Genomic adaptation in the CAZyome and specialised metabolism of the plant-associated Streptomyces violaceusniger clade

Abstract

Streptomycetes are Gram-positive actinobacteria largely represented in the plant root microbiota. The genetic determinants involved in the presence of Streptomyces in the rhizosphere are mostly unknown but can rely on the ability to release phytohormones, degrade plant cell-wall polysaccharides and produce specialised metabolites. Here we sequenced the genome of the rhizospheric and plant defence-stimulating strain Streptomyces sp. AgN23. We found out that it belongs to the soil and plant root dwelling S. violaceusniger clade. The genome annotation of AgN23 revealed the ability of the bacterium to synthesise auxin, a major regulator of plant development, to degrade plant cell wall with a large repertoire of carbohydrate degrading enzymes and to produce antimicrobials (rustmicin, mediomycin, niphimycin, nigericin) and plant bioactive compounds (nigericin, echosides, elaiophylin) through a set of biosynthetic gene clusters. We also found that these genomic features are well-conserved among members of the S. violaceusniger clade. In addition, AgN23 display original events of biosynthetic gene clusters acquisitions and losses which may account for its beneficial effect on plants. Taken together, our work supports the hypothesis that hydrolytic enzymes and specialised metabolites repertoires underpin the interaction of bacteria belonging to the S. violaceusniger clade with plant roots within the rhizosphere. Impact statementStreptomycetes are filamentous Gram-positive bacteria universally found around and within host plant tissues. These actinobacteria have been extensively investigated for their tremendous ability to produce diverse specialised metabolites (e.g., antibiotics). By contrast their impact on host plant physiology is widely neglected. Whether specific lineage of Streptomyces colonise host plant and what are the underlying molecular mechanisms is poorly documented. Here we report a chromosome-scale assembly of AgN23 genome, a Streptomyces sp. strain previously characterised for its ability to activate the plant immune system. This reference sequence enabled us to position AgN23 in the S. violaceusniger clade from which several representatives have been isolated worldwide from the rhizosphere of unrelated plants. Comparative genomic studies suggest that S. violaceusniger spp. produce a prominent CAZyome with expansion of plant cell wall degrading enzymes families and a conserved specialised metabolism acting on host plant physiology and its rhizospheric microbiota. These genomic features may underly S. violaceusniger spp. adaptation to the rhizopsheric niche. Data summaryThe raw reads sequences of AgN23 genome are available at NCBI on the Sequence Read Archive portal for PacBio and MiSeq data (SRR13990229 and SRR14028548 respectively). The Genome assembly is available on the NCBI nucleotide portal under the accession NZ_CP007153.1. This genome sequence was uploaded on the MicroScope platform for genome annotation and analysis (https://mage.genoscope.cns.fr/microscope/home/index.php) [1]. The RNA-seq raw reads are archived in the NCBI Bioproject PRJNA745930. The following eight supplementary tables are included in the online version of this article. Supplementary Information 1: Genomes used in this study. The accession number used from the NCBI portal, name, size, number of contigs as well as the level of completeness of the assembly are indicated. Supplementary Information 2: List of the single copy core genes used by autoMLST to build the phylogenetic tree in Figure 1. O_FIG O_LINKSMALLFIG WIDTH=161 HEIGHT=200 SRC="FIGDIR/small/465742v2_fig1.gif" ALT="Figure 1"> View larger version (38K): org.highwire.dtl.DTLVardef@e15f3org.highwire.dtl.DTLVardef@c34218org.highwire.dtl.DTLVardef@1286956org.highwire.dtl.DTLVardef@1bb5b3_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOFigure 1:C_FLOATNO Multi-locus sequence typing assigned AgN23 to the S. violaceusniger clade. Phylogenetic tree based on the multiple alignment of 85 single-copy homologous genes selected from genomic sequences with AutoMLST. The green node highlights the isolates considered to be from the same species (ANI>95%). The black node highlights the clade formed by isolates with ANI>90% as compared to AgN23. The black squares highlights the eight strains that were used for the BGC conservation study. The green logo indicates plant-isolated strains. Frankia alni ACN14a was used as outgroup, bootstrap=100. C_FIG Supplementary information 3: Annotation of AgN23 full chromosome. For each gene the frame of translation, sequence length and position on the chromosome are indicated. All genes were annotated according to the Microscope platform, see materials and methods. In addition, the expression for each gene is reported in transcripts per million (TPM) based on the the RNA-seq data from three biological replicates. Supplementary Information 4: Genomes having a Mash-based estimated ANI (Average Nucleotide Identity) superior or egal to 80% according to autoMLST. Supplementary Information 5: Prediction of the CAZyme encoding genes using HMMER dbCAN2. The genes are sorted according their CAZy families. For each gene, the begin position on the chromosome, the CAZy category, the annotation, the expression level in transcripts per million (TPM) and the predicted targets of the putative enzymes are described. Supplementary Information 6: Gene identified by antiSMASH in the region containing a biosynthetic gene cluster. Expression levels in transcripts per million (TPM) are indicated for each gene. Annotated central bioynthetic genes are indicated as Y. Those are the ones used for the calculation of mean BGC expression in Table 2. O_TBL View this table: org.highwire.dtl.DTLVardef@1362f24org.highwire.dtl.DTLVardef@50f792org.highwire.dtl.DTLVardef@1ad9681org.highwire.dtl.DTLVardef@1742e96org.highwire.dtl.DTLVardef@9e9b6c_HPS_FORMAT_FIGEXP M_TBL O_FLOATNOTable 2:C_FLOATNO O_TABLECAPTIONantiSMASH annotation of AgN23 chromosomal regions coding for Biosynthetic Gene Clusters. The functional category of each BGCs was determined by antiSMASH. The BGC type, best hit in the MIBiG database as its percentage of similarity to the query are indicated along with the bacterial strain from whom the cluster was described. The expression level of each BGCs was determined by doing the mean of the expression level in Transcripts Per Million (TPM) of the core biosynthetic genes of each BGC from the RNA-seq data (n=3). C_TABLECAPTION C_TBL Supplementary Information 7: Annotation of AgN23 genes putatively involved in biosynthetic pathways for Auxins related phytohomones. Expression levels in transcripts per million (TPM) are indicated for each gene. The genes were detected by blasting reference KEGG sequences for each KEGG ONTOLOGY against AgN23 genes. A cut off of 70% identity and 40% coverage was applied to detect positive hits. These biosynthetic pathways and the KEGG ONTOLOGY are indicated in column F and G. Supplementary Information 8: Inspection of BiG-FAM hits with AgN23 BGCs to identify homologous BGCs found outside the S. violaceusniger clade. BiG-FAM distance higher than 900 were excluded from the analysis. The authors confirm all supporting data, code and protocols have been provided within the article or through supplementary data files.