Metabolic functions of Pseudomonas fluorescens strains from Populus deltoides depend on rhizosphere or endosphere isolation compartment

Abstract

The bacterial microbiota of plants is diverse, with 1000s of operational taxonomic units (OTUs) associated with any individual plant. In this work, we used phenotypic analysis, comparative genomics, and metabolic models to investigate the differences between 19 sequenced Pseudomonas fluorescens strains. These isolates represent a single OTU and were collected from the rhizosphere and endosphere of Populus deltoides. While no traits were exclusive to either endosphere or rhizosphere P. fluorescens isolates, multiple pathways relevant for plant-bacterial interactions are enriched in endosphere isolate genomes. Further, growth phenotypes such as phosphate solubilization, protease activity, denitrification and root growth promotion are biased toward endosphere isolates. Endosphere isolates have significantly more metabolic pathways for plant signaling compounds and an increased metabolic range that includes utilization of energy rich nucleotides and sugars, consistent with endosphere colonization. Rhizosphere P. fluorescens have fewer pathways representative of plant-bacterial interactions but show metabolic bias toward chemical substrates often found in root exudates. This work reveals the diverse functions that may contribute to colonization of the endosphere by bacteria and are enriched among closely related isolates.

Keywords: Populus; endosphere; metabolic modeling; metabolism; microbiome; rhizosphere.

Figure 1

Phylogenetic tree of P. fluorescens isolates. Phylogenetic reconstruction based on the predicted protein sequences of 10 genes: acsA, aroE, dnaE, guaA, gyrB, mutL, ppsA, pyrE, recA, and rpoB. Red highlight, rhizosphere isolate, blue highlight: endosphere isolate. Two common reference isolates of Pseudomonas (Pf0-1 and R124) are shown for comparison. Full tree included in Figure S1. Scale bar is expected substitutions per site. Node labels indicate posterior probability, unlabeled nodes have values >0.99.

Figure 2

Phenotype screening of P. fluorescens isolates. (A) Siderophore production, solubilization of calcium phosphate (CaPO₄), denitrification by anaerobic growth with potassium nitrate, acyl-homoserine lactone (AHL) production, protease activity, and growth promotion of Arabidopsis roots. Gray boxes indicate positive result of assay and white boxes indicate negative result of assay; dark gray indicates high activity for protease assay. Red shading on strain designation indicates rhizosphere isolate, blue shading indicates endosphere isolate (B) Production of indole-3-acetic acid production (IAA) using the Salkowski method. Red bars are results for rhizosphere isolates, and blue bars are results for endosphere isolates (C) Inhibition of four tested strains by lawn assay. Gray boxes indicate inhibition of strain by isolate.

Figure 3

Core/pan genome summary and pathway identification. (A) Core/pan-analysis using ORTHOMCL clustering. There are 3255 genes shared between all isolates (black overlap), with 731 in all rhizosphere and some endosphere isolate genomes, and 52 in all endosphere and some rhizosphere isolate genomes. 3157 genes are shared between some rhizosphere and some endosphere isolate genomes. 268 and 3212 genes are shared between rhizosphere isolates or endosphere isolates only. The zeros indicate that there are no genes in all endosphere isolate genomes and no rhizosphere isolate genomes or vice versa. (B) Summary of manually identified pathways biased toward endosphere isolates. (C) Summary of manually identified pathways biased toward rhizosphere isolates. Gray boxes indicate presence of pathway in organisms.

Figure 4

Metabolic reconstruction summary for Pseudomonas fluorescens isolates. (A) Number of reactions in models grouped by rhizosphere and endosphere. The black circle indicates the number of reactions common to all models, 175 reactions are shared between some endosphere and rhizosphere, and one reaction is unique to rhizosphere while 105 are unique to endosphere isolates. (B) Model summaries for non-core reactions. Each row represents a non-core reaction and each column represents the model from a single isolate.

Figure 5

Compound groups and individual substrates that differentiate endosphere and rhizosphere isolates. (A) Carbon sources are listed vertically (compound group shown), and strains are listed horizontally (R, rhizosphere strains; E, endosphere strains). A gray square indicates metabolic activity in the presence of the substrate. Only differentially used substrates are shown, full table with specific compounds is included in Supplemental File. Highlighting on compound groups indicates significant bias toward rhizosphere (red) or endosphere (blue) isolates. (B) Three compounds that showed highest rhizosphere bias. (C) Five compounds that showed highest endosphere bias. (D) Growth curves for sole carbon sources with observed growth in M9 minimal media. Numbers represent the number of isolates that were shown to grow using the specified compound as the sole carbon source.