Volatile Compounds From Bacillus, Serratia, and Pseudomonas Promote Growth and Alter the Transcriptional Landscape of Solanum tuberosum in a Passively Ventilated Growth System

Abstract

The interaction of an array of volatile organic compounds (VOCs) termed bacterial volatile compounds (BVCs) with plants is now a major area of study under the umbrella of plant-microbe interactions. Many growth systems have been developed to determine the nature of these interactions in vitro. However, each of these systems have their benefits and drawbacks with respect to one another and can greatly influence the end-point interpretation of the BVC effect on plant physiology. To address the need for novel growth systems in BVC-plant interactions, our study investigated the use of a passively ventilated growth system, made possible via Microbox^® growth chambers, to determine the effect of BVCs emitted by six bacterial isolates from the genera Bacillus, Serratia, and Pseudomonas. Solid-phase microextraction GC/MS was utilized to determine the BVC profile of each bacterial isolate when cultured in three different growth media each with varying carbon content. 66 BVCs were identified in total, with alcohols and alkanes being the most abundant. When cultured in tryptic soy broth, all six isolates were capable of producing 2,5-dimethylpyrazine, however BVC emission associated with this media were deemed to have negative effects on plant growth. The two remaining media types, namely Methyl Red-Voges Proskeur (MR-VP) and Murashige and Skoog (M + S), were selected for bacterial growth in co-cultivation experiments with Solanum tuberosum L. cv. 'Golden Wonder.' The BVC emissions of Bacillus and Serratia isolates cultured on MR-VP induced alterations in the transcriptional landscape of potato across all treatments with 956 significantly differentially expressed genes. This study has yielded interesting results which indicate that BVCs may not always broadly upregulate expression of defense genes and this may be due to choice of plant-bacteria co-cultivation apparatus, bacterial growth media and/or strain, or likely, a complex interaction between these factors. The multifactorial complexities of observed effects of BVCs on target organisms, while intensely studied in recent years, need to be further elucidated before the translation of lab to open-field applications can be fully realized.

Keywords: GC/MS; biocontrol; plant growth-promotion; plant-bacteria interactions; transcriptomics; volatile organic compounds.

FIGURE 1

Plants growing in polyurethane foam imbibed with 50 ml M+S media. Bacteria release BVCs from respective growth media and plants are exposed to BVCs within the growth chamber. Gas exchange between the atmosphere and the growth chamber is mediated by the Microbox^® growth chamber technology. Plants are co-cultivated with respective physically separated bacteria for 4 weeks.

FIGURE 2

Growth response of plants exposed to BVCs from isolates growing on solid MR-VP media. Dry weight of plants is measured in (mg), error bars represent confidence interval of the mean 99% (n = 8). Isolates which share a common letter are not significantly different to one another (p > 0.01) according to one-way between groups ANOVA with post hoc Tukey test.

FIGURE 3

Growth response of plants exposed to BVCs from isolates growing on solid M+S media. Dry weight of plants is measured in (mg), error bars represent confidence interval of the mean 99% (n = 8). Isolates which share a common letter are not significantly different to one another (p > 0.01) according to one-way between groups ANOVA with post hoc Tukey test.

FIGURE 4

The effect on plant stem length of BVCs from isolates growing on solid MR-VP media. Stem length of plants is measured in (mm), error bars represent confidence interval of the mean (99%) (n = 8). Isolates which share a common letter are not significantly different to one another (p > 0.01) according to one-way between groups ANOVA with post hoc Tukey test.

FIGURE 5

The effect on plant stem length of BVCs from isolates growing on solid M+S media. Stem length of plants is measured in (mm), error bars represent confidence interval of the mean (99%) (n = 8). Isolates which share a common letter are not significantly different to one another (p > 0.01) according to one-way between groups ANOVA with post hoc Tukey test.

FIGURE 6

Plot of the first two principal components of a PCA generated from the expression data (count data with variance stabilizing transformation from DeSeq2) colored by condition showing that samples in each condition group together.

FIGURE 7

A heatmap of RNA-Seq expression z-scores computed for genes that are significantly differentially expressed Benjamini-Hochberg adjusted p-Value (p < 0.05) in any comparison. The genes (rows) and samples (columns) are clustered using the Pearson Correlation distance and complete linkage hierarchical clustering. The color code shows the row z-score, with a red color indicating higher expression of a gene and blue color indicating lower expression of a gene. This heatmap demonstrates that C (control) and M (control with axenic MR-VP media) groups, and L (LAC2/MR-VP volatile blend) and F (FZB24/MR-VP volatile blend) groups cluster together and that the genes involved have higher expression in the C and M groups compared with the L and F groups.

FIGURE 8

UpSetR plot showing numbers of downregulated differentially expressed genes (DEGs), (p < 0.05) and Log2 fold change (≥2.0). Any DEGs common to plant treatments are shown by connected dots. Cumulative numbers of DEGs are calculated by adding numbers associated with dots in each respective treatment row.

FIGURE 9

UpSetR plot showing numbers of upregulated differentially expressed genes (DEGs), (p < 0.05) and Log2 fold change (≥2.0). Any DEGs common to plant treatments are shown by connected dots. Cumulative numbers of DEGs are calculated by adding numbers associated with dots in each respective treatment row.