Biological hydrogen cyanide emission globally impacts the physiology of both HCN-emitting and HCN-perceiving Pseudomonas

ABSTRACT Bacterial volatile compounds have emerged as important chemical messengers between bacteria themselves as well as in their interactions with other organisms. One of the earliest examples of bioactive volatiles emitted by bacteria is hydrogen cyanide (HCN), which was long considered a mere respiratory toxin conferring competitive advantage to cyanide-producing strains. Using cyanide-deficient mutants in two Pseudomonas strains and global transcriptome analysis, we demonstrate that the impact of HCN is much more global than previously thought. We first observed that the lack of cyanogenesis (i.e., the ability to produce HCN) in emitting strains led to massive transcriptome reprogramming affecting diverse traits such as motility and biofilm formation (respectively inhibited vs promoted by HCN), or the production of siderophores, phenazines, and other antimicrobial compounds (repressed by HCN). We then exposed non-cyanogenic strains to biogenically emitted HCN from neighboring cells and observed similar transcriptome modulations and phenotypic changes, suggesting that HCN not only acts endogenously but also exogenously, remotely manipulating important traits involved in competition and virulence, e.g., siderophore production, in other organisms. Cyanogenesis in Pseudomonas has long been known to play a role in both the virulence of opportunistic pathogens and the efficient biocontrol activity of plant-beneficial strains; however, this impact was so far thought to occur solely through the inhibition of respiration. We demonstrate here new ecological roles for a small and fast-diffusing volatile compound, which opens novel avenues in our understanding of and ability to interfere with important processes taking place in pathogenic and beneficial Pseudomonas strains. IMPORTANCE Bacteria communicate by exchanging chemical signals, some of which are volatile and can remotely reach other organisms. HCN was one of the first volatiles discovered to severely impact exposed organisms by inhibiting their respiration. Using HCN-deficient mutants in two Pseudomonas strains, we demonstrate that HCN’s impact goes beyond the sole inhibition of respiration and affects both emitting and receiving bacteria in a global way, modulating their motility, biofilm formation, and production of antimicrobial compounds. Our data suggest that bacteria could use HCN not only to control their own cellular functions, but also to remotely influence the behavior of other bacteria sharing the same environment. Since HCN emission occurs in both clinically and environmentally relevant Pseudomonas, these findings are important to better understand or even modulate the expression of bacterial traits involved in both virulence of opportunistic pathogens and in biocontrol efficacy of plant-beneficial strains.

x-axis as time in minutes and the y-axis as peak intensity (arbitrary units).b, MS profile for peaks (i), (ii) and (iii).Peak (i) = 1 or 2-hydroxyphenazine, peak, (ii) = phenazine-1-carboxylic acid (PCA) and peak (iii) = 2-hydroxyphenazine carboxylic acid.The wild type strains and the respective HCN-deficient mutants were grown on LB medium for 48h.The volatile compounds they emitted were collected using a closed-loop stripping method and analysed using GC/MS.Dots represent mass features.Grey color represents mass features that were not statistically different between the wild types and their respective HCN-deficient mutants.A Student's T-test with a P-value cutoff set at 0.05 was used to find statistically different mass features between the wild types and the HCN-deficient mutants.On the x-axis, mass features are plotted against the -log10(raw P value) on the y-axis.Mass features present in LB medium controls were removed.Data represent mass features detected in three independent biological replicates.Note that HCN cannot be detected by standard GC/MS protocol and is therefore not depicted in the graph.analysis of transcriptomic data of the wild type strain (wt), the hcn mutant strain (Δhcn) and the hcn mutant strain exposed to HCN emitted by the wild type strain (EΔhcn) at time points 1, 2 and 3. T1; early log phase, T2; transition phase (exponential to stationary growth), T3; early stationary growth phase.b, Dysregulated genes in wt vs Δhcn, wt vs EΔhcn and Δhcn vs EΔhcn comparisons.Log2 (fold change) threshold was set to 1 and q-value to 0.05.c, Dysregulated genes between different sets of strains (wt vs. ∆hcn in purple, wt vs. E∆hcn in grey and ∆hcn vs. E∆hcn in green) were compiled into Venn diagrams for the three timepoints.Numbers indicate unique vs. shared dysregulated genes for the different comparisons, with a total of 42 dysregulated genes at timepoint 1, 253 at timepoint 2 and 414 at timepoint 3. d, COG category analysis of dysregulated genes in Δhcn compared to wt as shown in Figure 2c, with supplementary data from the comparison between wt and EΔhcn to highlight recovered vs. nonrecovered gene expression in EΔhcn.A gene was considered recovered if the log2 (fold change) expression value in 'wt vs EΔhcn' comparison was less than 50% of its value in 'wt vs Δhcn' comparison.

Figure S1 .
Figure S1.Hydrogen cyanide quantification.Wild type strains of Pseudomonas chlororaphis R47 and Pseudomonas putida R32 were inoculated in 6-well plates (two wells on the left and two on the right while the two centre wells were filled with NaOH to capture HCN) for 7h at 30 ˚C with shaking.After 7h, NaOH was collected and HCN was quantified as described in the experimental procedure section.Bars represent the average HCN concentration in 1 ml bacterial culture (OD600 = 1) using three independent experiments.Error bars represent standard errors.

Figure S2 .
Figure S2.HCN represses the production of phenazines.a, Phenazine quantification assay using an HPLC-UV method as described in the experimental methods section.Images represent Pseudomonas chlororaphis R47 wt and R47 Δhcn in liquid culture in King's B medium after 24h growth at 30 ˚C with shaking.Chromatograms represent the HPLC-UV profiles with the

Figure S3 .
Figure S3.hcnA expression starts at the early exponential growth phase in Pseudomonas chlororaphis R47.qPCR analysis of the expression of the hcnA gene from the HCN biosynthetic operon.rpoD was used as reference gene.RNA samples and time points used for this analysis were the same as those used for transcriptomic analysis (Figure2a).Bars represent the mean normalized expression compared to rpoD expression.Bars represent the average from three independent experiments for each time point.T1; early log phase, T2; transition phase (exponential to stationary), T3; early stationary phase.

Figure S4 .
Figure S4.Volatile-mediated exposure assay in 6-well plates.The two wells on the left and on the right, respectively, were inoculated with the Pseudomonas chlororaphis R47 wild type strain.Filter papers imbibed with copper (II) ethyl acetate and 4,4methylenebis(dimethylaniline) solution for HCN detection were placed in the two middle wells.The cultures were incubated at 30 ˚C for 18h with shaking before capturing the image.The blue color in the middle two wells reveals the presence of HCN.

Figure S6 .
Figure S6.Exposure to exogenous HCN from Pseudomonas chlororaphis R47 wild type leads to global transcriptomic reprogramming in the mutant.a, Principal component

Figure S7 .
Figure S7.Split plate exposure experimental setup to measure motility.The left side of the split petri dish was inoculated with three drops of bacterial suspension 6h before the inoculation on the right side of the plate as described in experimental procedures.The left side was filled with LB agar medium while the right side was filled with low agar medium.The plate on the left represents the wild type strain exposed to the wild type strain, the plate in the middle represents the HCN-deficient mutant exposed to the HCN-deficient mutant, while the plate on the right shows the HCN-deficient mutant exposed to the wild type.

Figure
Figure S8.24-well exposure experimental setup to measure biofilm formation.HCN producing bacteria were inoculated in the wells colored in purple, while the HCN receiving bacteria were kept in green wells.Grey wells indicate uninoculated wells.Wells marked by * were sampled for biofilm analysis.