Marine biofilms constitute a bank of hidden microbial diversity and functional potential

Abstract

Recent big data analyses have illuminated marine microbial diversity from a global perspective, focusing on planktonic microorganisms. Here, we analyze 2.5 terabases of newly sequenced datasets and the Tara Oceans metagenomes to study the diversity of biofilm-forming marine microorganisms. We identify more than 7,300 biofilm-forming 'species' that are undetected in seawater analyses, increasing the known microbial diversity in the oceans by more than 20%, and provide evidence for differentiation across oceanic niches. Generation of a gene distribution profile reveals a functional core across the biofilms, comprised of genes from a variety of microbial phyla that may play roles in stress responses and microbe-microbe interactions. Analysis of 479 genomes reconstructed from the biofilm metagenomes reveals novel biosynthetic gene clusters and CRISPR-Cas systems. Our data highlight the previously underestimated ocean microbial diversity, and allow mining novel microbial lineages and gene resources.

Fig. 1

Comparative taxonomic analysis of the 101 biofilm and 24 seawater metagenomes sequenced as part of the current study and the 67 previously published Tara Oceans metagenomes. a Venn diagram showing the distribution of all OTUs. b Venn diagram showing the distribution of abundant OTUs (OTUs with more than nine tags in at least one sample). c Phylum-level taxonomic structure revealing marked differences in the community profiles between the biofilm and seawater habitats. Relative abundance of the phyla was calculated based on 16S miTag numbers. Abundant phyla (the top 30 phyla in terms of maximum relative abundance) are shown with all other phyla grouped together as “Minor”. The order of samples in this figure is the same as that in Supplementary Data 1

Fig. 2

Venn diagram showing the distribution of OTUs across the biofilms in the current study (blue) and the Tara miTags (red) comprising 16S rRNA sequences from 243 seawater samples

Fig. 3

Alpha- and beta-diversity. a ACE, Chao1 diversity, and observed OTUs are significantly different between the biofilm-associated (blue) and seawater-derived (red) microbial communities. ***p-value < 0.001 (two-tailed Student’s t-tests after Shapiro–Wilk test to confirm the normal distribution of these data). In a boxplot, central line represents the median, bounds represent upper and lower quartiles, and whiskers represent maximum and minimum. b Jaccard similarity of the microbial communities illustrated by PCoA of the OTU matrix. These analyses were performed after normalizing the different metagenome-derived 16S rRNA gene data to the same library size, i.e. 10,000 sequences per sample

Fig. 4

Functional structuring of the biofilm core gene catalog (BCGC). a Taxonomic breakdown of BCGC genes, 97.3% of which belong to bacteria. b BCGC is a nonredundant database comprising genes that are present in more than 99 biofilms but which were not detected in any of the seawater samples. Venn diagram showing the total number of genes in the BCGC and the number of genes annotated by three databases. 61.1% of the BCGC genes could not be annotated. c Distribution of the annotated BCGC genes in different SEED categories and their taxonomic affiliation. The 40 most enriched SEED categories and their phylum-level affiliations are shown. d Reads mapping to BCGC genes in SEED categories related to extracellular polysaccharide biosynthesis, stress and antibiotic resistance, showing the abundance distribution of these functions across the 101 biofilms

Fig. 5

Novel Acidobacteria lineages. a Distribution of Acidobacteria classes in the biofilms and seawater. Based on 16S miTags, several rarely studied or unclassified groups are enriched in biofilms. b Phylogenetic analysis of Acidobacteria 16S rRNA gene sequences (over 1000 bp) extracted from assembled biofilm metagenomes. 16S rRNA gene sequences that are available for known Acidobacteria classes were downloaded from the NCBI database and used as references and Rokubacteria was used to root the tree. Black dots indicate bootstrap values above 60. These analyses provide further support for novel microbial lineages within ocean biofilms

Fig. 6

Functional potential of the biofilm-forming microbes as revealed by analysis of 479 microbial genomes. a Biosynthetic gene clusters found on the 479 genomes recovered from 101 biofilm metagenomes. Putative product types were assigned by antiSMASH. b Number of CRISPR arrays identified from biofilm-derived genomes. c CRISPR arrays reconstructed from two cyanobacterial genomes and two acidobacterial genomes. White boxes indicate repeats and diamonds indicate spacers. Conserved T, C, and G nucleotides are displayed in yellow, green, and pink, respectively. The putative CRISPR-associated genes are shown with different colors, most of which are RNases. RAMP repeat associated mysterious protein