presence of NH₄ ⁺. Similar results were obtained using NO₃ ⁻ as a bioavailable inorganic nitrogen source (see Figure S2).

FIGURE 4

Biofilm formation characteristics of V. diazotrophicus NS1. (A) Corrected biofilm formation in V. diazotrophicus NS1 and derivative mutants in different media. Vibrio diazotrophicus NS1 and V. diazotrophicus NS1 ΔnifH contained the pFD085 plasmid and V. diazotrophicus NS1 compl‐nifH corresponds to V. diazotrophicus NS1 ΔnifH containing pFD120. Presented here are the mean + SEM of at least three replicates. Significance was assessed by one‐way ANOVA with Tukey test for each mutant in the different media. (B) View of the top of the biofilm produced by V. diazotrophicus NS1 pFD086 after 24 h of incubation in a flow cell. (C) Stacked microscopy image showing the thickness of the biofilm produced after 24 h. (D) Average and maximal thickness of the biofilm produced by V. diazotrophicus NS1 pFD086 after 24 h. See Figure S4 for images corresponding to timepoint 48 h.

FIGURE 5

Comparative genomics of V. diazotrophicus NS1. (A) Phylogenomic tree of V. diazotrophicus NS1 and the 40 diazotroph MAGs from (Delmont et al., 2022) using 16 single‐copy core genes occurring in at least 40 of the 41 genomes. We used FastTree to compute a phylogenomic tree and anvi'o (Eren et al., 2021) for the visualization. (B) Pangenomics analysis of publicly available V. diazotrophicus. A total of 6928 gene clusters are represented in the figure with each darker colour in the genome's layer indicating the presence of gene cluster in the corresponding genome. The red selection highlights the 2884 single‐copy core genes. The heatmap corresponds to the ANI computed between all genomes and it is scaled from below 0.95 ANI (white) to 1 ANI (dark red). We generated a phylogenomics tree (shown above the ANI heatmap) using a subset of set of 99 single copy core gene with a maximum functional homogeneity of 0.95 using anvi'o. (C) Conserved synteny of the nif gene locus across multiple V. diazotrophicus genomes. The scale is based on the total size of the locus in V. diazotrophicus NS1. The double bars indicate that the locus continues on a different contig. Acc. Num: 65.7 M (POSL00000000), 65.10 M (POSM00000000), 60.27F (POSK00000000), 60.18 M (POSJ00000000), 60.6F (POSI00000000), 60.6B (POSH00000000), HF9B (JAATOR000000000) and 99A (PRJNA456207)